Volume 7, Issue 6, June – 2022 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Grading The Quality of Apple Fruits Using

Machine Learning Technique
[1]Rachana M Shet, [2]Priyadarshini V, [3]HOD harshavardhan Tiwari, [4]Dr. Harshavardhan Tiwari, B.E Student, Department
of Information Science and Engineering, Jyothy Institute of Technology,Bangalore [1][2],Professor, Department of Information
Science and Engineering, Banglore[5]

Abstract:- This project introduces fruit quality II. OBJECTIVES

identification system. The system design considers certain
features including the color and shape of the fruit, which  To detect and classify apple fruits disease.
increases the accuracy for pixel detection of the fruit,  To extract the images of apple fruits using image
using a color classification support vector machine processing based on colour, Shape and texture.
(SVM). Image processing provides solutions for  To analyse the performance of CNN, ANN,SVM
automated fruit size grading to provide accurate, reliable, algorithms using different Data sets.
consistent and quantitative information in addition to  To recognize the major diseases in different categories of
handling large quantities. apple fruits and Identify quality graded apple fruits.

Computer vision is a consistent and advanced III. METHODOLOGY

technology for image processing with favorable results
and immense potential. A computer vision has been Detection and classification of apple fruit disease.
strongly adopted in the heterogeneous sector including Disease infection in agricultural products such as apple fruit
agriculture. Fruit grading is a very important function as resulting in deterioration in the quality and quantity of
it gives higher value to the grower and improves agricultural products. It directly affects the financial
packaging, management and overdevelopment in the resources of the farmer and human health. Detection of
marketing system. diseases in apple fruit crop leads to reduction in yield and
quality at an early stage of development.
Keywords:- GPU P100, GEFORCE10, Tensor flow,
MATPLOT LIB, keras To extract images of apple fruits using image processing
based on color, size and texture. Automated apple fruit
I. INTRODUCTION identification or recognition using image processing is a key
element in preposition agriculture for object detection in large
India is a developing country. It is the sixth largest crop plots. Image analysis demonstrated difficulty in
economy in the world. The US ranks first in the world with a processing an image captured in the natural environment and
GDP of 19.39 trillion, followed by China with a GDP of 12.24 with a textured object of non-ideal geometric shape. To
trillion, Germany, Japan, the UK with a GDP of 12.24 trillion, analyze the performance of CNN, ANN, SVM algorithms
India ranks sixth worldwide. Agriculture is the backbone of using different data sets.
the Indian economy. Apple is emerging as a super crop all
over the world. The development of agriculture is essential to ANN (Artificial Neural Network): ANN consists of
meet the demand. India is a major country which exports three layers, an input layer, hidden layer and output layer. The
many agricultural products. Agricultural produce is important bias function and the load are all parameters associated with
for the quality of agricultural products until it reaches the neurons. .ANN is used which has high adaptability and allows
customer. Computer vision is the process of applying a to work with very Large set of input data. The neural network
variety of technologies to provide imaging based automated was described in a MATLAB simulation environment using
inspection, process control and robotic guidance in industrial the program return.
applications.
CNN (Convolution Neural Network): The training
process of the algorithm in the architecture of the
convolutional neural network consists of several stages. CNN
is a power full algorithm for image processing. These images
contain RGB (red, green and blue) data, MATLAB can be
used to import an image from a file into memory. Color
images are stored in a 3D array.

IJISRT22JUN715 www.ijisrt.com 947

 Volume 7, Issue 6, June – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
SVM (Support Vector Machine): The working of REFERENCES
Support Vector Machine is based on supervised learning
algorithm. SVM is a linear model for classification and [1]. A role of computer vision in fruits and vegetables
regression problems. The algorithm creates a line or among various horticulture products of agriculture
hyperplan that separates the data into classes. fields: A survey Mukesh Kumar Tripathi *, Dr.
Dhananjay D. Maktedar
Identification of major diseases in different categories [2]. Fruit Classification System Using Computer Vision: A
of apple fruits and identification of quality apple fruits. Review 1 Raja Sekar L, 2Ambika N, 3Divya V and
Conventional approaches to disease detection require 4Kowsalya T, 1Assistant Professor, 2,3,4UG Scholar,
observation of the form for continuous monitoring, either by 1,2,3,4Department of EIE, Bannari Amman Institute of
formers or by specialists. But it is very expensive and time Technology, Sathyamangalam, Erode, India
consuming. Existing techniques for disease detection have [3]. Identification, classification & grading of fruits using
used machine learning, image processing and classification machine learning & Computer intelligence: a review
based approaches to identify and detect diseases in March 2020
agricultural products. [4]. Grading and sorting technique of dragon fruits using
machine learning algorithms. Pallavi U. Patil, Sudhir B,
 Work Flow: Vinay J, Nagalkar, Sona B. Nikam, G.C.Wakchaure

Fig 1

IV. EXPECTED OUTCOME OF THE PROJECT

 Detection and classification of apple fruits in terms of

quality.
 Extraction of apple fruit images and image processing for
quality gradation.
 Performance analysis of different machine learning
algorithms.
 Recognition of different diseases of apple fruits and
identifying graded of apple fruits.

V. CONCLUSION

Machine learning algorithms such as artificial neural

networks (ANNs), support vector machines (SVMs), and
convolutional neural networks (CNNs) will use the
characteristics of the apple fruit to detect and sort that apple
fruit. The accuracy of the features is based on the image
processing operation that enhances the quality of the data set
images and the images captured by the camera.

IJISRT22JUN715 www.ijisrt.com 948