Abstract

In this research, a new vision system to characterize the recognition of vegetables in images has
been developed. It is always related to image processing, which can control the classification,
qualification and segmentation of images. It is a recognition system for super market and
grocery stores. From the captured images multiple recognition clues such as colour, shape, size,
texture and weight are extracted and analysed to classify and recognize the vegetables. The
results show that it has good robustness and a very high success. An important characteristic of
the proposed algorithm is that it is able to work with several elements inside the camera field
of view. This adds flexibility to the application in order to work in the country or in a
greenhouse, where the elements are very close to each other and the location of all them must
be obtained in real time. This approach is less complex and relatively very faster than other
approach. If the system can identify uniquely, the checkout process at superstores will be fast
and efficient.
Image segmentation is to partition an image into meaningful regions with respect to a
particular application. Object recognition is the task of finding a given object in an image or
video sequence.

Page | 1
 CONTENTS

List of figures 3

1)Topic 1: Introduction 4

1.1:Definations 4

2)Topic 2: Problem Statement 5

2.1:Motivation 6

3)Topic 3:Working 6

4)Topic 4:Applications 8

5)Topic 5:Methodology 9

5.1:Flow Chart 9

5.2:Matlab Coding 9

6)Topic 6:Conclusion 12

7)Topic 7:References 13

Page | 2
 List of figures

Figure no. Name Page no.

3.1 Original Image 6

3.2 Reference Image 7

3.3 Extracted Image 8

3.4 Flow Chart 9

Page | 3
1.Introduction
We present automatic recognition system (vegetable vision) to facilitate the checkout process
of a supermarket or grocery stores. Vegetable vision is done with the image processing and
image analyzing. Image processing is done by MATLAB. This system consists of an integrated
measurement and imaging technology with a user friendly interface. When we bring a
vegetable at checkout point, an image is taken, a variety of features such as color, shape, size,
density, texture etc are then extracted. These features are compared to stored data. Depending
on the certainty of the classification and recognition, the final decision is made by the system.
Vegetable quality is frequently referred to size, shape, mass, firmness, color and bruises from
which fruits can be classified and sorted. The classification technique is used to recognize the
vegetables shape, size, color and texture at a unique glance. Digital image classification uses
the spectral information represented by the digital numbers in one or more spectral bands, and
attempts to classify each individual pixel based on this spectral information.
Image segmentation is the foundation of object recognition and computer vision. And there is
some specifically given work (such as region extraction and image marking) to do after the main
operation of image segmentation for the sake of getting better visual effect. Two major
computer vision problems, image segmentation and object recognition, have been traditionally
dealt with using a strict, bottom-up ordering. Implementation between Segmentation and
Object recognition using the said techniques is studied.

1.1 Definitions
The following are the definitions relevant to this project:

1.1.1 Computer Vision

Computer vision is a field in Computer Science that includes methods for acquiring, processing,
analyzing, and understanding images from the real world in order to produce numerical or
symbolic information. A theme in the development of this field has been to duplicate the
abilities of human vision by electronically perceiving and understanding an image. Computer
vision has application like Object Detection, Object Recognition and Object Classification.

Page | 4
1.1.2 Detection:
Many points in an image have a nonzero value for the gradient, and not all of these points are
edges for a particular application. Thresholding is used for the detection of edge points.

1.1.3 Feature Extraction

When the input data to an algorithm is too large to be processed and it is suspected to be
notoriously redundant then the input data is transformed into a reduced representation set of
features, also named features vector. Transforming the input data into the set of features is
called feature extraction. The features used in this project are aspect ratio, height, width,
elongation, skewness, compactness, orientation, area, perimeter and extent.
1.1.4 Object recognition:
It is the task of finding a given object in an image or video sequence. For any object in an image,
there are many 'features' which are interesting points on the object that can be extracted to
provide a "feature" description of the object.

1.1.5 Image segmentation:

It is the process of partitioning/subdividing a digital image into multiple meaningful regions or
sets of pixels regions with respect to a particular application. The segmentation is based on
measurements taken from the image and might be grey level, colour, texture, depth or motion.
The result of image segmentation is a set of segments that collectively cover the entire image.

2.Problem Statement
The task here is to automatically detect and classify the vegetables image acquired from
Matlab. Assuming that the different images are present and some are overlapped on one
another.The task is to separate the images overlapped and recognize them .In this, we have use
vegetables image present in Matlab.

Page | 5
2.1 Motivation
The classification technique is used to recognize the vegetables shape, size, color and texture
at a unique glance. Digital image classification uses the spectral information represented by the
digital numbers in one or more spectral bands, and attempts to classify each individual pixel
based on this spectral information. Two major computer vision problems, image segmentation
and object recognition, have been traditionally dealt with using a strict, bottom-up ordering.

3.Working :
The steps of the process are illustrated here with some discussions of each step.
The speed of the object can be determined by using Matlab function. There are various steps
involve in this process which are as given :

a)Read the image

After reading the image in the Matlab workspace, the image must be read using (imread)
function which read the image from Matlab workspace.

Fig 3.1 Original image

b)Taking Reference image :
We must have one reference image of which we have compare with original image for feature
extraction .

Page | 6
 Fig3.2: Reference image
c)Convert the image to Black and white
The image is converted to black and white by using the function im2bw, which converts an
image to binary image. Each pixel takes a value that is a normalized intensity value that lies in
the range [0, 1].
d)Morphological Operation:
Strel( ) function creates a flat, disk-shaped structuring element, where R specifies the radius. R
must be a nonnegative integer. N must be 0, 4, 6 or 8. When N is greater than 0, the disk-
shaped structuring element is approximated by a sequence of N periodic-line structuring
elements. When N equals 0, no approximation is used, and the structuring element members
consist of all pixels whose centers are no greater than R away from the origin. If N is not
specified, the default value is4.

e)Bounding Box Extraction

We do this by constructing a bounding box. This bounding box is used to isolate the area of
interest in the image and is similar to window (or key region) processing. This can be easily
made by calling the (Bounding Box) property of the (regionprops) function.
Feature extraction is used to extract the feature and locate the position of the object in the
image , which will be estimated using function like regionprops.

Page | 7
f)Calculating pixel value and circularity:
After constructing the bounding box and as per the value of centroid, we have calculated the
pixel value by using impixel inbuilt command and with the help of circularity we have
distinguish different images by using imcrop to crop the desired image.

Fig3.3 : Extracted frame

4.Applications :
This technique of using matlab code can be use in various purpose which are given below .
Locate tumors and other pathologies
Measure tissue volumes
Surgery planning
Face recognition
Fingerprint recognition
Iris recognition and many more.

Page | 8
5. METHODOLOGY:
This section presents the technology inputs and processing steps needed in the vegetables
identification and recognization .The key stages in the vehicle identification and tracking
process. Each of these steps is described in detail below.

5.1FLOW CHART:

Fig5.1:Flow Chart

5.2 Matlab CODING :

clear all
close all
clc
a=imread('C:\Users\Desktop\Whiteimage.jpg');
b=imread('C:\Users\Desktop\vegetableimage.jpg');
c=rgb2gray(b);
d=im2bw(c);
out=im2double(a)-im2double(d);
se=strel('disk',6);
frameout=imdilate(out,se);
[L,n]=bwlabel(frameout,8);

Page | 9
s=regionprops(L,'Area','Centroid','BoundingBox','Perimeter');
imshow(b)
for i=1:n
C=s(i,1).Centroid;
BB=s(i,1).BoundingBox;
P=s(i,1).Perimeter;
A=s(i,1).Area;
xcorner=BB(1);
ycorner=BB(2);
xwidth=BB(3);
ywidth=BB(4);
Centroidx=C(1);
Centroidy=C(2);
rect=[xcorner ycorner xwidth ywidth];
rectangle('Position',[xcorner ycorner xwidth
ywidth],'Edgecolor','r');
hold on
plot(Centroidx,Centroidy,'b*');
hold off
Pi=impixel(b,Centroidx,Centroidy);
Cir=P^2/(4*pi*A);

if((235<Pi(1))&&(Pi(1)<239)&&(25<Pi(2))&&(Pi(2)<29)&&(34<Pi(3))&&(Pi(3
)<38)&&(1.0<Cir)&&(Cir<1.2))
p=imcrop(b,rect);
figure,imshow(p)
title('TOMATO');

elseif((188<Pi(1))&&(Pi(1)<191)&&(36<Pi(2))&&(Pi(2)<40)&&(44<Pi(3))&&(
Pi(3)<46)&&(2.6<Cir)&&(Cir<2.8))
p=imcrop(b,rect);
figure,imshow(p)
title('CHILLY');

elseif((119<Pi(1))&&(Pi(1)<123)&&(34<Pi(2))&&(Pi(2)<38)&&(74<Pi(3))&&(
Pi(3)<76)&&(1.3<Cir)&&(Cir<1.4))
p=imcrop(b,rect);
figure,imshow(p)
title('BRINJAL');

elseif((249<Pi(1))&&(Pi(1)<253)&&(180<Pi(2))&&(Pi(2)<184)&&(111<Pi(3))
&&(Pi(3)<115)&&(1.6<Cir)&&(Cir<1.8))
p=imcrop(b,rect);
figure,imshow(p)
title('POTATO');

elseif((140<Pi(1))&&(Pi(1)<144)&&(180<Pi(2))&&(Pi(2)<184)&&(98<Pi(3))&
&(Pi(3)<102)&&(1.96<Cir)&&(Cir<2.0))
p=imcrop(b,rect);
figure,imshow(p)
title('PEAS');

Page | 10
elseif((0<Pi(1))&&(Pi(1)<2)&&(95<Pi(2))&&(Pi(2)<99)&&(49<Pi(3))&&(Pi(3
)<52)&&(1.4<Cir)&&(Cir<1.5))
p=imcrop(b,rect);
figure,imshow(p)
title('BITTERGUARD');

elseif((215<Pi(1))&&(Pi(1)<219)&&(84<Pi(2))&&(Pi(2)<88)&&(40<Pi(3))&&(
Pi(3)<44)&&(1.2<Cir)&&(Cir<1.3))
p=imcrop(b,rect);
figure,imshow(p)
title('PUMPKIN');

elseif((241<Pi(1))&&(Pi(1)<245)&&(111<Pi(2))&&(Pi(2)<115)&&(36<Pi(3))&
&(Pi(3)<40)&&(2.0<Cir)&&(Cir<2.1))
p=imcrop(b,rect);
figure,imshow(p)
title('CARROT');

elseif((62<Pi(1))&&(Pi(1)<66)&&(98<Pi(2))&&(Pi(2)<102)&&(34<Pi(3))&&(P
i(3)<38)&&(1.4<Cir)&&(Cir<1.5))
p=imcrop(b,rect);
figure,imshow(p)
title('BRONCHI');

elseif((65<Pi(1))&&(Pi(1)<69)&&(46<Pi(2))&&(Pi(2)<50)&&(Pi(3)<2)&&(1.8
<Cir)&&(Cir<1.9))
p=imcrop(b,rect);
figure,imshow(p)
title('MAIZE');

elseif((235<Pi(1))&&(Pi(1)<239)&&(25<Pi(2))&&(Pi(2)<29)&&(34<Pi(3))&&(
Pi(3)<38)&&(1.2<Cir)&&(Cir<1.3))
p=imcrop(b,rect);
figure,imshow(p)
title('CAPSICUM');

elseif((251<Pi(1))&&(Pi(1)<255)&&(184<Pi(2))&&(Pi(2)<188)&&(71<Pi(3))&
&(Pi(3)<75)&&(1.4<Cir)&&(Cir<1.5))
p=imcrop(b,rect);
figure,imshow(p)
title('ONION');
end
end

Page | 11
 Conclusion

It is testified that Vegetable vision is an alternative to unreliable manual sorting of Vegetables.

The system can be used for vegetables grading by the external qualities of size, shape, color and
surface. The Vegetable vision system can be developed to quantify quality attributes of various
vegetables such as mangoes, cucumbers, tomatoes, potatoes, peaches and mushrooms
In this research, segmentation method and classification based on area thresholding method
are developed. Centroid and pixel value algorithms are combined to obtain the exactly
classified images. The system shows an effective and reliable classification of images captured
by a camera. The image segmentation algorithm is very useful method in the image processing
and it is very helpful for the subsequent processing.

Page | 12
 References

Heidemann, G., (2005), Unsupervised image categorization, Image and Vision

computing 23, (10), pp. 861876

W. Seng, S. Mirisaee, (2009), "A New Method for Fruits Recognition System", Electrical
Engineering and informatics, 2009. ICEEI '09. International Conference on., pp. 130-134.

Rocha, A., Hauagge, D., Wainer, J., Goldenstein, S. ,(2010), Automatic fruit and
vegetable classification from images, Elsevier, pp. 1-11.

Aya Muhtaseb, Sanaa Sarahneh, Hashem Tamimi, Fruit/Vegetable Recognition,

http://sic.ppu.edu/abstracts/090172_681636469_Fruit/vegetableRecognition
(paper).pdf

Luis Gracia, Carlos Perez-Vidal, Carlos Gracia, Computer Vision Applied to Flower, Fruit
and Vegetable processing, http://www.waset.org/journals/waset/v54/v54-86.pdf

R. Fisher, S. Perkins, A. Walker and E. Wolfart, "Edge Detectors", 2003.

G. Wenshuo, Y. Lei, Z. Xiaoguangand L. Huizhong, "An Improved Sobel Edge Detection",

IEEE, 2010.

M. A. Fari, "Study Of Image Segmentation By Using Edge Detection Techniques",

International Journal

Page | 13
Page | 14