ANN Based Technique For Vegetable Quality Detection
ANN Based Technique For Vegetable Quality Detection
ANN Based Technique For Vegetable Quality Detection
e-ISSN: 2278-2834,p- ISSN: 2278-8735.Volume 10, Issue 5, Ver. I (Sep - Oct .2015), PP 62-70
www.iosrjournals.org
Abstract :Use of image processing technique is increasing day by day in all fields. In agriculture it is also used
to check the quality of vegetables and fruits. Shape, colour and size are the image features which help in quality
detection of vegetables. In this paper proposed method is used to increase the accuracy of the vegetable quality
detection by using colour, shape, and size based method with combination of artificial neural network (ANN). It
grades and classifies vegetable images based on obtained feature values by using cascaded forward network.
The proposed system starts the process by capturing the vegetables image. Then, the image is transmitted to the
processing level where the vegetable features like colour, shape and size of vegetable samples are extracted.
After that by using artificial neural network vegetable images are going through the training and testing.
Artificial neural network detect the quality of vegetables by using the shape colour and size features provided at
the time of training and also the extracted features of vegetables and provides the result by comparing these
features. In this proposed paper neural network is used to detect shape, size and colour of vegetable and with
the combination of these three features the results obtained are very promising.
Keywords: Artificial neural network, Feature extraction, Graphical user interface, Image processing,
Vegetable Quality detection.
I.
Introduction
Agriculture and Horticulture plays an important role in economic development of India. Development
in agriculture is very slow as compared with development in electronic and automobile sector. Hence, there is
need to come up with some new techniques so as to forefront the agriculture sector. Still in India, the traditional
inspection of vegetables is performed by human experts. It was very time consuming and laborious task.
Grading of vegetables is necessary in evaluating agricultural produce, meeting quality standards and increasing
market value. The features that can be extracted from an image of any vegetable are its shape, colour and size.
These features help the user to grade the vegetables. In this Paper, an economic and safe way is used to analyze
the vegetable quality which is based on colour, shape and size. Vegetables should be tested via non-destructive
techniques because these are delicate materials. The most important physical property is vegetable size while
colour resembles visual property. It is also helpful in planning, packaging, transportation and marketing
operations. If the classification and grading is done through manual techniques, the process will be too slow and
sometimes it will be error prone. Human sorted the fruits and vegetables on the basis of colour, size, shape etc.
If these quality measures are mapped into automated system by using suitable programming language then the
work will be faster and error free. In recent years, computer machine vision and image processing techniques
have been found increasingly useful in the fruit industry, especially for applications in quality inspection and
shape sorting [1]. Colour and shape characteristics of vegetables are decisive for visual inspection. There are
several techniques which can be used to extract the morphological features from an image. An efficient
autonomous system for vegetable sorting must be able to adequately identify both parameters. Shape of
vegetables can easily be obtained from a digital image using classical techniques for image processing.
However, colour identification involves many physical and psychological concepts, asking it difficult to
properly model and process colour in an image. There are wide varieties of colour systems present for the
grading of vegetables based on colours. There are some techniques like Fuzzy logic, Neural Network; Based on
Colour Histogram, Genetic algorithm etc. [2]. Software development is highly important in this colour
classification system.The entire system is designed over matlab software to inspect the colour and size of the
vegetable. Colour of the vegetable is very important in classification but since due to the similarity of colours
between some vegetables, the size also helps in solving this kind of problems. The colour and size based
classification involves extracting the useful information from the vegetable surface and classify it to the
respective type. Artificial neural network (ANN) is used to detect shape, size and colour of vegetable samples.
Neural networks are encouraged from the brain system. Our brain learns from everyday life
and use it in future, neural network also works on same code. There are two stages in neural network i.e.
training and testing. It is composed of large number of highly interconnected processing elements working in
DOI: 10.9790/2834-10516270
www.iosrjournals.org
62 | Page
II.
Related Work
Nagganaur and Sannanki [1] presented the sorting and grading of fruits using image processing
techniques. The system starts the process by capturing the fruits image. Then the image is transmitted to the
matlab for feature extraction,classification and grading.both classification and grading realized by fuzzy logic
approach.
Shahzadpreet, Saurabh, Mohit, Sugam and hari singh [3] presenting appraisal on several procedures in
detection and isolating of rotten fruits. Some of methods used are neural networks,fuzzy logic,neuro-fuzzy and
support vector machines. Different physiognomies which tell us about apples superiority.
Mandeep kaur and Reecha sharma [4] presented a system in which fruit features like colour,shape and
size of fruit samples are extracted.In this paper artificial neural network is used to detect shape,size and colour
of fruit.
Bhanu pratap, Navneet, Sunil and Suriti [5] proposedan algorithm for fruit classification based on
shape,color and texture. Shape features are calculated by using edge detection.hsi,hsv can be used for color base
classification. GLCM is used to calculate texture featuresartificial neural network is used to classifies the fruits
by comparing shape,color and texture feature provided at the time of training.
Effendi, Ramli and Ghani [6] presented that the quality of fruit depends upon type of defects, skin
colour and size of fruit. In there research, they develop an image recognition system to identify the level of
maturity of Jatropha curcas fruit and classify it into various categories. The system is divided into two stages:
The first stage is a training stage that is to extract the characteristics from the pattern. The second stage is to
recognize the pattern by using the characteristics derived from the first task. Back propagation diagnosis model
is used to recognition the Jatropha curcas fruits. A back propagation diagnosis model (BPDM) is adopted to
recognize the image of the matured fruits. Colour indices associated with image pixels are used as input.
Patel, Jain and Joshi [9] presented the fruit detection using improved multiple features based
algorithm. To detect the fruit, an image processing algorithm is trained for efficient feature extraction. The
algorithm is designed with the aim of calculating different weights for features like intensity, colour, orientation
and edge of the input test image. Arivazhagan, Shebiah, Nidhyanandhan and Ganesan [7] presented an efficient
fusion of color and texture features for fruit recognition. The recognition is done by the minimum distance
classifier based upon the statistical and co-occurrence features derived from the wavelet transformed sub-bands.
Bindu Tiger and Toran verma [11] presented apple recognition techniques of normal and infected.
Proposed method classifies and recognizes apple images based on obtained features values by using two-layer
feed-forward network, with sigmoid hidden and output neurons. The toolbox supports feed forward networks,
radial basis networks, dynamic networks, self-organizing maps, and other proven network paradigms. This work
represents the MATLAB 7.8.0 software and the recognition of generated signals by artificial neural network
technique.
Sandoval, Prieto and Betancur [12] have proposed a machine vision based classification system to sort
coffee fruits (cherries) according their ripeness stage is presented. Eight categories were defined and they
include the entire coffee cherry ripeness process, from the initial stage (early green) to over ripe and dry stages.
A Bayesian classifier was implemented using a set of nine features which include color,shape and texture
computed on an image of the fruit.
Amir Alipasandi, Hosein Ghaffari and Saman Zohrabi Alibeyglu [21] introduced a system that is using
machine vision algorithm and neural network classifier to classify three varieties of peach fruit. Three cultivatrs,
namely, Anjiri peach cultivar and Shalil Nectarine cultivar, varieties of Iran and Elberta peach cultivar variety of
United States were randomly handpicked in two stages of growth, immature and mature. An image capturing
system was designed to provide an enclosed and uniform light illumination and to obtain standard images from
the samples.the images were sent to processing toolboxes of matlab software to visualize, acquire and process
the images directly from the computer. Some qualitative information is extracted from the objects to be analyzed
in the images. This information was used as inputs to the algorithms for classifying the objects into different
categories. In this study feature vector that consider as network input consist of 12 components of colour spaces
DOI: 10.9790/2834-10516270
www.iosrjournals.org
63 | Page
III.
Proposed Methodology
This proposed automated system is designed to overcome the problems of manual techniques. Different
vegetables images which are used in this experiment are captured under constant light source. Proposed
methodology is as follows firstly the image of vegetable is captured and after that from the captured image
various features such as shape, colour and size are extracted. Size features are extracted in height and width.
After the features are extracted then artificial neural network is used to detect the quality of vegetables. The
system consists of several steps like feature extraction, sorting and grading.
There are seven Steps for the vegetable quality detection in proposed methodology. These steps are as
following:
Step 1: Get image of vegetable.
Step 2: This image is loaded into the matlab.
Step 3: Extract the features of vegetable sample.
Step 4: Train the neural network.
Step 5: Select the vegetable sample for testing.
Step 6: Perform testing by using artificial neural network training module button.
Step 7: Artificial neural network based output
The first step is to getting the image of vegetable. Image of the vegetable samples are captured by using
regular digital camera with white background with the help of a stand. Then in the second step the image of the
vegetable is loaded into the matlab. In third step features of the vegetable samples are extracted. Features such
as colour, shape and size of the vegetable sample are extracted. In fourth step neural network is used for training
the data, after that in step fifth vegetable sample is selected for testing from database. In step sixth testing is
performed by using ANN training module button. Finally, in step seventh ANN based results are obtained [4].
There are different modules which will perform different operations on the image being loaded. The
modules are described as below:
3.1 Image capture
An image of the vegetable is captured by using any digital camera or any mobile phone camera, an
image is captured. This image is loaded into the matlab by using the function imread. This function reads the
image from the specified path. The image is stored in the matrix form of rows and columns. If it is a gray scale
image, then it is stored as an M-by-N array. If the file contains a true colour image or RGB image, then it is
stored as an M-by-N-by-3 array.
3.2 image preprocessing
image captured from digital camera or any mobile phone camera cannot be used directly because it has
lots of noise due to dust and light effect.image processing is done to improve the quality of image.Desired
vegetable image is obtained after filtering and this image can be used for features extraction.
3.3 Boundary extraction
Asit is a coloured input it needs to be converted to grayscale by function rgb2gray (image) and the
syntax is:I=rgb2gray(RGB).which converts the true colour image RGB to the gray scale intensity image I, and
then the image is converted to binary before it is used for further processing in which image consists of only two
colours namely black and white[4].
3.4 Geometric features extraction
This starts with the extracted boundary of the sample. The function used to trace the features is
regionprops. The main features extracted are Area, Major axis and Minor axis [4].
3.5 Colour, shape, and size features extraction
In this red, green and yellow colours are used for classification as there is a difference between the
vegetabless skin based on these colors. Hence these colours are helpful for sorting out the vegetables. The red
and green component is calculated by counting pixel values corresponding to the red and green colours and
yellow component is calculated by first converting the RGB image to CMY by using the function. Shape
features are calculatd by thresholding the image and remove noise from image. After that compute the
boundaries, perimeter and area of image.Separating one kind of sample from another, classification method is
DOI: 10.9790/2834-10516270
www.iosrjournals.org
64 | Page
A1
A2
Input layer
Hidden layer
A3
Output layer
DOI: 10.9790/2834-10516270
www.iosrjournals.org
65 | Page
DOI: 10.9790/2834-10516270
www.iosrjournals.org
66 | Page
.
Fig. 6 Testing section for fruit sample
Table 1: Parameters of ANN Results
S.NO.
1
IV.
ANN BASED
RESULTS
1
FINAL RESULTS
It shows that the
vegetable sample
best quality.
It shows that the
vegetable sample
medium quality.
It shows that the
vegetable sample
poor quality.
given
is of
given
is of
given
is of
This section presents experiments and quality detection of vegetable samples. In this paper two
different types of vegetables are used which are of different colour, shape and size. Results are based on
different dimensions like size, shape, height and width where size A represents height and size B represents
width. The ANN based results for these vegetable samples are shown below in fig. 7, fig. 8, fig. 9 and in fig. 10
respectively.
DOI: 10.9790/2834-10516270
www.iosrjournals.org
67 | Page
DOI: 10.9790/2834-10516270
www.iosrjournals.org
68 | Page
NO.
OF
VEGETABLE
SAMPLES
1
2
3
4
Sample
1
Sample
2
Sample
3
Sample
4
SIZE
A
1
V.
SHAPE
SIZE
B
1
CATEGORY OF
QUALITY OF
VEGETABLE
SAMPLES
Best quality
vegetable
Medium quality
vegetable
Best quality
vegetable
Bad quality
vegetable
The paper proposes a new technique for quality detection of vegetables. The technique is started by
capturing the vegetables image using regular digital camera or any mobile phone camera. The features are
efficiently extracted from the sampled image. The extracted features are based on the parameters like colour,
shape and size. The ANN technique is used for checking the quality. The quality is determined by using the
extracted features of vegetables and the values provided to the neural network during training .The proposed
technique accurately detects the quality of vegetables. The results are good for the four chosen vegetable
samples of different colour, shape and size. This kind of system can be employed in juice plants, fruit and
vegetable farms, packaging etc. In future the quality detection based on ANN should be compared with other
mechanical and automated techniques and a new parameter can also be added.
Acknowledgements
I would like to express my profound sense of gratitude and respect to all those who helped me directly
or indirectly through this phase of my research work. With deep sense of gratitude, I express my sincere thanks
to my esteemed and highly worthy guide Reecha Sharma Assistant Professor, Department of Electronics and
Communication Engineering, Punjabi University, Patiala for their valuable advice, motivation, guidance,
support, encouragement, efforts, timely help and the attitude with which they solved all of my queries regarding
research work. I am deeply indebted to all my friends who helped me with their encouragement, ample morale
support and valuable suggestions. I am also thankful to the authors whose work I have consulted and quoted in
the work.
DOI: 10.9790/2834-10516270
www.iosrjournals.org
69 | Page
[4].
[5].
[6].
[7].
[8].
[9].
[10].
[11].
[12].
[13].
[14].
[15].
[16].
[17].
[18].
[19].
[20].
[21].
Naganur and Sannakki, Fruits Sorting and Grading using Fuzzy Logic, International Journal of Advanced Research in Computer
Engineering & Technology (IJARCET) Volume 1, Issue 6, August 2012.
Sagare, Sunil N. Kore, Fruits Sorting and Grading Based on Color and Size, International Journal of Emerging Technologies i n
Computational and Applied Sciences (IJETCAS), IJETCAS 12-333.
ShahzadPreet Singh, Saurabh Bansal, Mohit Ahuja, Sugam Parnami and Hari Singh, Classification of Apples using Neural
Networks, International Conference on Recent Trends in Engineering Science and Management,Jawaharlal Nehru University New
Delhi ,15 march 2015.
Mandeep Kaur and Reecha Sharma, Quality Detection of Fruits by using ANN Technique, IOSR Journal of Electronics and
Communication Engineering (IOSR-JECE) Volume 10, Issue 4, Ver. II (Jul - Aug .2015).
Bhanu Pratap, Navneet Agarwal, Sunil Joshi and Suriti Gupta, Development of Ann Based Efficient Fruit Recognition Technique,
Global Journal of Computer Science and Technology Volume 14, Issue 5, year 2014.
Effendi,Z ;Ramli,R. ;Ghani,J.A. ;Rahman, Pattern Recognition System of Jatropha curcas Fruits using Back Propagation, signal
and image processing applications (ICSIPA), 2009 IEEE International Conference on Kuala Lumpur.
Zhao, J.T., J. Katupitiya, J., On-tree Fruit Recognition using Texture properties and Color data, International conference on
Intelligent Robots and Systems. 2005, IEEE: Edmonton Canada. p. 263-268.
Lopez Jose J., Cobos Maximo and Aguilera Emanuel, Computer-based Detection and Classification of flaws in Citrus Fruits,
International conference on natural image Processing (ICONIP-2009).
Hetal N. Patel, Dr. R.K.Jain, Dr. M.V.Joshi, Fruit Detection using Improved Multiple Features based Algorithm, International
journal of computer applications Volume 13-No.2, January 2011.
S.Arivazhagan, R.Newlin Shebiah,S.Selva Nidhyanandhan and L.Ganesan , Fruit Recognition using Color and Texture
Features,Journal of Emerging Trends in Computing and Information Sciences Volume 1, No.2,October 2010.
Bindu Tiger and Toran Verma, Identification and Classification of Normal and Infected Apples using Neural Network,
International Journal of Science and Research (IJSR) Volume 2 issue 6, June 2013.
Sandoval,Z ;Prieto,F ;Betancur,J, Digital Image Processing for Classification of Coffee Cherries, Electronics,Robotics and
Automotive Mechanics Conference (CERMA),2010 on Morelos.
John B. Njoroge, Kazunori Ninomiya, Naoshi Kondo and Ilideki Toita, Automated Fruit Grading using Image Processing, The
Society of Instrument and Control Engineers(SICE2002), Osaka, Japan, August 2002, pp 1346-1351.
Usmail Kavdir and Daniel E. Guyer, Apple Grading Using Fuzzy Logic, Turk J Agric (2003), 2003, pp 375-382.
Kondo, N, Quality evaluation of orange fruit using neural networks, In: Food Processing Automation IV Proceedings of the
FPAC Conference. ASAE, 2950 Niles Road, St. Joseph, MI 49085-9659, USA,
K. Nakano, Application of neural networks to the color grading of apples, Computers and Electronics in Agriculture, page 105
116, 1997.
Guo Feng and Cao Qixin, Study on Color Image Processing Based Intelligent Fruit Sorting System, Proceedings of the 5 World
Congress on Intelligent Control and Automation, Hangzhou, P.R. China, June 15-19, 2004, pp 4802-4805.
Ghazanfari A, Kusalik A, and Irudayaraj J, Application of Multi-Structure Neural Network to Sorting Pistachio Nuts,
International Journal of Neural System 8(1):55-61
Mohana S. H. and Prabhakar C.J., A Novel Technique for Grading of Dates using Shape and Texture Features, Machine Learning
and Applications: An International Journal (MLAIJ) Vol.1, No.2, and December 2014.
P.Deepa and Dr.S.N.Geethalakshmi, A Comparative Analysis of Feature Extraction Methods for Fruit Grading Classifications,
International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) 4(2), March-May 2013, pp. 221225.
Alipasandi, Ghaffari and Alibeyglu, Classification of three Varieties of Peach Fruit Using Artificial Neural Network Assisted with
Image Processing Techniques, International Journal of Agronomy and Plant Production. Vol., 4 (9), 2013.
DOI: 10.9790/2834-10516270
www.iosrjournals.org
70 | Page