Nothing Special   »   [go: up one dir, main page]
Next Article in Journal
How Does Herbicide Resistance Change Farmer’s Weed Management Decisions? Evidence from the Roundup Ready Experiment
Previous Article in Journal
Adsorption Properties of Fishbone and Fishbone-Derived Biochar for Cadmium in Aqueous Solution
Previous Article in Special Issue
Innovations in Agriculture for Sustainable Agro-Systems
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 14
Issue 11
10.3390/agronomy14112719
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Crops Disease Detection, from Leaves to Field: What We Can Expect from Artificial Intelligence

by
Youssef Lebrini
1 and
Alicia Ayerdi Gotor
2,*
1
Institut Polytechnique UniLaSalle, UPJV, B2R (GeNumEr), U2R 7511, 19 Rue Pierre Waguet, BP 30313, 60026 Beauvais, France
2
Institut Polytechnique UniLaSalle, AGHYLE, UP 2018.C101, UniLaSalle, 19 Rue Pierre Waguet, BP 30313, 60026 Beauvais, France
*
Author to whom correspondence should be addressed.
Agronomy 2024, 14(11), 2719; https://doi.org/10.3390/agronomy14112719
Submission received: 25 September 2024 / Revised: 9 November 2024 / Accepted: 14 November 2024 / Published: 18 November 2024
(This article belongs to the Special Issue Innovations in Agriculture for Sustainable Agro-Systems)
Browse Figures
Graphical abstract
"> Figure 1
<p>Temporal dynamics of research trends in crop disease detection topics.</p> "> Figure 2
<p>Flowchart of the methodology for collecting data about plant disease identification using artificial intelligence and image data.</p> "> Figure 3
<p>Thematic clustering of the topics identified from the keywords as well as their temporal evolution based on data collected from 2000 to 2024. The node size of a topic is proportionate to its relative frequency.</p> "> Figure 4
<p>Thematic clustering of the topics identified from the keywords based on data collected from 2000 to 2024. The node size of a topic is proportionate to its relative frequency.</p> "> Figure 5
<p>Sankey diagram analysis of global research trends in crop disease detection using artificial intelligence and image processing for the period between 2000 and 2024.</p> ">
Versions Notes

Abstract

:
Agriculture is dealing with numerous challenges of increasing production while decreasing the amount of chemicals and fertilizers used. The intensification of agricultural systems has been linked to the use of these inputs which nevertheless have negative consequences for the environment. With new technologies, and progress in precision agriculture associated with decision support systems for farmers, the objective is to optimize their use. This review focused on the progress made in utilizing machine learning and remote sensing to detect and identify crop diseases that may help farmers to (i) choose the right treatment, the most adapted to a particular disease, (ii) treat diseases at early stages of contamination, and (iii) maybe in the future treat only where it is necessary or economically profitable. The state of the art has shown significant progress in the detection and identification of disease at the leaf scale in most of the cultivated species, but less progress is done in the detection of diseases at the field scale where the environment is complex and applied only in some field crops.

Graphical Abstract">

Graphical Abstract

1. Introduction

The agriculture sector plays a crucial role in achieving the United Nations’ goals of feeding the society by producing food to feed the growing population and providing livelihoods for millions of people [1,2]. Achieving these goals needs increasing crops yields by modern technologies but should encourage sustainable land use and practices to protect the environment, such as reducing chemical spraying.

1.1. Impact of Crop Diseases on Production and Agricultural Sector

Crop diseases have a significant impact on crop production [3]. Plant diseases are caused by pathogens such as viruses, bacteria, fungi, insects, and nematodes which may act alone or in combination, whose effect depends on the species or varieties encountered, the plant stage of development, the meteorological conditions [4], and the agroecological context [5,6] known as the disease triangle (plant–pathogen–environment) [7]. These different pathogens may cause significant reductions in crop yields, reduce the visual aesthetic aspect of the produce [8,9,10], alter the organoleptic quality, or produce toxic molecules [11,12,13]. The quantity or quality reduction generates direct losses for farmers with less amounts produced, or indirectly with products that cannot be sold (such as products presenting a visual damage), or not being able to sell at the highest price or having products that do not reach the normal standard size. Lastly, some pathogens produce toxic contaminants rendering the product inedible. This leads to economic losses for farmers which affect the food chain and at the end may generate a problem of food security for the population [14].
If farmers want to control pests, they will face increased costs as they spend more money on pesticides and other chemical treatments to control crop diseases, further reducing their profits [14]. Overuse of chemical treatments can lead to the development of resistance in pathogens, making it difficult or impossible to control crop diseases [15]. Furthermore, the overuse of phytochemicals may increase the risk of finding chemical residues on food, leading to food contamination and inducing a cumulative risk [16].
Lastly, farmers exposed to these chemical products may develop diseases themselves [17,18]. To mitigate the impact of crop diseases on production and the agricultural sector, it is important to implement effective disease management practices such as using disease-resistant varieties, crop rotation, and cultural practices that improve soil health, as well as Integrated Pest Management (IPM) techniques, which involve using a combination of methods such as biological, cultural, and chemical controls. Additionally, the research and development of new technologies such as genetic engineering and biotechnology can also help to develop disease-resistant crops.

1.2. Data Acquisition to Assess Crop Diseases

The rapid advancement of image and data collection is transforming how we assess crop disease. The reduced cost of unmanned aerial vehicles (UAVs) and the increased density of satellites have accelerated their use for agricultural purposes. Remote sensing with satellites and camera-equipped UAVs allows large-scale detection and mapping, enabling early intervention and targeted resource allocation to minimize yield loss [19,20,21,22]. If the data are either acquired on a large scale or with proximal tools like handheld camera supported by a stick, an armor, or ground vehicles, or just with a smartphone, the information collected will depend on the detector integrated. Those data could be image based or non-image-based approaches. Using imagery data has many advantages, including the ability to cover large areas quickly, the ability to detect diseases at an early stage, and the ability to monitor disease progression over time [23]. Early disease detection is crucial for effective disease management as it allows farmers to act before the disease spreads and generates a bigger yield loss [24,25]. Furthermore, the use of imagery data can also provide a more accurate assessment of disease incidence and severity, which can help in decision-making and resource allocation [24].
There are several image approaches using the following: RGB, multispectral, hyperspectral, thermal, and fluorescence imaging [26]. RGB imaging uses only the red, green, and blue wavelengths and has been used to detect disease presence in leaves but also in fruits [27,28]; multispectral imaging [21,29] uses a discontinuous but larger panel of wavelengths and hyperspectral [30] imaging uses a continuous and larger panel of wavelength. Both methods have been used for detecting diseases as they provide more information compared to RGB imaging. Lastly, fluorescence and thermal imaging provide complementary information to the three precedent ones as indicators of plant metabolism and biotic and abiotic stress [31,32]. Methodologies that are non-imaging based have also been developed to detect diseases in crops such as fluorescence, visible, and infra-red spectrometry [23,26,33,34].
Sensors and Internet of Things (IoT) devices are also used at the field scale to measure factors such as temperature, humidity, and light intensity that can affect disease development [35]. Data collected by these sensors can be used to predict disease risk and to trigger automatic responses such as applying pesticides [35,36].

1.3. Data Analysis to Detect and Identify Crop Diseases

Spectral data analysis, which examines light reflected from crops, plays a crucial role in early disease detection by identifying subtle changes in leaf color or texture [37]. These data, gathered from various sources, including ground-based sensors, aerial platforms, or satellites, allows for a comprehensive monitoring of diseases, often the first signs of disease. Machine learning algorithms analyze these image data to recognize patterns and track trends in disease development. This approach not only aids in current disease management but also facilitates the prediction of future outbreaks and even builds models to optimize disease management strategies [38].
Using vegetation indices, such as the Normalized Difference Vegetation Index (NDVI), which reflects plant health, further enhances the ability to monitor and predict crop health [39]. By employing these indices, researchers and farmers can develop advanced predictions and management strategies that respond dynamically to crop health needs [20]. However, implementing these methods requires significant expertise and resources, with each approach needing adaptation based on specific crops, diseases, and regional factors.
Machine learning (ML) and artificial intelligence (AI) techniques deployed to analyze data collected by remote sensors is providing new insights in the field of agronomy by its powerful learning algorithms [40]. ML and AI are deployed for different agronomical purposes such as patterns identification and trends in disease development, predicting disease risk, and to develop models that can be used to optimize disease management strategies [40,41]. Such technologies have applications across various agronomic needs, making them versatile tools for enhancing crop health and yield management [42].
Despite their benefits, it is important to note that the implementation of these technologies is not a one-size-fits-all solution. The selection and use of these tools must be tailored to suit the specific requirements of the crop, disease, and region involved [43]. Practical considerations, such as the costs of technology, the availability of trained personnel, and local infrastructure, are crucial to ensuring these technologies are used effectively.
Tools for assessing crop disease presence have many advantages and can be a valuable resource for farmers, researchers, and policymakers [19]. However, it is important to note that these methods should be complemented with other information, such as weather data, ground-based sensor data, and agronomic knowledge, to create a more comprehensive view of crop health, disease progression, and treatment needs. Additionally, an awareness of potential dataset limitations and biases in analytical methods is essential to achieving accurate, reliable outcomes in disease detection and management.
The main goal of this review is to assess recent advancements in detecting and identifying crop diseases through remote sensing and machine learning, spanning from leaf-scale to field-scale, the various data collection vectors, data processing techniques, and exploring practical applications for optimizing chemical use by enabling targeted treatment only where necessary.

2. Methodology

2.1. Crop Disease Detection Systematic Search Approach (SSA)

To gain a comprehensive understanding of the current research directions and to identify potential gaps in our knowledge, a two-pronged approach was employed. Firstly, a bibliometric analysis was conducted using the R bibliometrix package (https://www.bibliometrix.org/home/index.php (accessed on 25 September 2024)) and VOSviewer software, version 1.6.20 [44,45]. This analysis leveraged data from the Scopus database to uncover emerging trends and frequently explored topics within the field of crop disease detection.
The creation of the network maps from VOSViewer was based on keyword co-occurrences, aimed at identifying the structural connections within the research landscape. Author keywords were selected as the unit of analysis, with the full counting method applied to provide equal weight to each co-occurrence. This approach allowed a detailed examination of keyword relationships, where node proximity on the map indicated term relatedness, while clusters suggested distinct research themes or subfields.
Regarding the precision of these clusters, the methodology ensures that identified clusters offer an initial framework for interpreting core topics within the field. However, due to limitations inherent in automated clustering, such as potential overlaps between thematic areas or the absence of nuanced context, further analysis was necessary. Consequently, a more detailed examination of individual papers was conducted. This secondary analysis involved a fine-grained review of selected papers based on topics in Figure 1, allowing for a more accurate interpretation of research themes, emerging trends, and the intellectual structure of the field.
This combined approach provides a more nuanced perspective on the research landscape, highlighting not only the trending topics but also the prevalent methodologies in plant disease detection based on artificial intelligence techniques.
Figure 1 unveils the evolving research trends in crop disease detection over the past decade (between 2012 and 2024). The term frequency on the y-axis indicates the prevalence of specific research topics within the analyzed body of literature. As the x-axis progresses through the years, a clear upward trajectory emerges, signifying a growing global focus on combating plant diseases.
Unsurprisingly, “plant disease detection” itself leads greatest throughout the ten-year period, highlighting its enduring importance. However, a closer look reveals a fascinating shift towards artificial intelligence-based (AI) techniques. The surging popularity of terms like “deep learning” and “convolutional neural networks” underscores the dominance of AI in contemporary research. The substantial rise in these terms, particularly when compared to established methods like “support vector machines”, suggests a paradigm shift towards AI-powered automated detection systems.
Intriguingly, Figure 1 also unveils a surge in interest for some exciting new research directions. The rise of “spectroscopy” signifies a growing interest in analyzing the unique spectral signatures of plants to identify diseases. Similarly, the increasing prevalence of “phenotyping” indicates the exploration of how observable plant characteristics like growth rate or size can be used to detect subtle health changes indicative of disease. Finally, the significant upward trend for “remote sensing” highlights the potential of using satellites or drones to monitor vast crop areas and detect potential disease outbreaks on the field scale.
The following analyses were then conducted in the literature from January 2012 to May 2024, considering that prior to this period, the research done were preliminary works and too scarce to be significant.

2.2. Data Retrieval and Extraction

Scopus was used to select metadata information for this study, as it is an authoritative database offering extensive coverage of the literature and is widely recognized for citations. Additionally, Scopus enables the exporting of up to 2000 publications at once, making it a commonly used reference database for several bibliometric analyses. To focus on the relevance of studies and increase consistency, a set of inclusion and exclusion criteria was established [46]. These criteria are essential to provide clear guidelines for this review and to control its scope and size. The inclusion and exclusion criteria of this review are provided within Table 1.
The inclusion and exclusion criteria are designed to ensure that the studies included in the review are relevant to the research question and of high quality. The inclusion criteria were data of a publication between 2000–2024, the peer-reviewed and grey literature, written in English, sourced from Scopus and Google Scholar, and focused on crop disease detection based on artificial intelligence and image processing. The inclusion criteria were defined to select studies that meet the specific requirements essential for this review and relevant to the research question.
On the other hand, the exclusion criteria were established to filter out studies that do not meet these requirements. Studies published before 2000, from websites and project pages, written in languages other than English, sourced from other search engines, and employing other crop disease biology-based detection methods such as the following: detecting the spores without symptoms (PCR) [47] or based on the thermal needs of the pest reproduction cycle (meteorological) [48] were excluded. By applying these criteria, the review ensures that the included studies are both relevant and of high quality, thereby enhancing the credibility and utility of the review.
Overall, the carefully established inclusion and exclusion criteria help maintain the focus and quality of the review, ensuring that the studies considered are pertinent to the research objectives and meet high standards. This structured approach not only streamlines the review process but also increases the reliability and relevance of the findings, ultimately contributing to a more robust and impactful study.
The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines was used to ensure a transparent and standardized approach in conducting this review. Specifically, we followed the PRISMA checklist to structure our methodology, reporting, and presentation of results. The guidelines helped us in ensuring the rigor and reproducibility of our review process [49]. In this review the Preferred Reporting Items for Reviews and Meta-Analyses (PRISMA) method was applied (Figure 2).
The systematic exploration of crop disease detection, sensing technologies, artificial intelligence (AI), and imaging techniques began with an initial search on SCOPUS, yielding a corpus of 1409 relevant papers. A first trial based on the title and the abstract has eliminated 36 documents out of the topic. For example, we considered, out of the scope of this review, studies where the goal was to predict the presence of the disease based on meteorological conditions combined with agronomical parameters like Kundu et al. [50] or Nie et al. [51]. The cleaned total corpus was used to make the trends, but then to extract information, a second trial was performed to eliminate reviews (45) with not enough data for our research and proceedings papers (737) which were not accessible, then other studies were eliminated because their main research objective was to improve algorithms without mentioning the crop or the disease clearly (56). Eight hundred and seventy-four (874) papers were excluded, leaving a focused set of 535 papers for further analysis. This subset formed the foundation for a structured literature review, meticulously organized into five pivotal topics. The first topic, “Acquisition Methods and Data Sources”, delved into the varied methodologies and sources employed in gathering data for crop disease detection via sensing technologies and AI. The second area, “Study Scale”, investigated the spectrum of scales adopted in research endeavors, encompassing laboratory experiments, field trials, and remote-sensing applications. “Algorithm Used” scrutinized the diverse array of AI algorithms implemented for disease detection in plants, encompassing machine learning, deep learning, and other computational techniques. Additionally, “Studied Disease” categorically explored the types of diseases scrutinized in the literature, shedding light on their characteristics and nuances. Lastly, the topic “Crop Type” delved into the specific plant species or crop types targeted within the reviewed studies. The meticulous organization into these five thematic domains facilitated a comprehensive analysis, offering insights into trends, patterns, and advancements within the realm of crop disease detection. This approach systematically delineated diverse approaches, highlighted emerging applications, identified research gaps, and paved the way for a holistic synthesis of the current state of research in this field, thereby providing a valuable framework for further investigation and scientific exploration.
Several working scales have been identified during the literature analysis, the scale report, the part of the crop that has been analyzed, and the environment background where data have been collected. Firstly, there are terms which only consider the data treating a single part of the crop, taking the data out of the field on a neutral or uniform background: “Fruit”, “Leaf”, “Panicle”, “Pulp”, and “Stem”. Secondly, there are data taken on a single plant called “Plant” or “Tree” with a more or less uniform background but without mixing several plants together. Thirdly, there is a mixture of parts of the plants with nearly no place to have a background: “Fruits”, “Leaves”, “Pods”, and “Spikes”. Four correspond to several plants together, with no place to have the background close to a crop field but the image is too close to catch the variability at the field level, which have been called “Plants”, and the last scale represents the data collection of the field crop called “Field” including all the plants parts, the background, and the variability within the field. There could also be a mixture of two plant parts like fruits and leaves together, which has been indicated with an “&”.

3. Results

3.1. Bibliometric Analysis

Figure 3 presents a thematic clustering visualization, that unveils the evolving landscape of crop disease detection research between 2000 and 2024. The information is gleaned from keywords extracted from titles, abstracts, and keywords of academic publications within the SCOPUS database. The size of each node in the cluster directly corresponds to the relative frequency of its associated keyword.
The clustering map predominantly shows topics from 2018 to 2024 because these years reflect a period of intensified research and significant advancements in crop disease detection through remote sensing and machine learning. Although the dataset spans from 2000 to 2024, earlier years had comparatively fewer publications or lower keyword frequencies, making them less prominent in the visualization. The clustering method highlights periods with dense, high-frequency keyword occurrences, hence the concentration on recent years where techniques like convolutional neural networks (CNNs) and deep learning became widely adopted and transformative in this field.
At the center of the clustering map, a prominent cluster emerges around key terms, including “plant disease”, “convolutional neural networks (CNNs)”, “deep learning”, and “feature extraction”. This central concentration underscores the dominance of CNNs and deep learning as dominant approaches for extracting diagnostically relevant features from imagery in the field of crop disease detection research.
Shifting to the left cluster, we encounter a thematic cluster dominated by terms like “image processing methods”, “image enhancement”, “pre-processing”, and “principal component analysis (PCA)”. These techniques likely serve as the preparatory steps before feeding images into CNNs for disease detection purposes.
On the other hand, the right cluster revolves around keywords such as “classification models”, “spectroscopy”, “plant pathology”, and “hyperspectral imaging”. This suggests that researchers are increasingly exploring spectroscopic techniques, in conjunction with classification models, as a powerful tool for detecting plant diseases.
Further exploration reveals a cluster positioned at the bottom right, centering on “IoT”, “machine learning”, “remote sensing”, and “precision agriculture”. This cluster signifies a fascinating trend: the incorporation of these emerging technologies into crop disease detection within the agricultural sector. This integration has the potential to revolutionize how we monitor and manage crop health across vast fields.
Finally, the placement of terms along the timeline provides valuable insights into their emergence and prominence over time. For instance, “k-means clustering” and “support vector machines (SVM)” appear on the left side of the map, hinting at their prevalence in earlier research. In stark contrast, terms like “deep learning” and “random forests” appear on the right, indicating a significant rise in their use in recent years. This visual representation underscores the dynamic nature of plant disease detection research, constantly adapting and embracing new advancements in the field.
To provide further and more in-depth insights into how the research in crop disease detection evolved, text mining was conducted to reveal prominent topics in the 535 sampled papers (Figure 4). The expressed topics that the papers claim to cover (i.e., keywords) fall in five main clusters: (1) the red cluster on plant disease detection methods; (2) the magenta cluster on the type of algorithms used for crop disease detection; (3) the yellow cluster of crop type where disease detection was performed; (4) the blue cluster for the dataset and learning techniques and (5) the green cluster for the study scale. These clusters, while informative, are based on co-occurrence patterns and inherently feature some overlap between keywords. For example, certain keywords may appear within a cluster even if their relevance is more closely aligned with another theme. This overlap reflects the interconnected nature of research areas and underscores the complexity of categorizing topics strictly by co-occurrence. Nonetheless, these clusters provide a structured view of the field, showing both dominant themes and how various aspects of crop disease detection research intersect and evolve. However, a more refined analysis is presented in the following sections, where results are obtained from a detailed data analysis of the selected papers.
The Sankey diagram in Figure 5 offers a compelling snapshot of the research landscape in crop disease detection using artificial intelligence and image processing. It visually depicts the interplay between the most active countries contributing to this field, the prevalent research themes, and the leading journals publishing this work. Countries like India, China, and the USA stand out on the left axis, indicating a strong research presence in this domain. The center axis showcases the core keywords, with “deep learning”, “machine learning”, and “plant disease detection” taking center stage. This reaffirms the dominance of AI techniques in contemporary research in this field. The width of the connecting lines between these keywords and countries like India suggests a significant focus on these AI-powered approaches in these regions.
The Sankey diagram provides a comprehensive visualization of the global research landscape in agricultural technology, with a particular focus on plant disease detection and the application of AI methods.
The rightmost column of the diagram lists prominent journals that serve as outlets for this research topic. The significant presence of journals such as “Computers and Electronics in Agriculture” and “IEEE Access” highlights a notable intersection between agricultural research and technological or engineering disciplines, underscoring the interdisciplinary nature of this research domain. The frequent publication in these outlets reflects the global research community’s commitment to cross-disciplinary approaches, integrating innovations in both biosciences and computational sciences to address complex challenges in agriculture. The thickness of the lines connecting these journals to keywords like “plant disease detection” and “deep learning” implies a strong emphasis on these specific image analysis techniques within the published works. This co-occurrence analysis provides valuable insights into the current trends and focus areas within this field.
The central column delineates key research themes and methodologies prevalent in this field of research. “plant disease detection” is identified as a major theme, reflecting the pressing need for solutions to mitigate crop loss due to diseases. Methodologies applied in this research area include “deep learning”, “machine learning”, “Convolutional Neural Networks (CNNs)”, and “image processing”. These connections illustrate that advanced AI and image-processing techniques are widely used across countries. The inclusion of various AI-based approaches indicates a trend towards integrating sophisticated machine learning models to address agricultural challenges, particularly in plant health monitoring and automated disease detection.
Keywords such as “plant disease detection” and “deep learning” exhibit connections to multiple countries and journals, underscoring their central role in agricultural research aimed at enhancing food security and crop resilience. The choice of publication venues demonstrates the importance of interdisciplinary journals that bridge agriculture, engineering, and computer science, facilitating the dissemination of innovative technologies applicable to agriculture.
The leftmost column lists the countries actively contributing to this field. The diversity in country participation underscores the international scope of research in agricultural technology.
By identifying the leading research hubs, the dominant themes, and the preferred publication channels, this Sankey diagram serves as a springboard for further exploration of specific research areas within the realm of AI-powered crop disease detection. The Sankey diagram allows researchers to identify potential collaborators from active countries, delve deeper into specific AI techniques highlighted by the keywords, and target relevant journals for their own research contributions.
In the recent publications included in our analysis, there is a notable prevalence of the term “plant” rather than “crop”. This observation can be attributed to the fact that many of the authors contributing to this body of research come from non-agronomic institutions. These researchers often publish their findings in journals that are not specifically focused on agricultural sciences. As a result, the terminology they use tends to emphasize a broader biological context, favoring “plant” over the more specific agricultural term “crop”. This trend is also reflected in Figure 3 and Figure 4, where the prominence of the word “plant” indicates a shift in focus toward a more general understanding of plant health and disease, rather than a focus solely on agricultural production.

3.2. Analysis of Disease in the Literature

The extraction of the data retrieved in this systematic review based on the research and review articles in review is summarized in Table 2, where we identified the crops concerned by the studies, the diseases detected or identified, and the study scale.

3.2.1. Study Scale

Based on the objectives and aims of each study, study scales are different. Most manuscripts retrieved have, as the main study scale, a single leaf (Table 2) which has been taken out of the plant to be placed in a neutral background that will be eliminated easily at the first image treatment avoiding any interference with data acquisition. In addition, most of the documents retrieved used an online database to find leaves that were already annotated (diseased and healthy), Plant Village (https://datasets.activeloop.ai/docs/ml/datasets/plantvillage-dataset/ accessed on 9 September 2024) being the most common. This joins the precedent point, alerting us about how the disease detection subject is mainly treated from an algorithmics or computational vision, with few agronomical points of view, since annotated photos without an environmental context is easier to treat. There are a few studies where photos were taken on an entire plant where several leaves, and stem, or fruits were included. Lastly, some research were based on UAV [52,60,75] or satellite [59,105] data acquisition, considering several plants at the same time with a complex environment that includes light/shadowing of some plants, ground presence, different maturity stages of plants, several layers of leaves with a differentiate disease progression, or the presence of weeds. These complex environments are difficult to treat, and the pre-treatment of the image is longer and difficult.

3.2.2. Crop and Disease

The most studied crops were tomato [146], rice [130], wheat [20] and maize [90] (Table 2), where other crops have been neglected or present little research. Another trend is that quite a lot of documents (not reported in Table 2) do not focus the research on a particular crop but were looking for the identification of one or several diseases [168,169,170], and that they do not care about the crop nor the disease and there is only research on distinguishing health and diseased leaves [21]. The common characteristic of those manuscripts is that the work focused on public photos available in databases like the Plant Village. In nearly all the cases, the analysis and the identification are made between the diseased leaf and a healthy leaf. At the level of UAV [20,171] and satellite acquisition, studies were able not only to identify the disease but also started to indicate the disease pressure in the field and others started to analyze the disease severity [158,172]. Most of the studies focused on leaves with fungal diseases, but a few also studied stems [109] or fruits [82,88] diseases and others have observed the impact of insects [156] or viruses [82] on the leaves. Major crops have been studied and regarding the literature, no disease in any crop has had a particular difficulty in being identified in a particular organ when detected outside the field with a proximal data acquisition using high quality images. When it comes to the detection of the diseases in the field, the number of crops evaluated is much reduced and the results are not advanced enough when acquired with either UAVs or with satellites. First studied in a large scale, they can detect the disease when it is largely expanded and can evaluate a disease severity level but they are not accurate enough to develop a decision support system based on those acquisitions.

3.2.3. Acquisition Method and Data Source

The efficiency of artificial intelligence methodologies in plant disease detection axes on the variety of data sources is employed. A selection of these sources is guided by the research objectives and the spatial scale of the targeted disease. For instance, high-resolution ground-based imagery facilitates the detection of early disease stages in specific plant types [162]. The detailed information captured allows for an accurate examination of variations in a plant’s spectral reflectance, potentially revealing disease signatures before visible symptoms manifest [155,165].
Conversely, monitoring vast fields necessitates a broader approach. In such cases, multispectral data acquired by UAVs proves advantageous [19]. These aerial platforms can be equipped with sensors that gather information beyond the visible spectrum, providing valuable insights into plant health that are imperceptible to the human eye. For example, near-infrared sensors can detect subtle changes in plant vigor, while thermal sensors can identify areas with abnormal temperature fluctuations, which are potentially indicative of disease presence [38,162,164].
The incorporation of data from supplementary sources can significantly enhance model performance. Data from weather stations and strategically placed soil sensors can contribute critical details regarding environmental factors like temperature, humidity, and nutrient levels, all of which significantly influence disease development and spread [85]. By strategically combining data from these diverse sources, plant disease detection can achieve a comprehensive and refined understanding of crop health, enabling earlier interventions and improved disease management strategies.

3.2.4. Algorithm Used and Annotation Tools

Further data processing steps are usually needed to make data ready for training. In a deep learning architecture, annotations or image labelling is a major step to supervise the algorithm and provide some regions of interest. These regions of interest generally are the objects and patterns to be detected in images. Table 3 provides a summary of the most used tools for labelling images.
These advantages and constraints can vary based on the specific requirements of the project and the preferences of the annotators. Some of these constraints may not be significant for certain research, while others may be deal breakers. Table 3 summarizes frequently used tools for image annotation.
Annotation tools exhibit significant heterogeneity in design and functionality to address diverse research requirements. User interfaces can be graphical (GUI), web-based, or command-line driven. Platform compatibility varies, with some operating tools being system-specific or requiring additional software libraries. Additionally, annotation formats differ, with support ranging from the widely used PASCAL VOC XML to COCO JSON [173]. Feature sets further distinguish these tools, with some offering collaborative annotation, automation capabilities, and integrated data management, while others prioritize the core annotation process. Finally, cost models diverge, with open-source options like LabelImg freely available and commercially developed tools like RectLabel requiring licensing fees.
Only a few of the references retrieved have used one of these annotation tools, for example [170], where most studies prefer to use already annotated images like those found in the Plant Village database. Labelling is a tedious step which seems to have not been largely combined with research developing the algorithms for disease detection.
When choosing an annotation tool, it is important to consider the specific requirements of the project, such as the format of the data, the type of annotation, and the preferences of the annotators.
Table 4 shows the use frequency of various algorithms used for plant disease prediction. Convolutional Neural Networks (CNNs) are the most frequent algorithm, cited in 37% of the references, followed by Generative Adversarial Networks (GANs) at 18%.
This suggests that CNNs and GANs are popular choices for researchers in this field. CNNs are a type of deep learning algorithm that have been very successful in image recognition tasks. They are well suited for analyzing images of plant leaves to detect diseases. GANs are a type of generative model that can be used to create synthetic data. This can be useful for training CNNs when there is a limited amount of real-world data available.
Other algorithms listed in the table include Support Vector Machines (SVMs) and VGG models, both at 16% and 13%, respectively. These are all machine learning algorithms that can be used for classification tasks. In the context of plant disease prediction, they can be used to classify images of plant leaves as healthy or diseased
Results also show the use of various deep learning object detection models, including R-CNN, Mask R-CNN, Faster R-CNN, and YOLO. These models are designed to detect and localize objects in images. In the context of plant disease prediction, they could be used to identify the specific regions of a plant leaf that are infected with disease.
At the field level, high-resolution images captured by handheld devices are analyzed using models like CNNs, known for their robust feature extraction in detecting specific leaf diseases, such as tomato blight or grape mildew [185,186]. Transfer learning, which adapts pre-trained CNNs to specific agricultural datasets, enhances detection efficiency. Additionally, Support Vector Machines (SVMs) and Random Forests (RF) offer reliable classification based on handcrafted features (e.g., color, texture), ideal for smaller datasets and simpler disease identification tasks. These models make field-level disease detection accessible even with limited computational resources.
Object detection models, such as YOLO and Faster R-CNN, localize disease patches for targeted pesticide application, while vegetation indices (e.g., NDVI, EVI) reveal early stress markers. Satellite-based detection, often using Deep Convolutional Networks (DCNs) and temporal Recurrent Neural Networks (RNNs), monitors regional disease trends with spectral and temporal data. GANs enhance low-resolution satellite data to improve detection precision and anomaly detection with vegetation indices that helps track large-scale disease outbreaks. Together, these ML and DL applications facilitate rapid, scalable, and data-driven disease management in agriculture.
Despite its advantages for large-scale monitoring, the use of satellite data in crop disease detection remains limited, particularly for detecting early-stage symptoms at the leaf scale. Satellite imagery, even at high resolution, struggles to capture fine-grained details, making it difficult to identify specific symptoms that are evident in close-up images, such as small lesions or discolorations on individual leaves. These leaf-level symptoms, crucial for early detection and intervention, are often imperceptible from space due to resolution constraints and atmospheric interference.

3.3. Uses of the Data Acquired

To date, few uses have been developed to use these ML or DL techniques except some mobile applications to help farmers in the identification of diseases at leaf level [187,188,189] whose advantages and limitations have already been reviewed [43]. A few other studies have started research on the possibility of combining ML with weather data and agronomic knowledge to find a tool that helps farmers with disease detection [190], and a few others have combined AI to detect diseases at the field level to decide when is more convenient to make the treatment [191]. However, to the date, there is a gap between the disease detection and the treatment by farmers, as most of the researchers found work at the leaf level. Research has achieved the recognition and identification of pathogens and disease symptoms compared to sane or not contaminated organs. However, they have not tested their accuracy at the field level, so we can expect that their use is limited. They could help low-skilled farmers to identify the disease, but it will require agronomical knowledge to determine if it is necessary to make a treatment, the product that needs to be applied, and the amount to be applied.

4. Perspectives and Challenges

Once the technology is developed, it will be extended to farmers’ fields. Then, other issues may rise. Technically, one major problem will be the quality of the acquired images, particularly with latent symptoms and small-sized lesions. There is a challenge in capturing the diseases present in the lower leaves of the canopy, in motion but also because of the presence of shadows or bad light depending on the time of acquisition and the weather conditions. The second challenge will be the speed of data treatment before disease management actions are taken on the field.
From an agronomic point of view, another major challenge will be the annotation of diseases on leaves in non-controlled environments such as field conditions. In those situations, the progression of the diseases relies on many factors such as the meteorological conditions and the crop density. The developmental stage of each individual plant in the crop’s field may not result in a uniform distribution of the disease in the field. Additionally, in general, there are several concomitant pests in a field combined with the progression of the senescence of older leaves due to natural optimization of allocation of nutrients to maximize photosynthesis in the upper leaves. Even for trained personnel, it is difficult to distinguish those symptoms combined in the canopy if the data (for annotations or for the data acquisition) are not collected with a high quality (proximal way), in which case the information will be less accurate, demanding a leaf-by-leaf annotation within a field which would be tedious and long.
On the contrary, machine learning tools do not require such detailed annotations because they are going to have a global grade of disease level (considering the global health status of the plant) with not many details, making the step of annotation quicker but less precise.
If the variable rate of application is considered, then there are several possibilities with different consequences which could be observed on field yield or produce quality, disease resistance development and disease spreading, conducting to a need for re-treatment. Once there is an identified risk on the parcel that may require a chemical treatment, it is necessary to determine the rules that decide if a person should apply or not apply the chemical product and which quantity will be necessary. Firstly, the choice may be to cover the whole parcel but with modulated application quantities depending on the risk of infection. Secondly, the choice may be to apply the product only where the disease is present or to apply only where the disease is present with or without a surrounding margin around the target to avoid proximal contamination. Lastly, the choice may be to use a reduction of the recommended doses in cases of low risk to prevent the development of resistance by pathogens due to mutations [192].
Field-level and drone-based imaging provide high-resolution, actionable insights, enabling early detection and targeted interventions. Despite current limitations in resolution and computational demand, drones fitted with multispectral sensors should be considered for monitoring extensive agricultural areas, while there is a need to further investigate the potential of satellite data for regional crop health monitoring.
The integration of advanced sensing techniques like fluorescence imaging, spectroscopy, and thermal imaging at the leaf level represents a promising direction for crop disease detection. When combined with ML and DL models, these methods could provide precise, early-stage disease identification, allowing for timely interventions. Future research should focus on optimizing these tools for field use, potentially transforming precision agriculture with enhanced disease monitoring and management capabilities at the plant level.
The field of crop disease detection can be significantly enhanced by the adoption of new ML and DL technologies. Federated learning allows for the training of models on decentralized data sources while preserving data privacy, enabling diverse agricultural datasets from multiple farms to improve disease detection accuracy. Additionally, transfer learning and domain adaptation can optimize model performance across different crops and environmental conditions, addressing the challenge of limited labelled datasets. Ensemble learning techniques can further enhance predictive performance by combining outputs from multiple models to reduce variance and bias. Advances in edge computing will facilitate real-time data processing from IoT sensors and drones, ensuring timely disease detection and response. Furthermore, multi-modal data integration merging visual images, spectral data, soil health information, and climatic conditions promises a comprehensive understanding of crop health dynamics. Finally, enhancing user-friendly mobile applications that leverage ML and DL algorithms can empower farmers to diagnose diseases using smartphones, offering immediate guidance for effective management practices. Collectively, these innovations hold great potential in improving crop health management, increasing yields, and promoting sustainable agricultural practices.

5. Conclusions

This review shows that there are many crops where artificial intelligence using deep learning or machine learning methods have been used to detect and identify diseases. Unfortunately, even if the algorithms developed presented a high accuracy, as they are mainly developed with the image of one leaf out of an agronomical context (other leaves, other plants, or varying environmental conditions), their use remains difficult in establishing a field diagnosis.
Nevertheless, some smartphone applications may help farmers to identify the disease present in their field as a decision support system. Future research is needed to extend their use to field applications and combine them with localized treatments.

Author Contributions

Y.L. and A.A.G. have contributed equally to the conceptualization, data curation, data validation, and writing. Y.L. has defined the methodology and made the formal analysis. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

A Zotero database with all the retrieved references is available on request to the authors.

Acknowledgments

Authors would like to thank D. Rizzo for the help in retrieving the last SCOPUS references. We are also thankful to Fataw Ibrahim for the English reviewing and comments given to improve this document. We would also like to express our gratitude to the APEX Lab Platform for its invaluable support.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. United Nations SDGS, United Nations. Available online: https://sdgs.un.org/goals (accessed on 24 June 2024).
  2. Whitcraft, A.K.; Becker-Reshef, I.; Justice, C.O.; Gifford, L.; Kavvada, A.; Jarvis, I. No Pixel Left behind: Toward Integrating Earth Observations for Agriculture into the United Nations Sustainable Development Goals Framework. Remote Sens. Environ. 2019, 235, 111470. [Google Scholar] [CrossRef]
  3. Gaunt, R.E. The Relationship Between Plant Disease Severity and Yield. Annu. Rev. Phytopathol. 1995, 33, 119–144. [Google Scholar] [CrossRef] [PubMed]
  4. Courson, E.; Petit, S.; Poggi, S.; Ricci, B. Weather and Landscape Drivers of the Regional Level of Pest Occurrence in Arable Agriculture: A Multi-Pest Analysis at the French National Scale. Agric. Ecosyst. Environ. 2022, 338, 108105. [Google Scholar] [CrossRef]
  5. Milosavljević, I.; Esser, A.D.; Crowder, D.W. Effects of Environmental and Agronomic Factors on Soil-Dwelling Pest Communities in Cereal Crops. Agric. Ecosyst. Environ. 2016, 225, 192–198. [Google Scholar] [CrossRef]
  6. Paredes, D.; Rosenheim, J.A.; Karp, D.S. The Causes and Consequences of Pest Population Variability in Agricultural Landscapes. Ecol. Appl. 2022, 32, e2607. [Google Scholar] [CrossRef]
  7. Poole, N.F.; Arnaudin, M.E. The Role of Fungicides for Effective Disease Management in Cereal Crops. Can. J. Plant Pathol. 2014, 36, 1–11. [Google Scholar] [CrossRef]
  8. Dewasme, C.; Mary, S.; Darrieutort, G.; Roby, J.-P.; Gambetta, G.A. Long-Term Esca Monitoring Reveals Disease Impacts on Fruit Yield and Wine Quality. Plant Dis. 2022, 106, 3076–3082. [Google Scholar] [CrossRef]
  9. Keinath, A.P. Effect of Fungicide Applications Scheduled to Control Gummy Stem Blight on Yield and Quality of Watermelon Fruit. Plant Dis. 2001, 85, 53–58. [Google Scholar] [CrossRef]
  10. Milošević, T.; Milošević, N.; Mladenović, J.; Jevremović, D. Impact of Sharka Disease on Tree Growth, Productivity and Fruit Quality of Apricot (Prunus armeniaca L.). Sci. Hortic. 2019, 244, 270–276. [Google Scholar] [CrossRef]
  11. Ferrigo, D.; Raiola, A.; Causin, R. Fusarium Toxins in Cereals: Occurrence, Legislation, Factors Promoting the Appearance and Their Management. Molecules 2016, 21, 627. [Google Scholar] [CrossRef]
  12. Oldenburg, E.; Höppner, F.; Ellner, F.; Weinert, J. Fusarium Diseases of Maize Associated with Mycotoxin Contamination of Agricultural Products Intended to Be Used for Food and Feed. Mycotoxin Res. 2017, 33, 167–182. [Google Scholar] [CrossRef] [PubMed]
  13. Siou, D.; Gélisse, S.; Laval, V.; Elbelt, S.; Repinçay, C.; Bourdat-Deschamps, M.; Suffert, F.; Lannou, C. Interactions between Head Blight Pathogens: Consequences for Disease Development and Toxin Production in Wheat Spikes. Appl. Environ. Microbiol. 2015, 81, 957–965. [Google Scholar] [CrossRef] [PubMed]
  14. Brown, J.F.; Keane, P. Assessment of Disease and Effects on Yield. In Plant Pathogens and Plant Diseases; Brown, J.F., Ogle, H.J., Eds.; Rockvale Publications: Armidale, Australia, 1997; pp. 315–329. [Google Scholar]
  15. Price, C.L.; Parker, J.E.; Warrilow, A.G.; Kelly, D.E.; Kelly, S.L. Azole Fungicides—Understanding Resistance Mechanisms in Agricultural Fungal Pathogens. Pest Manag. Sci. 2015, 71, 1054–1058. [Google Scholar] [CrossRef] [PubMed]
  16. Boobis, A.R.; Ossendorp, B.C.; Banasiak, U.; Hamey, P.Y.; Sebestyen, I.; Moretto, A. Cumulative Risk Assessment of Pesticide Residues in Food. Toxicol. Lett. 2008, 180, 137–150. [Google Scholar] [CrossRef] [PubMed]
  17. Damalas, C.A.; Koutroubas, S.D. Farmers’ Exposure to Pesticides: Toxicity Types and Ways of Prevention. Toxics 2016, 4, 1. [Google Scholar] [CrossRef]
  18. Dhananjayan, V.; Ravichandran, B. Occupational Health Risk of Farmers Exposed to Pesticides in Agricultural Activities. Curr. Opin. Environ. Sci. Health 2018, 4, 31–37. [Google Scholar] [CrossRef]
  19. Görlich, F.; Marks, E.; Mahlein, A.-K.; König, K.; Lottes, P.; Stachniss, C. Uav-Based Classification of Cercospora Leaf Spot Using Rgb Images. Drones 2021, 5, 34. [Google Scholar] [CrossRef]
  20. Bohnenkamp, D.; Behmann, J.; Mahlein, A.-K. In-Field Detection of Yellow Rust in Wheat on the Ground Canopy and UAV Scale. Remote Sens. 2019, 11, 2495. [Google Scholar] [CrossRef]
  21. Pourazar, H.; Samadzadegan, F.; Dadrass Javan, F. Aerial Multispectral Imagery for Plant Disease Detection: Radiometric Calibration Necessity Assessment. Eur. J. Remote Sens. 2019, 52, 17–31. [Google Scholar] [CrossRef]
  22. Ouhami, M.; Hafiane, A.; Es-Saady, Y.; El Hajji, M.; Canals, R. Computer Vision, IoT and Data Fusion for Crop Disease Detection Using Machine Learning: A Survey and Ongoing Research. Remote Sens. 2021, 13, 2486. [Google Scholar] [CrossRef]
  23. Zhang, J.; Huang, Y.; Pu, R.; Gonzalez-Moreno, P.; Yuan, L.; Wu, K.; Huang, W. Monitoring Plant Diseases and Pests through Remote Sensing Technology: A Review. Comput. Electron. Agric. 2019, 165, 104943. [Google Scholar] [CrossRef]
  24. Abbas, A.; Jain, S.; Gour, M.; Vankudothu, S. Tomato Plant Disease Detection Using Transfer Learning with C-GAN Synthetic Images. Comput. Electron. Agric. 2021, 187, 106279. [Google Scholar] [CrossRef]
  25. Saleem, M.H.; Potgieter, J.; Arif, K.M. Plant Disease Detection and Classification by Deep Learning. Plants 2019, 8, 468. [Google Scholar] [CrossRef] [PubMed]
  26. Gogoi, N.K.; Deka, B.; Bora, L.C. Remote Sensing and Its Use in Detection and Monitoring Plant Diseases: A Review. Agric. Rev. 2018, 39, 307–313. [Google Scholar] [CrossRef]
  27. Tian, S.; Xu, H. Nondestructive Methods for the Quality Assessment of Fruits and Vegetables Considering Their Physical and Biological Variability. Food Eng. Rev. 2022, 14, 380–407. [Google Scholar] [CrossRef]
  28. Padmavathi, K.; Thangadurai, K. Implementation of RGB and Grayscale Images in Plant Leaves Disease Detection—Comparative Study. Indian. J. Sci. Technol. 2016, 9, 1–6. [Google Scholar] [CrossRef]
  29. Cholachgudda, K.E.; Biradar, R.C.; Olivier Akansie, K.Y.; Devanagavi, G.D.; Sannabhadti, A.A. Design of a Multispectral and Thermal Data Acquisition System for High-Throughput Phenotyping of Plant Pathology. In Proceedings of the 2022 IEEE International Conference on Electronics, Computing and Communication Technologies, CONECCT 2022, Bangalore, India, 8–10 July 2022; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2022. [Google Scholar]
  30. Cheshkova, A.F. A Review of Hyperspectral Image Analysis Techniques for Plant Disease Detection and Identification. Vavilovskii Zhurnal Genet. Sel. 2022, 26, 202–213. [Google Scholar] [CrossRef]
  31. Pérez-Bueno, M.L.; Pineda, M.; Cabeza, F.M.; Barón, M. Multicolor Fluorescence Imaging as a Candidate for Disease Detection in Plant Phenotyping. Front. Plant Sci. 2016, 7, 1790. [Google Scholar] [CrossRef]
  32. Zhu, W.; Chen, H.; Ciechanowska, I.; Spaner, D. Application of Infrared Thermal Imaging for the Rapid Diagnosis of Crop Disease. IFAC-Pap. 2018, 51, 424–430. [Google Scholar] [CrossRef]
  33. Raji, S.N.; Subhash, N.; Ravi, V.; Saravanan, R.; Mohanan, C.N.; Nita, S.; Kumar, T.M. Detection of Mosaic Virus Disease in Cassava Plants by Sunlight-Induced Fluorescence Imaging: A Pilot Study for Proximal Sensing. Int. J. Remote Sens. 2015, 36, 2880–2897. [Google Scholar] [CrossRef]
  34. Yu, K.; Leufen, G.; Hunsche, M.; Noga, G.; Chen, X.; Bareth, G. Investigation of Leaf Diseases and Estimation of Chlorophyll Concentration in Seven Barley Varieties Using Fluorescence and Hyperspectral Indices. Remote Sens. 2013, 6, 64–86. [Google Scholar] [CrossRef]
  35. Kale, A.P.; Sonavane, S.P. IoT Based Smart Farming: Feature Subset Selection for Optimized High-Dimensional Data Using Improved GA Based Approach for ELM. Comput. Electron. Agric. 2019, 161, 225–232. [Google Scholar] [CrossRef]
  36. Al-Saddik, H.; Laybros, A.; Billiot, B.; Cointault, F. Using Image Texture and Spectral Reflectance Analysis to Detect Yellowness and Esca in Grapevines at Leaf-Level. Remote Sens. 2018, 10, 618. [Google Scholar] [CrossRef]
  37. Robi, S.N.A.M.; Ahmad, N.; Izhar, M.A.M.; Kaidi, H.M.; Noor, N.M. Utilizing UAV Data for Neural Network-Based Classification of Melon Leaf Diseases in Smart Agriculture. Intl. J. Adv. Comput. Sci. Appl. 2024, 15, 1212–1219. [Google Scholar] [CrossRef]
  38. Bao, W.; Huang, C.; Hu, G.; Su, B.; Yang, X. Detection of Fusarium Head Blight in Wheat Using UAV Remote Sensing Based on Parallel Channel Space Attention. Comput. Electron. Agric. 2024, 217, 108630. [Google Scholar] [CrossRef]
  39. Neupane, K.; Baysal-Gurel, F. Automatic Identification and Monitoring of Plant Diseases Using Unmanned Aerial Vehicles: A Review. Remote Sens. 2021, 13, 3841. [Google Scholar] [CrossRef]
  40. Abade, A.; Ferreira, P.A.; de Barros Vidal, F. Plant Diseases Recognition on Images Using Convolutional Neural Networks: A Systematic Review. Comput. Electron. Agric. 2021, 185, 106125. [Google Scholar] [CrossRef]
  41. Jafar, A.; Bibi, N.; Naqvi, R.A.; Sadeghi-Niaraki, A.; Jeong, D. Revolutionizing agriculture with artificial intelligence: Plant disease detection methods, applications, and their limitations. J. Sens. 2024, 15, 1356260. [Google Scholar] [CrossRef]
  42. Messina, G.; Modica, G. Applications of UAV Thermal Imagery in Precision Agriculture: State of the Art and Future Research Outlook. Remote Sens. 2020, 12, 1491. [Google Scholar] [CrossRef]
  43. Siddiqua, A.; Kabir, M.A.; Ferdous, T.; Ali, I.B.; Weston, L.A. Evaluating Plant Disease Detection Mobile Applications: Quality and Limitations. Agronomy 2022, 12, 1869. [Google Scholar] [CrossRef]
  44. Van Eck, N.J.; Waltman, L. Software Survey: VOSviewer, a Computer Program for Bibliometric Mapping. Scientometrics 2010, 84, 523–538. [Google Scholar] [CrossRef] [PubMed]
  45. Aria, M.; Cuccurullo, C. Bibliometrix: An R-Tool for Comprehensive Science Mapping Analysis. J. Informetr. 2017, 11, 959–975. [Google Scholar] [CrossRef]
  46. Selcuk, A.A. A Guide for Systematic Reviews: PRISMA. Department of Otorhinolaryngology, Health Sciences University, Derince Training and Research Hospital, Kocaeli, Turkey. Turk. Arch. Otorhinolaryngol. 2019, 57, 57–58. [Google Scholar] [CrossRef] [PubMed]
  47. Trippa, D.; Scalenghe, R.; Basso, M.F.; Panno, S.; Davino, S.; Morone, C.; Giovino, A.; Oufensou, S.; Luchi, N.; Yousefi, S.; et al. Next-generation Methods for Early Disease Detection in Crops. Pest Manag. Sci. 2024, 80, 245–261. [Google Scholar] [CrossRef]
  48. Sanghavi, K.; Sanghavi, M.; Rajurkar, A.M. Early Stage Detection of Downey and Powdery Mildew Grape Disease Using Atmospheric Parameters through Sensor Nodes. Artif. Intell. Agric. 2021, 5, 223–232. [Google Scholar] [CrossRef]
  49. Page, M.J.; McKenzie, J.E.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.; Brennan, S.E.; et al. The PRISMA 2020 Statement: An Updated Guideline for Reporting Systematic Reviews. BMJ 2021, 372, n71. [Google Scholar] [CrossRef]
  50. Kundu, N.; Rani, G.; Dhaka, V.S.; Gupta, K.; Nayak, S.C.; Verma, S.; Ijaz, M.F.; Woźniak, M. Iot and Interpretable Machine Learning Based Framework for Disease Prediction in Pearl Millet. Sensors 2021, 21, 5386. [Google Scholar] [CrossRef]
  51. Nie, X.; Wang, L.; Ding, H.; Xu, M. Strawberry Verticillium Wilt Detection Network Based on Multi-Task Learning and Attention. IEEE Access 2019, 7, 170003–170011. [Google Scholar] [CrossRef]
  52. Martínez-Heredia, J.M.; Gálvez, A.I.; Colodro, F.; Mora-Jiménez, J.L.; Sassi, O.E. Feasibility Study of Detection of Ochre Spot on Almonds Aimed at Very Low-Cost Cameras Onboard a Drone. Drones 2023, 7, 186. [Google Scholar] [CrossRef]
  53. Abdulridha, J.; Ehsani, R.; Abd-Elrahman, A.; Ampatzidis, Y. A Remote Sensing Technique for Detecting Laurel Wilt Disease in Avocado in Presence of Other Biotic and Abiotic Stresses. Comput. Electron. Agric. 2019, 156, 549–557. [Google Scholar] [CrossRef]
  54. Albayati, J.S.H.; Üstündağ, B.B. Evolutionary Feature Optimization for Plant Leaf Disease Detection by Deep Neural Networks. Int. J. Comput. Intell. Syst. 2020, 13, 12–23. [Google Scholar] [CrossRef]
  55. Amin, H.; Darwish, A.; Hassanien, A.E. Deep Learning-Based Apple Leaves Disease Identification Approach with Imbalanced Data; Lecture Notes on Data Engineering and Communications Technologies; Springer Science and Business Media Deutschland GmbH: Cham, Switzerland, 2022; Volume 113. [Google Scholar]
  56. Barthel, D.; Cullinan, C.; Mejia-Aguilar, A.; Chuprikova, E.; McLeod, B.A.; Kerschbamer, C.; Trenti, M.; Monsorno, R.; Prechsl, U.E.; Janik, K. Identification of Spectral Ranges That Contribute to Phytoplasma Detection in Apple Trees—A Step towards an on-Site Method. Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 2023, 303, 123246. [Google Scholar] [CrossRef] [PubMed]
  57. Omrani, E.; Khoshnevisan, B.; Shamshirband, S.; Saboohi, H.; Anuar, N.B.; Nasir, M.H.N.M. Potential of Radial Basis Function-Based Support Vector Regression for Apple Disease Detection. Meas. J. Int. Meas. Confed. 2014, 55, 512–519. [Google Scholar] [CrossRef]
  58. Uryasheva, A.; Kalashnikova, A.; Shadrin, D.; Evteeva, K.; Moskovtsev, E.; Rodichenko, N. Computer Vision-Based Platform for Apple Leaves Segmentation in Field Conditions to Support Digital Phenotyping. Comput. Electron. Agric. 2022, 201, 107269. [Google Scholar] [CrossRef]
  59. Navrozidis, I.; Alexandridis, T.K.; Dimitrakos, A.; Lagopodi, A.L.; Moshou, D.; Zalidis, G. Identification of Purple Spot Disease on Asparagus Crops across Spatial and Spectral Scales. Comput. Electron. Agric. 2018, 148, 322–329. [Google Scholar] [CrossRef]
  60. Blessy Annie Flora, J.; Radha, S.; Hemalatha, R.; Aasha Nandhini, S. Plant Disease Detection for Banana Using Long Range Wide Area Network. Int. J. Secur. Netw. 2021, 16, 129–134. [Google Scholar] [CrossRef]
  61. Hari, P.; Singh, M.P. A Lightweight Convolutional Neural Network for Disease Detection of Fruit Leaves. Neural Comput. Appl. 2023, 35, 14855–14866. [Google Scholar] [CrossRef]
  62. Zhang, S.; Li, X.; Ba, Y.; Lyu, X.; Zhang, M.; Li, M. Banana Fusarium Wilt Disease Detection by Supervised and Unsupervised Methods from UAV-Based Multispectral Imagery. Remote Sens. 2022, 14, 1231. [Google Scholar] [CrossRef]
  63. Thomas, S.; Wahabzada, M.; Kuska, M.T.; Rascher, U.; Mahlein, A.-K. Observation of Plant-Pathogen Interaction by Simultaneous Hyperspectral Imaging Reflection and Transmission Measurements. Funct. Plant Biol. 2017, 44, 23–34. [Google Scholar] [CrossRef]
  64. Elfatimi, E.; Eryiğit, R.; Shehu, H.A. Impact of Datasets on the Effectiveness of MobileNet for Beans Leaf Disease Detection. Neural Comput. Appl. 2024, 36, 1773–1789. [Google Scholar] [CrossRef]
  65. Kursun, R.; Bastas, K.K.; Koklu, M. Segmentation of Dry Bean (Phaseolus vulgaris L.) Leaf Disease Images with U-Net and Classification Using Deep Learning Algorithms. Eur. Food Res. Technol. 2023, 249, 2543–2558. [Google Scholar] [CrossRef]
  66. Sunil, C.K.; Jaidhar, C.D.; Patil, N. Cardamom Plant Disease Detection Approach Using EfficientNetV2. IEEE Access 2022, 10, 789–804. [Google Scholar] [CrossRef]
  67. Kalyani, G.; Sai Sudheer, K.; Janakiramaiah, B.; Narendra Kumar Rao, B. Hyperparameter Optimization for Transfer Learning-Based Disease Detection in Cassava Plants. J. Sci. Ind. Res. 2023, 82, 536–545. [Google Scholar] [CrossRef]
  68. Ramcharan, A.; Baranowski, K.; McCloskey, P.; Ahmed, B.; Legg, J.; Hughes, D.P. Deep Learning for Image-Based Cassava Disease Detection. Front. Plant Sci. 2017, 8, 1852. [Google Scholar] [CrossRef] [PubMed]
  69. Bobbinpreet; Aljahdali, S.; Sharma, T.; Goyal, B.; Dogra, A.; Mahajan, S.; Pandit, A.K. MRMR Based Feature Vector Design for Efficient Citrus Disease Detection. Comput. Mater. Contin. 2022, 72, 4771–4787. [Google Scholar] [CrossRef]
  70. Dananjayan, S.; Tang, Y.; Zhuang, J.; Hou, C.; Luo, S. Assessment of State-of-the-Art Deep Learning Based Citrus Disease Detection Techniques Using Annotated Optical Leaf Images. Comput. Electron. Agric. 2022, 193, 106658. [Google Scholar] [CrossRef]
  71. Liu, Y.; Xiao, H.; Xu, H.; Rao, Y.; Jiang, X.; Sun, X. Visual Discrimination of Citrus HLB Based on Image Features. Vib. Spectrosc. 2019, 102, 103–111. [Google Scholar] [CrossRef]
  72. Saini, A.K.; Bhatnagar, R.; Srivastava, D.K. Citrus Fruits Diseases Detection and Classification Using Transfer Learning. In Proceedings of the ACM International Conference Proceeding Series, Windhoek, Namibia, 9–12 August 2021; Association for Computing Machinery: New York, NY, USA, 2021; pp. 277–283. [Google Scholar]
  73. Sujatha, R.; Chatterjee, J.M.; Jhanjhi, N.Z.; Brohi, S.N. Performance of Deep Learning vs Machine Learning in Plant Leaf Disease Detection. Microprocess. Microsyst. 2021, 80, 103615. [Google Scholar] [CrossRef]
  74. Syed-Ab-Rahman, S.F.; Hesamian, M.H.; Prasad, M. Citrus Disease Detection and Classification Using End-to-End Anchor-Based Deep Learning Model. Appl. Intell. 2022, 52, 927–938. [Google Scholar] [CrossRef]
  75. Ramamurthy, K.; Thekkath, R.D.; Batra, S.; Chattopadhyay, S. A Novel Deep Learning Architecture for Disease Classification in Arabica Coffee Plants. Concurr. Comput. Pract. Exper. 2023, 35, e7625. [Google Scholar] [CrossRef]
  76. Elaraby, A.; Hamdy, W.; Alruwaili, M. Optimization of Deep Learning Model for Plant Disease Detection Using Particle Swarm Optimizer. Comput. Mater. Contin. 2022, 71, 4019–4031. [Google Scholar] [CrossRef]
  77. Katti, J.; Dharmale, G.; Ubale, S.A.; Deoghare, R.; Sutar, S.H. IoT Innovations in Cotton Plant Disease Detection for Sustainable Agriculture. Internat. J. Intel. Syst. Appl. Eng. 2024, 12, 651–658. [Google Scholar]
  78. Maurya, J.P.; Soni, D.; Devaraju, S.; Goyal, A. Implementation of Leaf Disease Detection Using One-Shot & Region Inception Image Recognition Technique. In Power Engineering and Intelligent Systems; Shrivastava, V., Bansal, J.C., Panigrahi, B.K., Eds.; Springer Science and Business Media Deutschland GmbH: Singapore, 2024; Volume 1098, pp. 435–448. [Google Scholar]
  79. Rai, C.K.; Pahuja, R. Classification of Diseased Cotton Leaves and Plants Using Improved Deep Convolutional Neural Network. Multimed. Tools Appl. 2023, 82, 25307–25325. [Google Scholar] [CrossRef]
  80. Liu, Y.; Wang, Z.; Wang, R.; Chen, J.; Gao, H. Flooding-Based MobileNet to Identify Cucumber Diseases from Leaf Images in Natural Scenes. Comput. Electron. Agric. 2023, 213, 108166. [Google Scholar] [CrossRef]
  81. Sabarre, A.B.L.; Navidad, A.N.S.; Torbela, D.S.; Adtoon, J.J. Development of Durian Leaf Disease Detection on Android Device. Int. J. Electr. Comput. Eng. 2021, 11, 4962–4971. [Google Scholar] [CrossRef]
  82. Chelladurai, V.; Karuppiah, K.; Jayas, D.S.; Fields, P.G.; White, N.D.G. Detection of Callosobruchus maculatus (F.) Infestation in Soybean Using Soft X-Ray and NIR Hyperspectral Imaging Techniques. J. Stored Prod. Res. 2014, 57, 43–48. [Google Scholar] [CrossRef]
  83. Boulent, J.; St-Charles, P.-L.; Foucher, S.; Théau, J. Automatic Detection of Flavescence Dorée Symptoms Across White Grapevine Varieties Using Deep Learning. Front. Artif. Intell. 2020, 3, 564878. [Google Scholar] [CrossRef]
  84. Elsherbiny, O.; Elaraby, A.; Alahmadi, M.; Hamdan, M.; Gao, J. Rapid Grapevine Health Diagnosis Based on Digital Imaging and Deep Learning. Plants 2024, 13, 135. [Google Scholar] [CrossRef]
  85. Hernández, I.; Gutiérrez, S.; Ceballos, S.; Palacios, F.; Toffolatti, S.L.; Maddalena, G.; Diago, M.P.; Tardaguila, J. Assessment of Downy Mildew in Grapevine Using Computer Vision and Fuzzy Logic. Development and Validation of a New Method. Oeno One 2022, 56, 41–53. [Google Scholar] [CrossRef]
  86. Huang, Z.; Qin, A.; Lu, J.; Menon, A.; Gao, J. Grape Leaf Disease Detection and Classification Using Machine Learning. In Proceedings of the Proceedings—IEEE Congress on Cybermatics: 2020 IEEE International Conferences on Internet of Things, iThings 2020, IEEE Green Computing and Communications, GreenCom 2020, IEEE Cyber, Physical and Social Computing, CPSCom 2020 and IEEE Smart Data, SmartData 2020, Rhodes, Greece, 2–6 November 2020; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2020; pp. 870–877. [Google Scholar]
  87. Sood, S.; Singh, H. A Comparative Study of Grape Crop Disease Classification Using Various Transfer Learning Techniques. Multimed. Tools Appl. 2024, 83, 4359–4382. [Google Scholar] [CrossRef]
  88. Mostafa, A.M.; Kumar, S.A.; Meraj, T.; Rauf, H.T.; Alnuaim, A.A.; Alkhayyal, M.A. Guava Disease Detection Using Deep Convolutional Neural Networks: A Case Study of Guava Plants. Appl. Sci. 2022, 12, 239. [Google Scholar] [CrossRef]
  89. Agustiono, W.; Syarief, M.; Anamisa, D.; Damayanti, F.; Yusuf, M. A Conceptual Model of Mobile Expert System for Integrated Pest and Disease Control: The Case of MyCorn. J. Phys. Conf. Ser. 2020, 1569, 022057. [Google Scholar] [CrossRef]
  90. Akanksha, E.; Sharma, N.; Gulati, K. OPNN: Optimized Probabilistic Neural Network Based Automatic Detection of Maize Plant Disease Detection. In Proceedings of the 6th International Conference on Inventive Computation Technologies, ICICT 2021, Coimbatore, India, 20–22 January 2021; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2021; pp. 1322–1328. [Google Scholar]
  91. Giraddi, S.; Desai, S.; Deshpande, A. Deep Learning for Agricultural Plant Disease Detection; Springer: Singapore, 2020; Volume 601. [Google Scholar]
  92. Nandhini, S.; Ashokkumar, K. Identification of Maize Plant Diseases Based on Linear Vector Quantization with Neural Network. J. Uncertain. Syst. 2022, 15, 22410057. [Google Scholar] [CrossRef]
  93. PAN, S.-Q.; QIAO, J.-F.; WANG, R.; YU, H.-L.; WANG, C.; TAYLOR, K.; PAN, H.-Y. Intelligent Diagnosis of Northern Corn Leaf Blight with Deep Learning Model. J. Integr. Agric. 2022, 21, 1094–1105. [Google Scholar] [CrossRef]
  94. Panigrahi, K.P.; Das, H.; Sahoo, A.K.; Moharana, S.C. Maize Leaf Disease Detection and Classification Using Machine Learning Algorithms; Springer: Singapore, 2020; Volume 1119. [Google Scholar]
  95. Hu, R.; Zhang, S.; Wang, P.; Xu, G.; Wang, D.; Qian, Y. The Identification of Corn Leaf Diseases Based on Transfer Learning and Data Augmentation. In Proceedings of the 3rd International Conference on Computer Science and Software Engineering, Beijing, China, 22–24 May 2020; Association for Computing Machinery: New York, NY, USA, 2020; pp. 58–65. [Google Scholar]
  96. Ahmed, S.I.; Ibrahim, M.; Nadim, M.; Rahman, M.M.; Shejunti, M.M.; Jabid, T.; Ali, M.S. MangoLeafBD: A Comprehensive Image Dataset to Classify Diseased and Healthy Mango Leaves. Data Brief 2023, 47, 108941. [Google Scholar] [CrossRef]
  97. Priyadharshini, M.K.; Sivakami, R.; Janani, M. Sooty Mould Mango Disease Identification Using Deep Learning. Int. J. Innov. Technol. Explor. Eng. 2019, 8, 402–405. [Google Scholar]
  98. Rajpoot, V.; Dubey, R.; Mannepalli, P.K.; Kalyani, P.; Maheshwari, S.; Dixit, A.; Saxena, A. Mango Plant Disease Detection System Using Hybrid BBHE and CNN Approach. Trait. Du. Signal 2022, 39, 1071–1078. [Google Scholar] [CrossRef]
  99. El Abidine, M.Z.; Merdinoglu-Wiedemann, S.; Rasti, P.; Dutagaci, H.; Rousseau, D. Machine Learning-Based Classification of Powdery Mildew Severity on Melon Leaves; Springer: Cham, Switzerland, 2020; Volume 12119. [Google Scholar]
  100. Salam, A.; Naznine, M.; Jahan, N.; Nahid, E.; Nahiduzzaman, M.; Chowdhury, M.E.H. Mulberry Leaf Disease Detection Using CNN-Based Smart Android Application. IEEE Access 2024, 12, 83575–83588. [Google Scholar] [CrossRef]
  101. Khairi, N.A.S.; Yusof, N.A.; Abdullah, J.; Seman, I.A.; Ithnin, N.; Abdrahman, S.F.A. β-Cyclodextrin Functionalized Reduced Graphene Oxide-Gold Nanoparticles for Electrochemical Detection of Stigmasterol in Ganoderma Boninense-Infected Oil Palm Leaves. IEEE Sens. J. 2024, 24, 9341–9352. [Google Scholar] [CrossRef]
  102. Alshammari, H.H.; Taloba, A.I.; Shahin, O.R. Identification of Olive Leaf Disease through Optimized Deep Learning Approach. Alex. Eng. J. 2023, 72, 213–224. [Google Scholar] [CrossRef]
  103. Alshammari, H.H.; Alkhiri, H. Optimized Recurrent Neural Network Mechanism for Olive Leaf Disease Diagnosis Based on Wavelet Transform. Alex. Eng. J. 2023, 78, 149–161. [Google Scholar] [CrossRef]
  104. Alberto, R.; Isip, M.; Biagtan, A. Evaluation of Vegetation Indices (VIs) to Detect Twister Disease of Onion Using Sentinel-2 Imagery. In Proceedings of the 40th Asian Conference on Remote Sensing, ACRS 2019: Progress of Remote Sensing Technology for Smart Future, Daejeon, Republic of Korea, 14–18 October 2019; Korean Society of Remote Sensing (KSRS): Seoul, Republic of Korea; Asian Association on Remote Sensing: Colombo, Sri Lanka, 2020. [Google Scholar]
  105. Isip, M.F.; Alberto, R.T.; Biagtan, A.R. Exploring Vegetation Indices Adequate in Detecting Twister Disease of Onion Using Sentinel-2 Imagery. Spat. Inf. Res. 2020, 28, 369–375. [Google Scholar] [CrossRef]
  106. Haq, Q.M.I.; Mabood, F.; Naureen, Z.; Al-Harrasi, A.; Gilani, S.A.; Hussain, J.; Jabeen, F.; Khan, A.; Al-Sabari, R.S.M.; Al-khanbashi, F.H.S.; et al. Application of Reflectance Spectroscopies (FTIR-ATR & FT-NIR) Coupled with Multivariate Methods for Robust in Vivo Detection of Begomovirus Infection in Papaya Leaves. Spectrochim. Acta—Part A Mol. Biomol. Spectrosc. 2018, 198, 27–32. [Google Scholar] [CrossRef]
  107. Singh, K.; Kumar, S.; Kaur, P. Support Vector Machine Classifier Based Detection of Fungal Rust Disease in Pea Plant (Pisam sativam). Int. J. Inf. Technol. 2019, 11, 485–492. [Google Scholar] [CrossRef]
  108. Wei, X.; Johnson, M.A.; Langston, D.B., Jr.; Mehl, H.L.; Li, S. Identifying Optimal Wavelengths as Disease Signatures Using Hyperspectral Sensor and Machine Learning. Remote Sens. 2021, 13, 2833. [Google Scholar] [CrossRef]
  109. Ahmad, M.; Abdullah, M.; Moon, H.; Han, D. Plant Disease Detection in Imbalanced Datasets Using Efficient Convolutional Neural Networks with Stepwise Transfer Learning. IEEE Access 2021, 9, 140565–140580. [Google Scholar] [CrossRef]
  110. Agustika, D.K.; Mercuriani, I.; Purnomo, C.W.; Hartono, S.; Triyana, K.; Iliescu, D.D.; Leeson, M.S. Fourier Transform Infrared Spectrum Pre-Processing Technique Selection for Detecting PYLCV-Infected Chilli Plants. Spectrochim. Acta—Part A Mol. Biomol. Spectrosc. 2022, 278, 121339. [Google Scholar] [CrossRef]
  111. Hasbollah, A.F.K.; Zin, Z.M.; Ibrahim, N.; Suleiman, R.F.R. Green Chilli Leaf Disease Detection Using Convolution Neural Networks. J. Green Eng. 2020, 10, 13005–13019. [Google Scholar]
  112. Tej, B.; Bouaafia, S.; Hajjaji, M.A.; Mtibaa, A. AI-Based Smart Agriculture 4.0 System for Plant Diseases Detection in Tunisia. Signal Image Video Process. 2024, 18, 97–111. [Google Scholar] [CrossRef]
  113. Vasavi, P.; Punitha, A.; Rao, T.V. Chili Leaf Disease Detection Using Deep Feature Extraction. J. Intell. Syst. Internet. Thing. 2023, 9, 222–230. [Google Scholar] [CrossRef]
  114. Bhagat, S.; Kokare, M.; Haswani, V.; Hambarde, P.; Taori, T.; Ghante, P.H.; Patil, D.K. Advancing Real-Time Plant Disease Detection: A Lightweight Deep Learning Approach and Novel Dataset for Pigeon Pea Crop. Smart Agric. Technol. 2024, 7, 100408. [Google Scholar] [CrossRef]
  115. Ahmad, J.; Jan, B.; Farman, H.; Ahmad, W.; Ullah, A. Disease Detection in Plum Using Convolutional Neural Network under True Field Conditions. Sensors 2020, 20, 5569. [Google Scholar] [CrossRef] [PubMed]
  116. Kaushik, I.; Prakash, N.; Jain, A. Plant Disease Detection Using a Depth-Wise Separable-Based Adaptive Deep Neural Network. Multimed. Tools Appl. 2024. [Google Scholar] [CrossRef]
  117. Lanjewar, M.G.; Morajkar, P.; Payaswini, P. Modified Transfer Learning Frameworks to Identify Potato Leaf Diseases. Multimed. Tools Appl. 2024, 83, 50401–50423. [Google Scholar] [CrossRef]
  118. Sarah, M.; Abdlemadjid, M.; Sarah, B.; Yacine, H.; Miloud, C.E.M. Evaluating the Effect of Super-Resolution for Automatic Plant Disease Detection: Application to Potato Late Blight Detection. Multimed. Tools Appl. 2024, 83, 78469–78487. [Google Scholar] [CrossRef]
  119. Sofuoglu, C.I.; Birant, D. Potato Plant Leaf Disease Detection Using Deep Learning Method. Tarim. Bilim. Derg. 2024, 30, 153–165. [Google Scholar] [CrossRef]
  120. Kong, W.; Zhang, C.; Cao, F.; Liu, F.; Luo, S.; Tang, Y.; He, Y. Detection of Sclerotinia Stem Rot on Oilseed Rape (Brassica napus L.) Leaves Using Hyperspectral Imaging. Sensors 2018, 18, 1764. [Google Scholar] [CrossRef]
  121. Ahmad Effendi, A.F.A.; Md Isa, M.N.; Ahmad, M.I.; Che Husin, M.F.; Md Naziri, S.Z. Image Processing for Paddy Disease Detection Using K-Means Clustering and GLCM Algorithm. Int. J. Nanoelectron. Mater. 2021, 14, 253–263. [Google Scholar]
  122. Almasoud, A.S.; Abdelmaboud, A.; Eisa, T.A.E.; Al Duhayyim, M.; Elnour, A.A.H.; Hamza, M.A.; Motwakel, A.; Zamani, A.S. Artificial Intelligence-Based Fusion Model for Paddy Leaf Disease Detection and Classification. Comput. Mater. Contin. 2022, 72, 1391–1407. [Google Scholar] [CrossRef]
  123. Azim, M.A.; Islam, M.K.; Rahman, M.M.; Jahan, F. An Effective Feature Extraction Method for Rice Leaf Disease Classification. Telkomnika (Telecommun. Comput. Electron. Control) 2021, 19, 463–470. [Google Scholar] [CrossRef]
  124. Chen, W.-L.; Lin, Y.-B.; Ng, F.-L.; Liu, C.-Y.; Lin, Y.-W. RiceTalk: Rice Blast Detection Using Internet of Things and Artificial Intelligence Technologies. IEEE Internet Things J. 2020, 7, 1001–1010. [Google Scholar] [CrossRef]
  125. Kumar, R.; Baloch, G.; Pankaj, P.; Buriro, A.B.; Bhatti, J. Fungal Blast Disease Detection in Rice Seed Using Machine Learning. Int. J. Adv. Comput. Sci. Appl. 2021, 12, 248–258. [Google Scholar] [CrossRef]
  126. Narmadha, R.P.; Sengottaiyan, N.; Kavitha, R.J. Deep Transfer Learning Based Rice Plant Disease Detection Model. Intell. Autom. Soft Comput. 2022, 31, 1257–1271. [Google Scholar] [CrossRef]
  127. Patidar, S.; Pandey, A.; Shirish, B.A.; Sriram, A. Rice Plant Disease Detection and Classification Using Deep Residual Learning; Springer: Singapore, 2020; Volume 1240. [Google Scholar]
  128. Sharma, R.; Singh, A. Big Bang–Big Crunch-CNN: An Optimized Approach towards Rice Crop Protection and Disease Detection. Arch. Phytopathol. Plant Prot. 2022, 55, 143–161. [Google Scholar] [CrossRef]
  129. Upadhyay, S.K.; Kumar, A. A Novel Approach for Rice Plant Diseases Classification with Deep Convolutional Neural Network. Int. J. Inf. Technol. 2022, 14, 185–199. [Google Scholar] [CrossRef]
  130. Agrawal, M.; Agrawal, S. Rice Plant Diseases Detection Using Convolutional Neural Networks. Int. J. Eng. Syst. Model. Simul. 2023, 14, 30–42. [Google Scholar] [CrossRef]
  131. Minaei, S.; Jafari, M.; Safaie, N. Design and Development of a Rose Plant Disease-Detection and Site-Specific Spraying System Based on a Combination of Infrared and Visible Images. J. Agric. Sci. Technol. 2018, 20, 23–36. [Google Scholar]
  132. Abdu, A.M.; Mokji, M.M.; Sheikh, U.U. Machine Learning for Plant Disease Detection: An Investigative Comparison between Support Vector Machine and Deep Learning. IAES Int. J. Artif. Intell. 2020, 9, 670–683. [Google Scholar] [CrossRef]
  133. Rao, A.R.; Anji Reddy, V.; Raju, G.V.N. Plant Disease Detection Using Machin Learning Algorithms. YMER 2022, 21, 425–430. [Google Scholar] [CrossRef]
  134. Dubey, A. Agricultural Plant Disease Detection and Identification. Int. J. Electr. Eng. Technol. 2020, 11, 354–363. [Google Scholar]
  135. Arjoune, Y.; Sugunaraj, N.; Peri, S.; Nair, S.V.; Skurdal, A.; Ranganathan, P.; Johnson, B. Soybean Cyst Nematode Detection and Management: A Review. Plant Methods 2022, 18, 110. [Google Scholar] [CrossRef] [PubMed]
  136. Feng, L.; Chen, S.; Feng, B.; Liu, F.; He, Y.; Lou, B. Early Detection of Soybean Pod Anthracnose Based on Spectrum Technology. Nongye Gongcheng Xuebao/Trans. Chin. Soc. Agric. Eng. 2012, 28, 139–144. [Google Scholar] [CrossRef]
  137. Lay, L.; Lee, H.S.; Tayade, R.; Ghimire, A.; Chung, Y.S.; Yoon, Y.; Kim, Y. Evaluation of Soybean Wildfire Prediction via Hyperspectral Imaging. Plants 2023, 12, 901. [Google Scholar] [CrossRef]
  138. Ganesh Babu, R.; Chellaswamy, C. Different Stages of Disease Detection in Squash Plant Based on Machine Learning. J. Biosci. 2022, 47, 9. [Google Scholar] [CrossRef]
  139. Belattar, S.; Abdoun, O.; Haimoudi, E.K. Performance Analysis of the Application of Convolutional Neural Networks Architectures in the Agricultural Diagnosis. Indones. J. Electr. Eng. Comput. Sci. 2022, 27, 156–162. [Google Scholar] [CrossRef]
  140. Jiang, Q.; Wu, G.; Tian, C.; Li, N.; Yang, H.; Bai, Y.; Zhang, B. Hyperspectral Imaging for Early Identification of Strawberry Leaves Diseases with Machine Learning and Spectral Fingerprint Features. Infrared Phys. Technol. 2021, 118, 103898. [Google Scholar] [CrossRef]
  141. You, J.; Jiang, K.; Lee, J. Deep Metric Learning-Based Strawberry Disease Detection with Unknowns. Front. Plant Sci. 2022, 13, 891785. [Google Scholar] [CrossRef]
  142. Mahlein, A.K.; Hillnhütter, C.; Mewes, T.; Dehne, H.W.; Steiner, U.; Oerke, E.C. Hyperspectral Imaging of Foliar Sugar Beet Diseases and Automatic Classification by the Spectral Angle Mapper Algorithm. In Proceedings of the Precision Agriculture 2011—Papers Presented at the 8th European Conference on Precision Agriculture 2011, ECPA 2011, Prague, Czech Republic, 11–14 July 2011; Czech Centre for Science and Society: Prague, Czech Republic, 2011; pp. 264–272. [Google Scholar]
  143. Zhou, R.; Kaneko, S.; Tanaka, F.; Kayamori, M.; Shimizu, M. Disease Detection of Cercospora Leaf Spot in Sugar Beet by Robust Template Matching. Comput. Electron. Agric. 2014, 108, 58–70. [Google Scholar] [CrossRef]
  144. Daphal, S.D.; Koli, S.M. Transfer Learning Approach to Sugarcane Foliar Disease Classification with State-of-the-Art Sugarcane Database. In Proceedings of the 2021 International Conference on Computational Intelligence and Computing Applications, ICCICA 2021, Nagpur, India, 18–19 June 2021; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2021. [Google Scholar]
  145. Hemalatha, N.K.; Brunda, R.N.; Prakruthi, G.S.; Prabhu, B.V.B.; Shukla, A.; Narasipura, O.S.J. Sugarcane Leaf Disease Detection through Deep Learning. In Deep Learning for Sustainable Agriculture; Elsevier: Amsterdam, The Netherlands, 2022; pp. 297–323. [Google Scholar]
  146. Roy, A.M.; Bose, R.; Bhaduri, J. A Fast Accurate Fine-Grain Object Detection Model Based on YOLOv4 Deep Neural Network. Neural Comput. Appl. 2022, 34, 3895–3921. [Google Scholar] [CrossRef]
  147. Verma, M.; Anand, A.S.; Srivastava, A. Plant Disease Detection Using CNN Through Segmentation and Balancing Techniques; Springer Science and Business Media Deutschland GmbH: Singapore, 2022; Volume 427. [Google Scholar]
  148. Bakr, M.; Abdel-Gaber, S.; Nasr, M.; Hazman, M. Tomato Disease Detection Model Based on Densenet and Transfer. Appl. Comput. Sci. 2022, 18, 56–70. [Google Scholar] [CrossRef]
  149. Chowdhury, M.E.H.; Rahman, T.; Khandakar, A.; Ayari, M.A.; Khan, A.U.; Khan, M.S.; Al-Emadi, N.; Reaz, M.B.I.; Islam, M.T.; Ali, S.H.M. Automatic and Reliable Leaf Disease Detection Using Deep Learning Techniques. AgriEngineering 2021, 3, 294–312. [Google Scholar] [CrossRef]
  150. Hammou, D.R.; Boubaker, M. Tomato Plant Disease Detection and Classification Using Convolutional Neural Network Architectures Technologies; Springer Science and Business Media Deutschland GmbH: Singapore, 2022; Volume 237. [Google Scholar]
  151. Islam, M.S.; Sultana, S.; Farid, F.A.; Islam, M.N.; Rashid, M.; Bari, B.S.; Hashim, N.; Husen, M.N. Multimodal Hybrid Deep Learning Approach to Detect Tomato Leaf Disease Using Attention Based Dilated Convolution Feature Extractor with Logistic Regression Classification. Sensors 2022, 22, 6079. [Google Scholar] [CrossRef] [PubMed]
  152. Ouhami, M.; Es-Saady, Y.; Hajji, M.E.; Hafiane, A.; Canals, R.; Yassa, M.E. Deep Transfer Learning Models for Tomato Disease Detection; Springer: Cham, Switzerland, 2020; Volume 12119. [Google Scholar]
  153. Ruszczak, B.; Smykała, K.; Dziubański, K. The Detection of Alternaria Solani Infection on Tomatoes Using Ensemble Learning. J. Ambient. Intell. Smart Environ. 2020, 12, 407–418. [Google Scholar] [CrossRef]
  154. Wadadare, S.S.; Fadewar, H.S. Deep Learning Convolution Neural Network for Tomato Leaves Disease Detection by Inception; Springer Science and Business Media Deutschland GmbH: Singapore, 2022; Volume 303. [Google Scholar]
  155. Wang, D.; Vinson, R.; Holmes, M.; Seibel, G.; Bechar, A.; Nof, S.; Tao, Y. Early Detection of Tomato Spotted Wilt Virus by Hyperspectral Imaging and Outlier Removal Auxiliary Classifier Generative Adversarial Nets (OR-AC-GAN). Sci. Rep. 2019, 9, 4377. [Google Scholar] [CrossRef]
  156. Pardede, H.F.; Suryawati, E.; Zilvan, V.; Ramdan, A.; Kusumo, R.B.S.; Heryana, A.; Yuwana, R.S.; Krisnandi, D.; Subekti, A.; Fauziah, F.; et al. Plant Diseases Detection with Low Resolution Data Using Nested Skip Connections. J. Big Data 2020, 7, 57. [Google Scholar] [CrossRef]
  157. Khan, M.A.; Ali, M.; Shah, M.; Mahmood, T.; Ahmad, M.; Jhanjhi, N.Z.; Bhuiyan, M.A.S.; Jaha, E.S. Machine Learning-Based Detection and Classification of Walnut Fungi Diseases. Intell. Autom. Soft Comput. 2021, 30, 771–785. [Google Scholar] [CrossRef]
  158. Su, W.-H.; Zhang, J.; Yang, C.; Page, R.; Szinyei, T.; Hirsch, C.D.; Steffenson, B.J. Automatic Evaluation of Wheat Resistance to Fusarium Head Blight Using Dual Mask-RCNN Deep Learning Frameworks in Computer Vision. Remote Sens. 2021, 13, 26. [Google Scholar] [CrossRef]
  159. Cuenca-Romero, C.; Apolo-Apolo, O.E.; Rodríguez Vázquez, J.N.; Egea, G.; Pérez-Ruiz, M. Tackling Unbalanced Datasets for Yellow and Brown Rust Detection in Wheat. Front. Plant Sci. 2024, 15, 1392409. [Google Scholar] [CrossRef]
  160. Deng, J.; Zhou, H.; Lv, X.; Yang, L.; Shang, J.; Sun, Q.; Zheng, X.; Zhou, C.; Zhao, B.; Wu, J.; et al. Applying Convolutional Neural Networks for Detecting Wheat Stripe Rust Transmission Centers under Complex Field Conditions Using RGB-Based High Spatial Resolution Images from UAVs. Comput. Electron. Agric. 2022, 200, 107211. [Google Scholar] [CrossRef]
  161. Ennadifi, E.; Laraba, S.; Vincke, D.; Mercatoris, B.; Gosselin, B. Wheat Diseases Classification and Localization Using Convolutional Neural Networks and GradCAM Visualization; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2020. [Google Scholar]
  162. Feng, Z.; Song, L.; Duan, J.; He, L.; Zhang, Y.; Wei, Y.; Feng, W. Monitoring Wheat Powdery Mildew Based on Hyperspectral, Thermal Infrared, and RGB Image Data Fusion. Sensors 2022, 22, 31. [Google Scholar] [CrossRef]
  163. Gao, C.; Guo, W.; Yang, C.; Gong, Z.; Yue, J.; Fu, Y.; Feng, H. A Fast and Lightweight Detection Model for Wheat Fusarium Head Blight Spikes in Natural Environments. Comput. Electron. Agric. 2024, 216, 108484. [Google Scholar] [CrossRef]
  164. Moshou, D.; Bravo, C.; Oberti, R.; West, J.; Bodria, L.; McCartney, A.; Ramon, H. Plant Disease Detection Based on Data Fusion of Hyper-Spectral and Multi-Spectral Fluorescence Imaging Using Kohonen Maps. Real-Time Imaging 2005, 11, 75–83. [Google Scholar] [CrossRef]
  165. Terentev, A.; Badenko, V.; Shaydayuk, E.; Emelyanov, D.; Eremenko, D.; Klabukov, D.; Fedotov, A.; Dolzhenko, V. Hyperspectral Remote Sensing for Early Detection of Wheat Leaf Rust Caused by Puccinia triticina. Agriculture 2023, 13, 1186. [Google Scholar] [CrossRef]
  166. Wójtowicz, A.; Piekarczyk, J.; Czernecki, B.; Ratajkiewicz, H. A Random Forest Model for the Classification of Wheat and Rye Leaf Rust Symptoms Based on Pure Spectra at Leaf Scale. J. Photochem. Photobiol. B Biol. 2021, 223, 112278. [Google Scholar] [CrossRef] [PubMed]
  167. Yu, K.; Anderegg, J.; Mikaberidze, A.; Karisto, P.; Mascher, F.; McDonald, B.A.; Walter, A.; Hund, A. Hyperspectral Canopy Sensing of Wheat Septoria Tritici Blotch Disease. Front. Plant Sci. 2018, 9, 1195. [Google Scholar] [CrossRef]
  168. Demilie, W.B. Plant Disease Detection and Classification Techniques: A Comparative Study of the Performances. J. Big Data 2024, 11, 5. [Google Scholar] [CrossRef]
  169. Ferentinos, K.P. Deep Learning Models for Plant Disease Detection and Diagnosis. Comput. Electron. Agric. 2018, 145, 311–318. [Google Scholar] [CrossRef]
  170. Fu, R.; Sha, Y.; Xu, X.; Liu, S.-B. Advancements in the Loop-Mediated Isothermal Amplification Technique for the Rapid Detection of Plant Viruses in Various Crops. Physiol. Mol. Plant. Pathol. 2024, 130, 102229. [Google Scholar] [CrossRef]
  171. Delgado, C.; Benitez, H.; Cruz, M.; Selvaraj, M. Digital Disease Phenotyping. In Proceedings of the International Geoscience and Remote Sensing Symposium (IGARSS), Athens, Greece, 7–12 July 2024; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2019; pp. 5702–5705. [Google Scholar]
  172. Verma, S.; Chug, A.; Singh, A.P.; Singh, D. Plant Disease Detection and Severity Assessment Using Image Processing and Deep Learning Techniques. SN Comput. Sci. 2024, 5, 83. [Google Scholar] [CrossRef]
  173. Maski, P.; Thondiyath, A. Plant Disease Detection Using Advanced Deep Learning Algorithms: A Case Study of Papaya Ring Spot Disease. In Proceedings of the 2021 6th International Conference on Image, Vision and Computing, ICIVC 2021, Qingdao, China, 23–25 July 2021; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2021; pp. 49–54. [Google Scholar]
  174. Albattah, W.; Javed, A.; Nawaz, M.; Masood, M.; Albahli, S. Artificial Intelligence-Based Drone System for Multiclass Plant Disease Detection Using an Improved Efficient Convolutional Neural Network. Front. Plant Sci. 2022, 13, 808380. [Google Scholar] [CrossRef]
  175. Falaschetti, L.; Manoni, L.; Di Leo, D.; Pau, D.; Tomaselli, V.; Turchetti, C. A CNN-Based Image Detector for Plant Leaf Diseases Classification. HardwareX 2022, 12, e00363. [Google Scholar] [CrossRef] [PubMed]
  176. Karthik, R.; Hariharan, M.; Anand, S.; Mathikshara, P.; Johnson, A.; Menaka, R. Attention Embedded Residual CNN for Disease Detection in Tomato Leaves. Appl. Soft Comput. J. 2020, 86, 105933. [Google Scholar] [CrossRef]
  177. Sharma, P.; Berwal, Y.P.S.; Ghai, W. Performance Analysis of Deep Learning CNN Models for Disease Detection in Plants Using Image Segmentation. Inf. Process. Agric. 2020, 7, 566–574. [Google Scholar] [CrossRef]
  178. Gerasimchuk, M.; Uzhinskiy, A. R-CCN Plant Diseases Detector Using Triples Loss and Siamese Neural Networks. Phys. Part. Nucl. Lett. 2022, 19, 570–573. [Google Scholar] [CrossRef]
  179. Wang, H.; Shang, S.; Wang, D.; He, X.; Feng, K.; Zhu, H. Plant Disease Detection and Classification Method Based on the Optimized Lightweight YOLOv5 Model. Agriculture 2022, 12, 931. [Google Scholar] [CrossRef]
  180. Saleem, M.H.; Potgieter, J.; Arif, K.M. A Performance-Optimized Deep Learning-Based Plant Disease Detection Approach for Horticultural Crops of New Zealand. IEEE Access 2022, 10, 89798–89822. [Google Scholar] [CrossRef]
  181. Gandhi, R.; Nimbalkar, S.; Yelamanchili, N.; Ponkshe, S. Plant Disease Detection Using CNNs and GANs as an Augmentative Approach. In Proceedings of the 2018 IEEE International Conference on Innovative Research and Development, ICIRD 2018, Bangkok, Thailand, 1–12 May 2018; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2018; pp. 1–5. [Google Scholar]
  182. Zhao, Y.; Chen, Z.; Gao, X.; Song, W.; Xiong, Q.; Hu, J.; Zhang, Z. Plant Disease Detection Using Generated Leaves Based on DoubleGAN. IEEE/ACM Trans. Comput. Biol. Bioinform. 2022, 19, 1817–1826. [Google Scholar] [CrossRef]
  183. Alatawi, A.A.; Alomani, S.M.; Alhawiti, N.I.; Ayaz, M. Plant Disease Detection Using AI Based VGG-16 Model. Int. J. Adv. Comput. Sci. Appl. 2022, 13, 718–727. [Google Scholar] [CrossRef]
  184. Agrawal, R.; Verma, P.; Garg, P.; Jain, D.; Raikwar, S. Plant Disease Analysis and Identification Using Multi SVM Approach. Int. J. Adv. Sci. Technol. 2020, 29, 4848–4856. [Google Scholar]
  185. Dalal, T.; Singh, M. Review Paper on Leaf Diseases Detection and Classification Using Various Cnn Techniques; Springer Science and Business Media Deutschland GmbH: Singapore, 2021; Volume 140. [Google Scholar]
  186. Karunanidhi, B.; Nandhini, B.; Ruth Jeba Kumari, S.; Sirpikadevi, M.; Sugassini, M. Plant Disease Detection and Classification Using Deep Learning CNN Algorithms. In Proceedings of the 2022 IEEE International Conference on Electronics, Computing and Communication Technologies, CONECCT 2022, Bangalore, India, 9–11 July 2021; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2022. [Google Scholar]
  187. Diah, N.M.; Jahim, M.L.M.; Roslan, N.A.M.; Ibrahim, Z.; Abdullah, A. Development of Mobile Application for Plant Disease Recognition Using Convolutional Neural Network Method. In Proceedings of the 2021 6th IEEE International Conference on Recent Advances and Innovations in Engineering, ICRAIE 2021, Kedah, Malaysia, 1–3 December 2021; Institute of Electrical and Electronics Engineers Inc.: Piscataway, NJ, USA, 2022. [Google Scholar]
  188. Janarthan, S.; Thuseethan, S.; Rajasegarar, S.; Yearwood, J. P2OP—Plant Pathology on Palms: A Deep Learning-Based Mobile Solution for in-Field Plant Disease Detection. Comput. Electron. Agric. 2022, 202, 107371. [Google Scholar] [CrossRef]
  189. Rishiikeshwer, B.S.; Shriram, T.A.; Raju, J.S.; Hari, M.; Santhi, B.; Brindha, G.R. Farmer-Friendly Mobile Application for Automated Leaf Disease Detection of Real-Time Augmented Data Set Using Convolution Neural Networks. J. Comput. Sci. 2020, 16, 158–166. [Google Scholar] [CrossRef]
  190. Angin, P.; Anisi, M.H.; Göksel, F.; Gürsoy, C.; Büyükgülcü, A. Agrilora: A Digital Twin Framework for Smart Agriculture. J. Wirel. Mob. Netw. Ubiquitous Comput. Dependable Appl. 2020, 11, 77–96. [Google Scholar] [CrossRef]
  191. Daniya, T.; Vidyadhari, C.; Aluri, S. Rice Plant Disease Detection Using Sensing Recognition Strategy Based on Artificial Intelligence. J. Mob. Multimed. 2022, 18, 705–722. [Google Scholar] [CrossRef]
  192. van den Bosch, F.; Paveley, N.; Shaw, M.; Hobbelen, P.; Oliver, R. The Dose Rate Debate: Does the Risk of Fungicide Resistance Increase or Decrease with Dose? Plant Pathol. 2011, 60, 597–606. [Google Scholar] [CrossRef]
Agronomy 14 02719 g001
Figure 1. Temporal dynamics of research trends in crop disease detection topics.
Figure 1. Temporal dynamics of research trends in crop disease detection topics.
Agronomy 14 02719 g001
Agronomy 14 02719 g002
Figure 2. Flowchart of the methodology for collecting data about plant disease identification using artificial intelligence and image data.
Figure 2. Flowchart of the methodology for collecting data about plant disease identification using artificial intelligence and image data.
Agronomy 14 02719 g002
Agronomy 14 02719 g003
Figure 3. Thematic clustering of the topics identified from the keywords as well as their temporal evolution based on data collected from 2000 to 2024. The node size of a topic is proportionate to its relative frequency.
Figure 3. Thematic clustering of the topics identified from the keywords as well as their temporal evolution based on data collected from 2000 to 2024. The node size of a topic is proportionate to its relative frequency.
Agronomy 14 02719 g003
Agronomy 14 02719 g004
Figure 4. Thematic clustering of the topics identified from the keywords based on data collected from 2000 to 2024. The node size of a topic is proportionate to its relative frequency.
Figure 4. Thematic clustering of the topics identified from the keywords based on data collected from 2000 to 2024. The node size of a topic is proportionate to its relative frequency.
Agronomy 14 02719 g004
Agronomy 14 02719 g005
Figure 5. Sankey diagram analysis of global research trends in crop disease detection using artificial intelligence and image processing for the period between 2000 and 2024.
Figure 5. Sankey diagram analysis of global research trends in crop disease detection using artificial intelligence and image processing for the period between 2000 and 2024.
Agronomy 14 02719 g005
Table 1. Inclusion and exclusion criteria used to define the search area of the review.
Table 1. Inclusion and exclusion criteria used to define the search area of the review.
Inclusion CriteriaExclusion Criteria
2000–2024Before 2000
Peer-review and grey literatureWebsites and project pages
EnglishBeyond English
ScopusOther search engines
Keywords: plant disease detection, artificial intelligence, image, sensingOther plant disease biology-based detection methods
Table 2. Examples of studies performed on crop disease detection via machine learning techniques listed by crop type.
Table 2. Examples of studies performed on crop disease detection via machine learning techniques listed by crop type.
CropDiseaseStudy Scale *Acquisition Methods **Citations
Almond (Prunus amygdalus)Leaf blotch (Polystigma ochraceum)LeavesUAV (RGB)[52]
Avocado (Persea americana)Laurel wilt (Raffaelea lauricola); Phytophtora root rot (Phytophthora spp.)TreeMultispectral camera[53]
Apple (Malus domestica)Alternaria leaf blotch (Alternaria spp.), Apple black spot (Venturia inaequalis), and apple leaf miner (Lyonetia clerkella)Leaf,
Tree		Photo (Plant Village)
Own photo database (VIS)		[54,55,56,57,58]
Asparagus (Asparagus officinalis)Purple spot disease (Stemphylium vesicarium)PlantPhoto (VIS & IR) and Satellite[59]
Banana (Musa)Banana fusarium wilt (Fusarium oxysporum f. sp. cubense)PlantUAV[60,61,62]
Barley (Hordeum vulgare)Powdery mildew (Blumeria graminis f. sp. hordei)LeafRGB[63]
Bean (Phaseolus vulgaris)Rust (Uromyces phaseoli var. typica) and Angular leaf spot (Pseudocercospora griseola)LeafPhoto (VIS)[64,65]
Cardamon (Elettaria cardamomum)Colletotrichum Blight (Colletotrichum gloeosporioides) and Phyllosticta Leaf Spot (Phyllosticta capitalensis)LeavesPhoto (VIS)[66]
Cassava (Manihot esculenta)Brown leaf spot (Mycosphaerella henningsii), Red mite damage (Tetranychus urticae), Green mite damage (Mononychellus tanajoa), Cassava brown streak virus, Cassava mosaic virusLeafPhoto (VIS)
Photo (Kaggle)		[67,68]
CitrusCitrus black spot (Phyllosticta citricarpa), Citrus bacterial canker (Xanthomonas citri subsp. Citri), Huanglongbing citrus greening (Candidatus Liberibacter asiaticus)Leaves,
Leaf		Photo (RGB)
Hyperspectral		[69,70,71,72,73,74]
Coffee (Coffea arabica)Leaf miner (Leucoptera caffeine) and Rust (Hemileia vastatrix)LeavesUAV [75]
Cotton (Gossypium hirsutum)SeveralLeaf,
Plants		Photo (VIS)
Photo (Plant Village)
UAV		[76,77,78,79]
Cucumber (Cucumis sativus)SeveralLeafPhoto (Plant Village)[80]
Durian (Durio zibethinus)SeveralLeafPhoto (VIS)[81]
Eggplant (Solanum melongena)Fruit rot, Alternaria leaf spot (Alternaria sp.), Little leaf of Brinjal (phytoplasma), Mosaic virus, Collar rot (Sclerotinia sclerotiorum)Leaves,
Fruit,
Plant		Photo (VIS)[82]
Grapevine (Vitis vinifera)Grapevine flavescence dorée phytoplasma, Yellows, Esca (Phaeomoniella chlamydospora, Phellinus punctatus, Fomitiporia mediterranea, Phaeoacremonium minimum), Downy mildew (Plasmopara viticola)Leaves
Leaf		RGB; Spectroradiometer
Photo (in vitro)
Photo (Plant Village)		[36,83,84,85,86,87]
Guava (Psidium guajava L.)Guava rust (Austropuccinia psidii), Scabby fruit canker (Pestalotia psidii), Mummy disease (Gloeosporium Psidii)Fruits & LeavesPhoto (VIS)[61,88]
Maize (Zea mais L.)Northern corn leaf blight (Exserohilum turcicum), Southern corn leaf blight (Bipolaris maydis), Common rust (Puccinia sorghi)LeafPhoto (VIS)[89,90,91,92,93,94,95]
Mango (Mangifera indica L.)Sooty mould (Capnodium salicinum)Leaf, LeavesPhoto (Plant Village, leaf snap)[96,97,98]
Melon (Cucumis melo L.)Powdery mildew (Sphaerotheca fuliginea)LeavesPhoto (VIS)
UAV		[37,99]
Mulberry (Morus nigra)Leaf rust (Peridiospora mori) and Leaf spot (Mycosphaerella mori)LeafPhoto (VIS)[100]
Oil palm (Elaeis guineensis)Basal stem rot of oil palm (Ganoderma boninense) LeafPhoto (VIS)
FTIR and Raman spectroscopy		[8,101]
Olive tree (Olea europaea)SeveralLeafPhoto (VIS)[102,103]
Onion (Allium cepa L.)Onion Smudge (Colletotrichum circinans) Satellite (VIS-NIR)[104,105]
Papaya (Carica papaya L.)Begomovirus (Geminiviridae)LeafNIR and FT-IR ATR[106]
Pea (Pisum sativum L.)Rust disease (Uromyces viciae-fabae Pers. de-Bary)LeafMicroscopic images[107]
Peanuts (Arachis hypogaea)Peanut stem rot (Athelia rolfsii)LeavesUV, VIS, NIR, Thermal[108]
Pepper (Capsicum spp.)Pepper yellow leaf curl virus (PepYLClV),
Several		Leaf, Pulp, StemFT-IR
Photo (VIS, Plant Village + pepper diseased dataset)		[109,110,111,112,113]
Pigeon pea (Cajanus cajan)Fusarium wilt (Fusarium udum), Pigeonpea sterility mosaic virus (PPSMV), Ashy stem blight (Macrophomina phaseolina), Phytophthora blight (Phytophthora drechsleri f. sp. cajani)LeafPhoto (VIS)[114]
Plum (Prunus subg. Prunus)SeveralLeaves, Leaves & FruitPhoto (VIS)[115]
Potato (Solanum tuberosum)Early blight (Alternaria solani), Late blight of potato (Phytophthora infestans)LeafPhoto (Plant Village)[116,117,118,119]
Rapeseed (Brassica napus L.)Sclerotinia stem rot (Sclerotinia sclerotiorum)LeavesHyperspectral[120]
Rice (Oryza sativa)Bacterial leaf blight (Xanthomonas oryzae pv. oryzae), Brown spot of rice (Cochliobolus miyabeanus), Tungrovirus oryzae, Entyloma oryzae (leaf smut of rice)Leaves,
Panicle		Photo (VIS)[121,122,123,124,125,126,127,128,129,130]
Rose (Rosa sp.)Powdery mildew of rose (Podosphaera pannosa) and Gray mold of roses (Botrytis cinerea)LeafThermal and visible images[131]
SolanumBlight (n.d.), severalLeafPhoto (Plant Village)[132,133,134]
Soybean (Glycine max.)Nematodes cyst nematode,
Anthracnose of soybean (Colletotrichum truncatum)
Wildfire (Pseudomonas syringae pv. tabaci)		Field,
Pods,
Leaves,		Satellite (hyperspectral)
VIS+NIR		[135,136,137]
Squash (Cucurbita)Cucurbit powdery mildew (Podosphaera xanthii)PlantUAV (multispectral)[138]
Strawberry (Fragaria x ananassa)Anthracnose (Colletotrichum fragariae), Gray mold of strawberries (Botrytis cinerea)LeafPhoto (VIS)
Hyperspectral		[139,140,141]
Sugar beet (Beta vulgaris)Cercospora leaf spot (Cercospora beticola)PlantsUAV[19,142,143]
Sugarcane (Saccharum officinarum)Orange rust disease of sugarcane (Puccinia kuehnii), Sugarcane yellow leaf virus (ScYLV), Eye spot disease of sugarcane (Helminthosporium saccahari), Brown leaf spot of sugarcane (Cercospora longipies), Red rot of sugarcane (Colletotrichum falcatum)LeafPhoto (VIS)[144,145]
Tomato (Solanum lycopersicum)10 diseases, early blight (Alternaria solani),
Tomato Spotted Wilt Virus (TSWV)		LeafPhoto (Plant Village)
Own database
VIS + NIR
Hyperspectral		[24,146,147,148,149,150,151,152,153,154,155]
Green Tea (Camellia sinensis)Blister blight of tea (Exobasidium vexans), Leafhopper, Caterpillars, Mosquito, Yellow miteLeafPhoto (VIS)[156]
Walnut (Juglans regia L.)Diseased (n.d.)LeafPhoto (VIS)[157]
Wheat (Triticum aestivum L.)Wheat stripe (yellow) rust (Puccinia striiformis f. sp. tritici), Wheatbrown rust (Puccinia triticina)
Fusarium head blight (Fusarium graminearum)		Field,
Spikes,
Leaves		UAV VIS
Hyperspectral + Fluorescence
Photo (VIS)		[20,158,159,160,161,162,163,164,165,166,167]
* Fruit, leaf, panicle, pulp, stem: a single organ photo in a neutral environment; fruits, leaves, spikes: several same parts of a plant pictured together; plant, tree: a single entire plant of a crop in an image with a neutral background; plants: several entire plants together; field: photo take above the crop, leaves, stems and soil included in the photo.** RGB: Red Green Blue; VIS: visible; NIR spectroscopy; Near Infrared; FT-IR: Fourier-Transform Infrared spectroscopy; ATR: Attenuated Total Reflection; UAV: Unmanned Aerial Vehicle; n.d.: not defined.
Table 3. Frequently used image annotation tools.
Table 3. Frequently used image annotation tools.
ToolSourceAdvantages/Constraints
LabelImghttps://github.com/heartexlabs/labelImg (accessed on 10 September 2024)Python based, open-source, needs programming skills
VGG Image Annotator (VIA)https://www.robots.ox.ac.uk/~vgg/software/via/ (accessed on 10 September 2024)Web-based tool, open-source, requires internet connection, some issues with large dataset
Labelboxhttps://labelbox.com/customers/genetech-customer-story/ (accessed on 10 September 2024)Python based, commercial, complex feature set
SLOTHhttps://github.com/cvhciKIT/sloth (accessed on 10 September 2024)Python based, open-source, not supported an all platforms, some issues with large dataset
Hastyhttps://hasty.ai/v2 (accessed on 10 September 2024) Cloud Annotations GUI, Commercial, auto label function
IBM Cloud Annotations Toolhttps://developer.ibm.com/blogs/ibm-cloud-annotations-tool-eases-the-process-of-ai-data-labeling/ (accessed on 10 September 2024)Cloud Annotations GUI, auto label function
RectLabelhttps://github.com/ryouchinsa/Rectlabel-support (accessed on 10 September 2024) Support only Linux environment, commercial
Labelmehttps://developer.ibm.com/blogs/ibm-cloud-annotations-tool-eases-the-process-of-ai-data-labeling/ (accessed on 10 September 2024)Python, based, open-source, some issues with large dataset
Scalehttps://scale.com/image (accessed on 10 September 2024)Cloud Annotations GUI, auto label function, commercial
SUPERVISELYhttps://supervise.ly/ (accessed on 10 September 2024)Cloud Annotations GUI, auto label function, commercial
Table 4. Frequently used algorithms in plant disease detection within the references in the constituted database.
Table 4. Frequently used algorithms in plant disease detection within the references in the constituted database.
AlgorithmsUse Frequency (%)Example References
CNN37[174,175,176,177]
DCNN3[88]
R-CNN4[178]
Mask R-CNN1[158]
Faster R-CNN2[70]
YOLO3[70,146,179]
RFCN3[180]
GAN18[24,181,182]
VGG13[183]
SVM16[184]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lebrini, Y.; Ayerdi Gotor, A. Crops Disease Detection, from Leaves to Field: What We Can Expect from Artificial Intelligence. Agronomy 2024, 14, 2719. https://doi.org/10.3390/agronomy14112719

AMA Style

Lebrini Y, Ayerdi Gotor A. Crops Disease Detection, from Leaves to Field: What We Can Expect from Artificial Intelligence. Agronomy. 2024; 14(11):2719. https://doi.org/10.3390/agronomy14112719

Chicago/Turabian Style

Lebrini, Youssef, and Alicia Ayerdi Gotor. 2024. "Crops Disease Detection, from Leaves to Field: What We Can Expect from Artificial Intelligence" Agronomy 14, no. 11: 2719. https://doi.org/10.3390/agronomy14112719

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop