Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Transformer based fruits disease classification

Transformer based fruits disease classification As technology progresses, automation is increasingly gaining importance, and one area where it holds significant potential is fruit classification. This research endeavors to develop a fruit classification system utilizing computer vision techniques. To this end, this work modified the adoption of deep learning-based approaches, specifically a customized Convolutional Neural Network (CNN) and a Vision Transformer. The dataset employed in this study consists of 12 distinct fruit classes, encompassing six categories of healthy fruits and six categories of unhealthy fruits. A total of 12000 data from 12 different classes have been used for training and testing. Extensive preprocessing techniques were applied to enhance the performance of the models, accompanied by data augmentation methods. These steps aimed to ensure the models’ ability to effectively capture and differentiate the unique characteristics and features of various fruits. To optimize the models’ performance, a series of systematic research trials were conducted, involving the adjustment of various parameters. Through this iterative process, the models were fine-tuned to achieve the highest average classification accuracy. Notably, the vision transformer approach emerged as the most successful, attaining an outstanding average accuracy of 98.05%. The findings of this research demonstrate the efficacy of deep learning methodologies in fruit classification tasks, particularly the Vision Transformer architecture. The modified system exhibits remarkable accuracy, highlighting its potential for accurate and reliable fruit classification. By automating the fruit classification process through image classification techniques, this research contributes to the broader goal of streamlining and optimizing fruit quality assessment. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Multimedia Tools and Applications Springer Journals

Transformer based fruits disease classification

Download PDF
21 pages

Loading next page...
 
/lp/springer-journals/transformer-based-fruits-disease-classification-nMjsIJU0h4

References (31)

Publisher
Springer Journals
Copyright
Copyright © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024
ISSN
1380-7501
eISSN
1573-7721
DOI
10.1007/s11042-024-19172-1
Publisher site
See Article on Publisher Site

Abstract

As technology progresses, automation is increasingly gaining importance, and one area where it holds significant potential is fruit classification. This research endeavors to develop a fruit classification system utilizing computer vision techniques. To this end, this work modified the adoption of deep learning-based approaches, specifically a customized Convolutional Neural Network (CNN) and a Vision Transformer. The dataset employed in this study consists of 12 distinct fruit classes, encompassing six categories of healthy fruits and six categories of unhealthy fruits. A total of 12000 data from 12 different classes have been used for training and testing. Extensive preprocessing techniques were applied to enhance the performance of the models, accompanied by data augmentation methods. These steps aimed to ensure the models’ ability to effectively capture and differentiate the unique characteristics and features of various fruits. To optimize the models’ performance, a series of systematic research trials were conducted, involving the adjustment of various parameters. Through this iterative process, the models were fine-tuned to achieve the highest average classification accuracy. Notably, the vision transformer approach emerged as the most successful, attaining an outstanding average accuracy of 98.05%. The findings of this research demonstrate the efficacy of deep learning methodologies in fruit classification tasks, particularly the Vision Transformer architecture. The modified system exhibits remarkable accuracy, highlighting its potential for accurate and reliable fruit classification. By automating the fruit classification process through image classification techniques, this research contributes to the broader goal of streamlining and optimizing fruit quality assessment.

Journal

Multimedia Tools and Applications – Springer Journals

Published: Jan 1, 2025

Keywords: Fruits classification; Computer vision; Vision transformer; Deep learning

There are no references for this article.