Grading and quality testing of food grains using neural

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 02 Issue: 11 | Nov-2013, Available @ http://www.ijret.org 545
GRADING AND QUALITY TESTING OF FOOD GRAINS USING NEURAL
NETWORK
Nandini Sidnal1
, Uttam V. Patil2
, Pankaja Patil3
1
Professor, 2
Assistant Professor, Computer Science and Engineering KLE Dr MSSCET, Belgaum, Karnataka, India,
3
Assistant Professor, Dept of Master of Computer Application, GIT Belgaum, Karnataka, India
sidnal.nandini@gmail.com, mail_uttam@yahoo.com, pankaja_kadalagi@yahoo.co.in
Abstract
The quality of food grains is referred to the every aspect of the profit of supply and marketing. The varietals purity is one of the factors
whose inspection is more difficult and more complicated than that of other factors. In the present grain-handling system, grain type
and quality are rapidly assessed by visual inspection. This evaluation process is, however, tedious and time consuming. The decision-
making capabilities of a grain inspector can be seriously affected by his/her physical condition such as fatigue and eyesight, mental
state caused by biases and work pressure, and working conditions such as improper lighting, climate, etc. The farmers are affected by
this manual activity. Hence, these tasks require automation and develop imaging systems that can be helpful to identify quality of
grain images. A model of quality grade testing and identification is built which is based on appearance features such as the
morphological and colour with technology of computer image processing and neural network. The morphological and colour features
are presented to the neural network for training purposes. The trained network is then used to identify the unknown grain types,
impurities and its quality.
Keywords: Grain quality, image processing, neural network
-----------------------------------------------------------------------***-----------------------------------------------------------------------
1. INTRODUCTION
Grains are the prime crop for our country’s peasants to
increase their agricultural income. The automation level of
testing grain quality is low and most work is done by
manpower. The workload is so mass that it will lead to
workers‟ fatigue and need them to have ample testing
experience. And it also makes the testing more costly and long
to be made. With the development of import and export trade
this contradiction is more and more outstanding [1]. During
grain handling operations, information on grain type and grain
quality is required at several stages before the next course of
operation can be determined and performed. The varietals
purity is one of the factors whose inspection is more difficult
and more complicated than that of other factors. In the present
grain handling system, grain type and quality are rapidly
assessed by visual inspection. This evaluation process is,
however, tedious and time consuming. The decision-making
capabilities of a grain inspector can be seriously affected by
his/her physical condition such as fatigue and eyesight, mental
state caused by biases and work pressure, and working
conditions such as improper lighting, climate, etc. [2].The
farmers are affected by this manual activity. Hence, these
tasks require automation and develop systems that can be
helpful to identify grain images, rectify it & then being
analyzed.
1.1 Problem Definition
In the food industry there are various food stuffs in the form of
grains. The food grain types and their quality are rapidly
assessed through visual inspection by human inspectors. The
decision-making capabilities of human-inspectors are
subjected to external influences. A model of quality testing
and identification is built which is based on appearance
features such as the shape and color with technology of image
processing and neural network. The system will take an image
of grain and identifies the grains and impurities if any and
gives the quality of the grain.
2. DESIGN AND IMPLEMENTATION DETAILS
2.1 System Architecture & Design
The work finds application in grain handling operations at
Agriculture Produce Market Committee (APMC), which is an
important organization that deals with farmers, produce,
measures quality and finally values the produce [3]. Different
food grains like wheat, horse gram, corn and rice are
considered in the study. Considering example of rice, rice is
one of the most important cereal grain crops. The quality of
rice has distinct effect on the yield of rice, so the proper
inspection of rice quality is very important. During grain
handling operations, information on grain type and grain
quality is required at several stages before the next course of

__________________________________________________________________________________________
operation can be determined and performed. The varietals
purity is one of the factors whose inspection is more difficult
and more complicated than that of other factors. In the present
grain handling system, grain type and quality are rapidly
assessed by visual inspection. This evaluation process is,
however, tedious and time consuming. The decision-making
capabilities of a grain inspector can be seriously affected by
his/her physical condition such as fatigue and eyesight, mental
state caused by biases and work pressure, and working
conditions such as improper lighting, climate, etc. Hence,
these tasks require automation and develop imaging systems
that can be helpful to identify rice grain images, rectify it &
then being analyzed [5].
Fig -1: Block diagram
2.2 The Work Process of System
• Capturing the image of grains/ images
acquisition
• Image pre-processing
• Image Segmentation and Identifying region of
Interest
• Feature Extraction
• Training and Testing
2.2.1 Image Acquisition
Food grain images acquisition is considered as the most
critical step of the grain recognition system, as it determines
the final grain image quality, which has drastic effects on
overall system performance. A total of around 400 food grain
images are acquired. The images are acquired with a color
Digital Camera. Sony cyber-shot DSCS750 digital camera is
used to capture images. The acquired image is of 2592x1944
resolution. 2.5inch by 3.5 inch area is marked on background
of maroon color. Food grains are positioned beneath the focus
of a camera on that marked area of the background and the
image is captured.
2.2.2 Image Pre-processing
The images acquired with a color Digital Camera is resized to
a resolution of 640 by 480 saved as JPEG image. Further to
remove noise the image is converted to black and white image
and patches in an image are considered using matlab function
bwlabel, Label connected components in binary image. The
patches with size less than 70 pixels are assumed to be noise
and are ignored. The other patches are the region of interest.
2.2.3 Image Segmentation and Identifying Region of
Interest
Connected components labeling scans an image and groups its
pixels into components based on pixel connectivity, i.e. all
pixels in a connected component share similar pixel intensity
values and are in some way connected with each other. Once
all groups have been determined, each pixel is labeled with a
gray level or a color (color labeling) according to the
component it was assigned to. Connected component labeling
works by scanning an image, pixel-by-pixel (from top to
bottom and left to right) in order to identify connected pixel
regions, i.e. regions of adjacent pixels which share the same
set of intensity values V. (For a binary image Original image
is converted to black and white image and patches in an image
are considered using matlab function bwlabel, Label
connected components in binary image. The patches with size
more than 70 pixels are assumed to be the region of interest.
Once the patches (grains and impurities) are identified the top,
bottom, left and right is calculated to crop the segment from
original image.
2.2.4 Feature Extraction
The color features such as mean value of R, the mean value of
G, the mean value of B are calculated and four Morphological
feature namely Area, Major axis length, Minor axis length and
Aspect ratio are extracted and stored in knowledgebase, so a
total of 7 features as listed in table 1 are stored suitably for
later usage in training PNN.
Table -1: Morphological and color features used for grade and
grain classification

__________________________________________________________________________________________
2.2.5 Training and Testing Phase
Seven Morphological and color features are extracted from the
images and are stored in the knowledgebase. A probabilistic
neural network (PNN) is predominantly a classifier that maps
any input pattern to a number of classifications. The
probabilistic neural network is trained with eight different
types of food grains namely Sonamasuri rice, Basmati rice,
Horsegram brown, Horsegram white,yellow corn,Gujrat
Wheat, OrangeCorn and Khapli Wheat, with 200 samples.
Testing is the final step of the grain recognition system. In
case of testing phase an image from the testing set (different
from the training set) is selected and its features are extracted
as training images. Then those features are matched with the
feature from the database created for training image. It will
compute the shortest distance between the trained images
stored in the database and the image that was chosen for
testing. Then the image with the shorted distance is chosen to
be recognized. This process is repeated for all the testing
images in order to know the accuracy of the system.
3. THE ALGORITHM: RECOGNITION AND
CLASSIFICATION OF FOOD GRAIN
Input: Original Color Image
Output: Classified food grains along with Quality
Step1: Acquire the food grain images.
Step2: Enhance image to remove noise and resizing.
Step3: Identify Patches and Do the image segmentation.
Step4: Extract Color and morphological features.
Step5: Use these features to recognize and classify the food
grain image samples using Neural network
4. RESULTS AND ANALYSIS
4.1 Training Phase
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Features Extraction (TRAINING PHASE)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Enter Grain Type
1. Rice 2. Wheat 3.Corn 4.Horse Gram 5. Impurity 1
Enter Grain Grade
1. Basamati 2.Sona Masuri 1
Select the Grain image for training phase
Fig.2 Snapshot to select the Grain image
Fig.3: Original Resized Image of basmati rice
Snapshot of the selected basmati rice image after Background
subtraction
Fig.4: After Background subtraction

__________________________________________________________________________________________
Fig.5: Black and white Image to find ROI
Snapshot of the selected basmati rice image after segmentation
Fig.6: After Segmentation
The four Morphological feature namely and color features are
extracted and stored in knowledgebase, so a total of 7 features
for each segment of fig 6
4.2. Testing Phase
Select the Grain image for testing phase
Fig.7 Snapshot to select the Grain image during testing phase
The selected image for rice with impurities for testing phase
Fig.9: Original Resized Image of basmati rice
Output: The grain and the impurity is identified and its
quality is
Fig 10: The result after testing
5. RESULTS
More than 160 images were used to test the system and it was
found that accuracy of identifying grain and its grade is
100%.Where as the accuracy of identifying quality of grains
was 80-90% accurate for each grain type. The testing results
are as shown below in table 2. The result analysis is as shown
in figure 11

__________________________________________________________________________________________
Table 2: Consolidated results of all grains with accuracy of
testing
It is very clear from the graph below that accuracy of
identifying grain and its grade is 100%.Where as the accuracy
of identifying quality of grains is in the range of 80-90% for
each grain type.
Fig 11: Results of all grains with accuracy of testing
CONCLUSIONS
The system identifies the quality of grains which is based on
appearance features such as the shape and color, with
technology of image processing and neural network. More
than 160 Images were used to test the system and it was found
that accuracy of identifying grain and its grade is
100%.Quality testing of grains was 80-90% accurate for each
grain type. A very simple method is proposed for
classification of food grains which require limited features and
thus overcoming the disadvantages like tediousness and time
consumption. The quality of food grain includes not only
exterior quality but also the interior one such as the texture,
the nutrient component, the protein, moisture of the grain and
producing area. The latter is not researched in this work
because of the limitation of machine vision. It is difficult to
test the features by visual image.
REFERENCES
[1]. HAN Zhong-zhi; LI Yan-zhao; LIU Jing; ZHAO You-
gang, Quality Grade Testing of Peanut Based on Image
Processing, 2010,IEEE
[2]. Sanjivani Shantaiya, Mrs.Uzma Ansari, Identification Of
Food Grains And Its Quality Using Pattern lassification,
Special Issue of IJCCT Vol. 2 Issue 2, 3, 4; 2010 for
International Conference [ICCT-2010], 3rd December 2010
[3]. B. S. ANAMI, D. G. SAVAKAR,Improved Method for
Identification and Classification of Foreign Bodies Mixed
Food Grains Image Samples, ICGSTAIML Journal, ISSN:
1687-4846, Volume 9, Issue I, February 2009
[4]. Basavaraj .S. Anami and Vishwanath.C.Burkpalli, Texture
based Identification and Classification of Bulk Sugary Food
Objects, ICGST-GVIP Journal, ISSN: 1687-398X, Volume 9,
Issue 4, August 2009
[5]. Basavaraj S. Anami, Dayanand G. Savakar, Effect of
Foreign Bodies on Recognition and Classification of Bulk
Food Grain Image Samples, Journal of Applied Computer
Science, no. 6 (3) /2009, Suceava.
[6]. D. M. Hobson, R. M. Carter, Y. Yan, Characterisation and
Identification of Rice Grains through Digital Image
Analysis,2007,IMTC
[7]. ABDELLAOUI Mehrez , DOUIK Ali, Hybrid method for
cereal grain identification using morphological and color
features, 1-4244-0395-2/06/$20.00 ©2006 IEEE.
[8]. LIU Zhao-yan ,CHENG Fang , YING Yi-bin , RAO Xiu-
qin ,Identification of rice seed varieties using neural network,
Journal of Zhejiang University SCIENCE

Grading and quality testing of food grains using neural

More Related Content

What's hot (18)

Viewers also liked (20)

Similar to Grading and quality testing of food grains using neural (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Grading and quality testing of food grains using neural