YOLOv8n-WSE-Pest: A Lightweight Deep Learning Model Based on YOLOv8n for Pest Identification in Tea Gardens
<p>Collection of original images of pests.</p> "> Figure 2
<p>Partial original image samples.</p> "> Figure 3
<p>Data augmentation.</p> "> Figure 4
<p>Examples of low-quality images. Note: (<b>a</b>) illustrates severe overexposure and blur, which can obscure essential details; (<b>b</b>) presents an instance where the subject is unidentifiable, potentially due to occlusion or poor resolution; (<b>c</b>) shows a case of missing targets, which is critical, as it pertains to the absence of the object of interest within the frame; (<b>d</b>) displays reflective or mutilated images, which can lead to misinterpretation by the detection model.</p> "> Figure 5
<p>Improved YOLOv8n network architecture. Note: <a href="#applsci-14-08748-f005" class="html-fig">Figure 5</a> illustrates the network architecture of YOLOv8n-WSE-pest in this study. The upper part depicts the hierarchical structure of the network, while the right side elucidates the working principle of the original components within the network. Specifically, the Binary Cross-Entropy (BCE) loss function is employed for binary classification tasks, quantifying the discrepancy between the predicted and actual class labels. The Distribution Focal Loss (DFL) transforms the bounding box regression problem in object detection into a sequence prediction problem, thereby enhancing the detection accuracy in scenarios where the targets exhibit boundary ambiguity or occlusion. Additionally, the shortcut operation within the Bottleneck component facilitates skip connections in feature maps, contributing to feature fusion and the stability of network training.</p> "> Figure 6
<p>WIoU-v3 schematic diagram. Note: On the <b>left</b> side of the figure are anchor boxes of three different qualities. On the <b>right</b> side, the method is demonstrated to more accurately assess the quality of anchor boxes by predicting the relative position and size differences between the predicted boxes and the ground truth boxes.</p> "> Figure 7
<p>Diagram of SCConv’s overall structure.</p> "> Figure 8
<p>SRU schematic diagram.</p> "> Figure 9
<p>CRU schematic diagram.</p> "> Figure 10
<p>EMA algorithm principle diagram.</p> "> Figure 11
<p>Model loss function: (<b>a</b>–<b>c</b>) represent the trend graphs of box_loss, cls_loss, and dfl_loss for the YOLOv8n-WSE-pest model, respectively; (<b>d</b>–<b>f</b>) represent the trend graphs of box_loss, cls_loss, and dfl_loss for the original YOLOv8n model, respectively; (<b>a</b>,<b>d</b>) compare the box_loss between the validation set and the training set; (<b>b</b>,<b>e</b>) represent the cls_loss comparison between the validation set and the training set; (<b>c</b>,<b>f</b>) depict the dfl_loss comparison between the validation set and the training set.</p> "> Figure 12
<p>Comparison between YOLOv8n-WSE-pest vs. original YOLOv8n. Note: This graphical representation showcases a comparative assessment of the foundational YOLOv8 model against a tailored variant, YOLOv8-WSE-pest, which is specifically adapted for pest recognition tasks. The depicted analysis spans three pivotal performance metrics—precision, recall, and F1 score—across a spectrum of confidence threshold values. Each distinct curve within the illustration corresponds to a separate pest category, while the overarching blue line signifies the aggregated average performance across all pest types considered. Progressing along the horizontal axis, which denotes the incremental confidence thresholds, the vertical axis records the respective scores of the outlined evaluative criteria. This meticulous comparison serves to highlight the augmented effectiveness of the YOLOv8-WSE-pest model in consistently identifying and classifying diverse pest species under a broad range of confidence levels, thereby affirming its advancement in specialized detection capabilities.</p> "> Figure 13
<p>Grad-CAM results for pest identification.</p> "> Figure 14
<p>Comparison of actual image detection.</p> "> Figure 15
<p>AP recognized using different models in different categories.</p> ">
Abstract
:1. Introduction
2. Materials and Methods
2.1. Data Collection
2.2. Data Augmentation
2.3. YOLOv8 Network Improvement
2.3.1. Improvement of Loss Function
2.3.2. Spatial and Channel Reconstruction Convolution for Feature Redundancy
2.3.3. Efficient Multi-Scale Attention Module with Cross-Spatial Learning
2.3.4. Experimental Setup and Evaluation Metrics for YOLOv8n-WSE-Pest Model Accuracy
3. Results
3.1. Analysis of Model Training Results
3.2. Experimental Analysis of Detection Model
3.2.1. Ablation Experiment
3.2.2. Comparative Model Experiment
4. Discussion
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Xia, Y.; Wang, Z.; Cao, Z.; Chen, Y.; Li, L.; Chen, L.; Zhang, S.; Wang, C.; Li, H.; Wang, B. Recognition Model for Tea Grading and Counting Based on the Improved YOLOv8n. Agronomy 2024, 14, 1251. [Google Scholar] [CrossRef]
- Wang, Z.; Yang, C.; Che, R.; Li, H.; Chen, Y.; Chen, L.; Yuan, W.; Yang, F.; Tian, J.; Wang, B. Assisted Tea Leaf Picking: The Design and Simulation of a 6-DOF Stewart Parallel Lifting Platform. Agronomy 2024, 14, 844. [Google Scholar] [CrossRef]
- Jiang, C.; Zeng, Z.; Huang, Y.; Zhang, X. Chemical compositions of Pu’er tea fermented by Eurotium cristatum and their lipid-lowering activity. LWT 2018, 98, 204–211. [Google Scholar] [CrossRef]
- Chen, N.; Zhu, J.; Zheng, L. Light-YOLO: A Study of a Lightweight YOLOv8n-Based Method for Underwater Fishing Net Detection. Appl. Sci. 2024, 14, 6461. [Google Scholar] [CrossRef]
- Zhang, Z.; Zhou, C.; Xu, Y.; Huang, X.; Zhang, L.; Mu, W. Effects of intercropping tea with aromatic plants on population dynamics of arthropods in Chinese tea plantations. J. Pest Sci. 2017, 90, 227–237. [Google Scholar] [CrossRef]
- Somnath, R.; Gautam, H.; Narayanannair, M.; Kavya, D.; Mukhopadhyay, R.S.; Ananda, M.; Azariah, B. Use of plant extracts for tea pest management in India. Appl. Microbiol. Biot. 2016, 100, 4831–4844. [Google Scholar]
- Liao, Y.; Yu, Z.; Liu, X.; Zeng, L.; Cheng, S.; Li, J.; Tang, J.; Yang, Z. Effect of Major Tea Insect Attack on Formation of Quality-Related Nonvolatile Specialized Metabolites in Tea (Camellia sinensis) Leaves. J. Agr. Food Chem. 2019, 67, 6716–6724. [Google Scholar] [CrossRef]
- Chen, X.; Hassan, M.M.; Yu, J.; Zhu, A.; Han, Z.; He, P.; Chen, Q.; Li, H.; Ouyang, Q. Time series prediction of insect pests in tea gardens. J. Sci. Food Agric. 2024, 104, 5614–5624. [Google Scholar] [CrossRef]
- Blackie, H.M.; MacKay, J.W.; Allen, W.J.; Smith, D.H.V.; Barrett, B.; Whyte, B.I.; Murphy, E.C.; Ross, J.; Shapiro, L.; Ogilvie, S.; et al. Innovative developments for long-term mammalian pest control. Pest Manag. Sci. 2014, 70, 345–351. [Google Scholar] [CrossRef]
- Schellhorn, N.A.; Parry, H.R.; Macfadyen, S.; Wang, Y.; Zalucki, M.P. Connecting scales: Achieving in-field pest control from areawide and landscape ecology studies. Insect Sci. 2015, 22, 35–51. [Google Scholar] [CrossRef]
- Baruah, P.M.; Bardaloi, S.; Bordoloi, S. A comparative survey of the pest prevalence and chemical control practices in the tea gardens of Sonitpur district of Assam. Int. J. Phys. Soc. Sci. 2015, 5, 22–32. [Google Scholar]
- He, J.; Zhang, S.; Yang, C.; Wang, H.; Gao, J.; Huang, W.; Wang, Q.; Wang, X.; Yuan, W.; Wu, Y.; et al. Pest recognition in microstates state: An improvement of YOLOv7 based on Spatial and Channel Reconstruction Convolution for feature redundancy and vision transformer with Bi-Level Routing Attention. Front. Plant Sci. 2024, 15, 1327237. [Google Scholar] [CrossRef] [PubMed]
- Sun, L.; Cai, Z.; Liang, K.; Wang, Y.; Zeng, W.; Yan, X. An intelligent system for high-density small target pest identification and infestation level determination based on an improved YOLOv5 model. Expert Syst. Appl. 2024, 239, 122190. [Google Scholar] [CrossRef]
- Li, H.; Shi, H.; Du, A.; Mao, Y.; Fan, K.; Wang, Y.; Shen, Y.; Wang, S.; Xu, X.; Tian, L.; et al. Symptom recognition of disease and insect damage based on Mask R-CNN, wavelet transform, and F-RNet. Front. Plant Sci. 2022, 13, 922797. [Google Scholar] [CrossRef] [PubMed]
- Magsi, F.H.; Cai, X.; Luo, Z.; Li, Z.; Bian, L.; Xiu, C.; Fu, N.; Li, J.; Hall, D.R.; Chen, Z. Identification, synthesis, and field evaluation of components of the female-produced sex pheromone of Helopeltis cinchonae (Hemiptera: Miridae), an emerging pest of tea. Pest Manag. Sci. 2024. [Google Scholar] [CrossRef] [PubMed]
- Jiang-Lin, Q.; Xiu-Hao, Y.; Zhong-Wu, Y.; Ji-Tong, L.; Xiu-Feng, L. New technology for using meteorological information in forest insect pest forecast and warning systems. Pest Manag. Sci. 2017, 73, 2509–2518. [Google Scholar]
- Banik, A.; Chattopadhyay, A.; Ganguly, S.; Mukhopadhyay, S.K. Characterization of a tea pest specific Bacillus thuringiensis and identification of its toxin by MALDI-TOF mass spectrometry. Ind. Crops Prod. 2019, 137, 549–556. [Google Scholar] [CrossRef]
- Li, J.; Zhou, Y.; Zhou, B.; Tang, H.; Chen, Y.; Qiao, X.; Tang, J. Habitat management as a safe and effective approach for improving yield and quality of tea (Camellia sinensis) leaves. Sci. Rep. 2019, 9, 433. [Google Scholar] [CrossRef]
- Ju, C.C.; Yu, H.Y.; Shuo, L.Y.; Yu, C.C.; Min, H.Y. An AIoT Based Smart Agricultural System for Pests Detection. IEEE Access 2020, 8, 180750–180761. [Google Scholar]
- Gao, D.; Sun, Q.; Hu, B.; Zhang, S. A Framework for Agricultural Pest and Disease Monitoring Based on Internet-of-Things and Unmanned Aerial Vehicles. Sensor 2020, 20, 1487. [Google Scholar] [CrossRef]
- Ali, M.A.; Sharma, A.K.; Dhanaraj, R.K. Heterogeneous features and deep learning networks fusion-based pest detection, prevention and controlling system using IoT and pest sound analytics in a vast agriculture system. Comput. Electr. Eng. 2024, 116, 109146. [Google Scholar] [CrossRef]
- Brunelli, D.; Albanese, A.; d’Acunto, D.; Nardello, M. Energy Neutral Machine Learning Based IoT Device for Pest Detection in Precision Agriculture. IEEE Internet Things Mag. 2019, 2, 10–13. [Google Scholar] [CrossRef]
- Kiobia, D.O.; Mwitta, C.J.; Fue, K.G.; Schmidt, J.M.; Riley, D.G.; Rains, G.C. A Review of Successes and Impeding Challenges of IoT-Based Insect Pest Detection Systems for Estimating Agroecosystem Health and Productivity of Cotton. Sensor 2023, 23, 4127. [Google Scholar] [CrossRef] [PubMed]
- Wang, Y.; Xu, R.; Bai, D.; Lin, H. Integrated Learning-Based Pest and Disease Detection Method for Tea Leaves. Forests 2023, 14, 1012. [Google Scholar] [CrossRef]
- Xiaohu, Z.; Jingcheng, Z.; Yanbo, H.; Yangyang, T.; Lin, Y. Detection and discrimination of disease and insect stress of tea plants using hyperspectral imaging combined with wavelet analysis. Comput. Electron. Agric. 2022, 193, 106717. [Google Scholar]
- Qingwen, G.; Chuntao, W.; Deqin, X.; Qiong, H. Automatic monitoring of flying vegetable insect pests using an RGB camera and YOLO-SIP detector. Precis. Agric. 2022, 24, 436–457. [Google Scholar]
- Cheng, X.; Zhang, Y.; Chen, Y.; Wu, Y.; Yue, Y. Pest identification via deep residual learning in complex background. Comput. Electron. Agric. 2017, 141, 351–356. [Google Scholar] [CrossRef]
- Yunong, T.; Shihui, W.; En, L.; Guodong, Y.; Zize, L.; Min, T. MD-YOLO: Multi-scale Dense YOLO for small target pest detection. Comput. Electron. Agric. 2023, 213, 108233. [Google Scholar]
- Dong, S.; Teng, Y.; Jiao, L.; Du, J.; Liu, K.; Wang, R. ESA-Net: An efficient scale-aware network for small crop pest detection. Expert Syst. Appl. 2024, 236, 121308. [Google Scholar] [CrossRef]
- Tang, Z.; Chen, Z.; Qi, F.; Zhang, L.; Chen, S. Pest-YOLO: Deep Image Mining and Multi-Feature Fusion for Real-Time Agriculture Pest Detection; IEEE: New York, NY, USA, 2021; pp. 1348–1353. [Google Scholar]
- Chen, Z.; Zhou, H.; Lin, H.; Bai, D. TeaViTNet: Tea Disease and Pest Detection Model Based on Fused Multiscale Attention. Agronomy 2024, 14, 633. [Google Scholar] [CrossRef]
- Shihao, Z.; Hekai, Y.; Chunhua, Y.; Wenxia, Y.; Xinghui, L.; Xinghua, W.; Yinsong, Z.; Xiaobo, C.; Yubo, S.; Xiujuan, D.; et al. Edge Device Detection of Tea Leaves with One Bud and Two Leaves Based on ShuffleNetv2-YOLOv5-Lite-E. Agronomy 2023, 13, 577. [Google Scholar] [CrossRef]
- Chataut, R.; Phoummalayvane, A.; Akl, R. Unleashing the Power of IoT: A Comprehensive Review of IoT Applications and Future Prospects in Healthcare, Agriculture, Smart Homes, Smart Cities, and Industry 4.0. Sensor 2023, 23, 7194. [Google Scholar] [CrossRef] [PubMed]
- Tan, S.-Y.; Hong, F.; Ye, C.; Wang, J.-J.; Wei, D. Functional characterization of four Hsp70 genes involved in high-temperature tolerance in Aphis aurantii (Hemiptera: Aphididae). Int. J. Biol. Macromol. 2022, 202, 141–149. [Google Scholar] [CrossRef] [PubMed]
- Laijin, L. Be on Guard against one of the Destructive Pest: Xyleborus fornicatus. Plant Prot. Technol. Ext. 1999, 19, 23. [Google Scholar]
- Guo, Z.M.; Cheng, F.Y.; Ma, M.J.; Tu, X.L. Advances in green control technology of Empoasca pirisuga Matumura in tea region of south Hubei province. Hubei Agric. Sci. 2019, 58, 9–12. [Google Scholar]
- Deka, B.; Babu, A. Tea Pest Management: A Microbiological Approach. Appl. Microbiol. Open Access 2021, 7, 206. [Google Scholar]
- Ivan, B.; Oscar, A.; Cristina, C.; Josep, P.; Stéphane, B.; Lorena, G.; Nuria, A. Development of a multi-primer metabarcoding approach to understanding trophic interactions in agroecosystems. Insect Sci. 2021, 29, 1195–1210. [Google Scholar]
- Li, X.Y.; Xie, M.; Dong, P.X.; Liang, X.C. Morphology of Pyrocoelia pygidialis Pic(Coleoptera: Lampyridae) with Notes on Its Biology. J. Insect Classif. 2008, 30, 300–308. [Google Scholar]
- Hussain, M. YOLO-v1 to YOLO-v8, the rise of YOLO and its complementary nature toward digital manufacturing and industrial defect detection. Machines 2023, 11, 677. [Google Scholar] [CrossRef]
- Luo, Z.; Wang, C.; Qi, Z.; Luo, C. LA_YOLOv8s: A lightweight-attention YOLOv8s for oil leakage detection in power transformers. Alex. Eng. J. 2024, 92, 82–91. [Google Scholar] [CrossRef]
- Chenghao, W.; Zhongqiang, L.; Ziyuan, Q. Transformer oil leakage detection with sampling-WIoU module. J. Supercomput. 2023, 80, 7349–7368. [Google Scholar]
- Yi-Fan, Z.; Weiqiang, R.; Zhang, Z.; Zhen, J.; Liang, W.; Tieniu, T. Focal and efficient IOU loss for accurate bounding box regression. Neurocomputing 2022, 506, 146–157. [Google Scholar]
- Tong, Z.; Chen, Y.; Xu, Z.; Yu, R. Wise-IoU: Bounding box regression loss with dynamic focusing mechanism. arXiv 2023, arXiv:2301.10051. [Google Scholar]
- Li, J.; Wen, Y.; He, L. in Scconv: Spatial and Channel Reconstruction Convolution for Feature Redundancy. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada, 17–24 June 2023; pp. 6153–6162. [Google Scholar]
- Hu, J.; Fan, T.; Tang, X.; Yang, Z.; Ren, Y. Nonlinear relations of urban morphology to thermal anomalies: A cross-time comparative study based on Grad-CAM and SHAP. Ecol. Indic. 2024, 162, 112024. [Google Scholar] [CrossRef]
Pest Species | Training Dataset | Verification Dataset | Test Dataset |
---|---|---|---|
Toxoptera aurantii | 801 | 264 | 267 |
Xyleborus fornicatus Eichhoffr | 795 | 269 | 271 |
Empoasca pirisuga Matumura | 792 | 258 | 262 |
Malthodes discicollis Baudi di Selve | 806 | 273 | 265 |
Total | 3194 | 1064 | 1065 |
Algorithm | WIoU-v3 | SCConv | EMA | Characteristics |
---|---|---|---|---|
YOLOv8n | Baseline | |||
YOLOv8n-W | Precision focused | |||
YOLOv8n-S | Efficiency | |||
YOLOv8n-E | High performance | |||
YOLOv8n-WS | Balanced | |||
YOLOv8n-WE | Enhanced | |||
YOLOv8n-SE | Simplified | |||
YOLOv8n-WSE-pest | Optimized |
Algorithm | Precision/% | Recall/% | mAP50/% | Layers | Parameters | Gradients | GFLOPs |
---|---|---|---|---|---|---|---|
YOLOv8n | 91.96 | 88.54 | 95.77 | 225 | 3,157,200 | 3,157,184 | 8.9 |
YOLOv8n-W | 92.36 | 91.03 | 96.56 | 225 | 3,157,200 | 3,157,184 | 8.9 |
YOLOv8n-S | 92.98 | 91.31 | 96.78 | 236 | 3,131,180 | 3,131,164 | 8.3 |
YOLOv8n-E | 93.65 | 95.17 | 97.88 | 233 | 11,169,184 | 11,169,168 | 29.1 |
YOLOv8n-WS | 92.67 | 91.93 | 96.47 | 236 | 3,131,180 | 3,131,164 | 8.3 |
YOLOv8n-WE | 94.14 | 96.41 | 97.16 | 233 | 11,169,184 | 11,169,168 | 29.1 |
YOLOv8n-SE | 94.29 | 95.75 | 97.42 | 244 | 3,131,852 | 3,131,836 | 8.4 |
YOLOv8n-WSE-pest | 95.08 | 94.19 | 97.95 | 244 | 3,131,852 | 3,131,836 | 8.4 |
Model Name | AP of Toxoptera aurantii/% | AP of Xyleborus fornicatus Eichhoffr/% | AP of Empoasca pirisuga Matumura/% | AP of Malthodes discicollis Baudi di Selve/% | mAP/% |
---|---|---|---|---|---|
Faster-RCNN | 84.62 | 83.78 | 82.76 | 83.25 | 83.61 |
SSD | 89.84 | 88.23 | 88.74 | 89.57 | 89.10 |
YOLOv8n | 94.07 | 95.53 | 96.12 | 97.36 | 95.77 |
YOLOv8n-WSE-pest | 97.25 | 98.11 | 97.82 | 98.62 | 97.95 |
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. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Li, H.; Yuan, W.; Xia, Y.; Wang, Z.; He, J.; Wang, Q.; Zhang, S.; Li, L.; Yang, F.; Wang, B. YOLOv8n-WSE-Pest: A Lightweight Deep Learning Model Based on YOLOv8n for Pest Identification in Tea Gardens. Appl. Sci. 2024, 14, 8748. https://doi.org/10.3390/app14198748
Li H, Yuan W, Xia Y, Wang Z, He J, Wang Q, Zhang S, Li L, Yang F, Wang B. YOLOv8n-WSE-Pest: A Lightweight Deep Learning Model Based on YOLOv8n for Pest Identification in Tea Gardens. Applied Sciences. 2024; 14(19):8748. https://doi.org/10.3390/app14198748
Chicago/Turabian StyleLi, Hongxu, Wenxia Yuan, Yuxin Xia, Zejun Wang, Junjie He, Qiaomei Wang, Shihao Zhang, Limei Li, Fang Yang, and Baijuan Wang. 2024. "YOLOv8n-WSE-Pest: A Lightweight Deep Learning Model Based on YOLOv8n for Pest Identification in Tea Gardens" Applied Sciences 14, no. 19: 8748. https://doi.org/10.3390/app14198748