Improving Sugarcane Disease Identification with L1-Regularized Transfer Learning Approach

DOI

10.2991/978-94-6463-544-7_11How to use a DOI?

Keywords

sugarcane; DenseNet121; InceptionNetv3; Deep Learning

Abstract

Sugarcane is a crucial cash crop in India, significantly impacting the country’s economy through its extensive use in the sugar industry. However, various diseases are hindering sugarcane production. Effective identification and management of these diseases can substantially improve yields. In this study, we developed a modified and regularized transfer learning model to diagnose sugarcane diseases early and accurately by processing leaf images. We enhanced two well-known pre-trained models, DenseNet121 and InceptionV3, by adding nine extra layers at the end. These include three layers of batch normalization, three dropout layers, and three dense layers using LASSO regularization. The models were trained using an appropriate optimizer and early-stop regularization. The customized models effectively identified sugarcane diseases. Our evaluation metrics included recall, precision, f1-score, roc curves, and confusion matrices. The results indicate that the modified and regularized InceptionV3 outperformed previous approaches, achieving 97% of the matrices. Similarly, the modified DenseNet121 also showed improved performance, with 96% accuracy, precision, recall, and F1-score.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Sagar Sidana
PY  - 2024
DA  - 2024/10/17
TI  - Improving Sugarcane Disease Identification with L1-Regularized Transfer Learning Approach
BT  - Proceedings of the 2nd International Conference on Emerging Technologies and Sustainable Business Practices-2024 (ICETSBP 2024)
PB  - Atlantis Press
SP  - 152
EP  - 166
SN  - 2352-5428
UR  - https://doi.org/10.2991/978-94-6463-544-7_11
DO  - 10.2991/978-94-6463-544-7_11
ID  - Sidana2024
ER  -