Scribd
Upload
7 views2 pages

Upload 02

The study evaluates the quality assessment of rice using various convolutional neural networks (CNN) and machine learning techniques, highlighting the need for automated grain identification due to the limitations of manual classification. A dataset of 500 rice images was analyzed, revealing that GoogleNet achieved the highest CNN accuracy of 64.4%, while texture feature-based methods yielded even better results with accuracies of 99.2% and 93.4%. The research emphasizes the importance of digital image processing in improving the efficiency and accuracy of rice quality grading.

Uploaded by

Robert
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views2 pages

Upload 02

The study evaluates the quality assessment of rice using various convolutional neural networks (CNN) and machine learning techniques, highlighting the need for automated grain identification due to the limitations of manual classification. A dataset of 500 rice images was analyzed, revealing that GoogleNet achieved the highest CNN accuracy of 64.4%, while texture feature-based methods yielded even better results with accuracies of 99.2% and 93.4%. The research emphasizes the importance of digital image processing in improving the efficiency and accuracy of rice quality grading.

Uploaded by

Robert
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

ISSN: 2347-4688, Vol. 12, No.(2) 2024, pg.

762-772

Current Agriculture Research Journal


www.agriculturejournal.org

Quality Assessment of Rice using Convolution Neural


Networks and other Machine Learning Techniques-
A comparative Study
NILUTPAL PHUKON1, KSH ROBERT SINGH1*, TAKHELLAMBAM SONAMANI SINGH2,
SUBIR DATTA1, SUBHASISH DEB1, USHAM ROBINCHANDRA SINGH3
and GHANESHWORI THINGBAIJAM4

1
Department of Electrical Engineering, Mizoram University, Mizoram, India.
Department of Computer Science and Engineering, MIT, Manipur University, Manipur, India.
2

3
Department of Computer Science, South East Manipur College, Chandel, Manipur, India.
4
Department of Electronics and communication Engineering, MIT, Manipur University, Manipur, India.

Abstract
Rice holds its position as the most widely cultivated crop worldwide,
its demand steadily rising alongside population growth. The precise
identification of grains, especially wheat and rice, holds significant Article History
importance as their ultimate utilization depends on the quality of grains Received: 20 June 2024
prior to processing. Traditionally, grain identification tasks have been Accepted: 03 August 2024
predominantly manual, relying on experienced grain inspectors and
consuming considerable time. However, this manual classification Keywords
Artificial Intelligence;
process is susceptible to variations influenced by individual perception, Computer Vision;
given the subjective nature of human image interpretation. Consequently, Image Classification;
there is an urgent need for an automated recognition system capable Image Recognition;
Machine Learning;
of accurately identifying grains under diverse environmental conditions, Neural Networks;
necessitating the application of digital image processing techniques. Object Detection.
In this study, we focus on grading five distinct varieties of rice based on
their quality, employing a range of convolution neural networks (CNN)
namely, Efficientnetb0, Googlenet, MobileNetV2, Resnet50, Resnet101,
and ShuffleNet. The performance of CNN towards identification and
grading of rice grain is also compared to that of other parametric and
Non-parametric classifiers namely, Linear Discriminant Analysis (LDA),
K-Nearest Neighbor (K-NN), Naive Bayes (NB), and Back Propagation
Neural Network (BPNN) using Gray Level Co-occurrence Matrix (GLCM)
and Gray Level Run Length Matrix (GLRLM) based texture features.

CONTACT Ksh Robert Singh [email protected] Department of Electrical Engineering, Mizoram University, Mizoram, India.

© 2024 The Author(s). Published by Enviro Research Publishers.


This is an Open Access article licensed under a Creative Commons license: Attribution 4.0 International (CC-BY).
Doi: http://dx.doi.org/10.12944/CARJ.12.2.21
PHUKON et al., Curr. Agri. Res., Vol. 12(2) 762-772 (2024) 763

The image dataset comprises five grades of rice, each containing 100
images, resulting in a comprehensive collection of 500 samples for
analysis. It is observed that, Convolution neural networks can grade
five different qualities of rice with highest accuracy of 64.4% in case
of GoogleNet. Results show that, rice grading using texture features
performs better with highest accuracy of 99.2% (using GLCM) and 93.4%
(using GLRLM).

Introduction However, these works also highlights the tedious


Rice is important because it gives us energy and processes of arranging the grain kernels in a manner
nutrients. Moreover, it is a big part of our culture not to overlapped or touch each other. Classification
and celebrations. It helps farmers make a living of grain based on bulk grain image using color and
and supports local economies. Since it is cheap texture features is also discussed.17 It is learnt that
and can be used in lots of different ways, many some of the tedious tasks for arranging the grain
people rely on it for food. Rice shows where we kernels in a non-touching pattern is not required
come from and how we are connected to nature. in this work as the images are taken on bulk grain
Therefore, understanding the significance of rice instead of single grain kernel. This work is able to
quality helps us to recognize its role in our daily achieve a classification accuracy of 90% using Back
lives and in society. Propagation Neural Network (BPNN). Classification
of different varieties of rice grain and plant seeds
Assessing rice quality is crucial for several reasons. kernels using image processing and Convolution
It ensures the rice we eat is safe and nutritious, Neural Networks (CNN) is also presented.18-24 It is
checking for essential nutrients and absence of found that only few of the works have compares
harmful substances. Secondly, it guarantees the the performance to that of pre-train CNN models.
rice meets our taste preferences in appearance, Identification and classification of rice grain (on
texture, and cooking properties. Additionally, single kernel image) using different deep learning
it influences market prices, supports farmers, and methods are also presented.25-26 The accuracy
aids industries in producing consistent products. achieved in these work ranges from 84% to 99%.
Proper assessment also helps in maintaining It is also observed that some of the classification
storage conditions and complying with food safety tasks carried out in the literature were based on
regulations, contributing to a reliable rice industry, few numbers of class labels and therefore the task
and meeting consumer demands. involve are not that challenging.

Rice grains and corn seeds classification using It is also observed from the literature that, manual
different image pre-processing techniques and identification and classification of grains is quite
machine learning algorithms such as K-Nearest tedious, time consuming and the result is subjective
Neighbor (K-NN), Artificial Neural Network in nature. Therefore, a robust classification process
(ANN), Support Vector Machine (SVM), etc., are based on digital image processing and computer
discussed.1-14 These works also address different vision is becoming so important to overcome the
aspects for assessing rice grain such as detecting above problems that are encountered in manual
defects, varieties classification, quality grading etc. inspection process. It is learnt that classification
The accuracy achieved in the above literature’s using single grain kernel are not that easy as
ranges from 39% to 100%. It is also observed that compared to classification on bulk grain images.
information related to dataset (training/testing) Because, the former involves arranging the grain
is not properly highlighted in most of the work. kernels in such a way that, they do not touch each or
Classification of food grains using image processing overlap. It is also learnt that most of the classification
techniques and Probabilistic Neural Network (PNN) task carried out in the last few years were based
is conducted with accuracies 96% and 100%.15-16 on extracting certain attributes or features from the

You might also like