NOCL Based Automatic Leaf Disease Detection with High Accuracy

Sindhuja V, Latha B, Dhanushiya S, Dharani A C, Swetha D, Udhika V

2025

Abstract

Aim: The study introduces the Neural Optimization and Classification Logic (NOCL) method, a new deep learning method to directly diagnose plant leaf diseases, which aims to reduce the mortality rate of plants and improve agricultural productivity, as well as promote sustainable agriculture. Materials and Methods: The proposed Neural Optimization and Classification Logic (NOCL) model was implemented using Python and TensorFlow designs and tested on a dataset with diseases and healthy leaves. Group 1 refers to the NOCL method that integrates advanced optimization techniques and classification logic with 1000 testing counts. Group 2 refers to the traditional CNN method. The NOCL-based architecture shows direct detection capability, providing a 25% improvement in classification accuracy and a 35% reduction in computational complexity, compared to the CNN method. Result: The proposed method uses the Ka15e leaf disease dataset, which contains a total of 78,456 images, which includes 75% as training data and 25% testing data. The NOCL method achieves 96% of accuracy, F1 score and recall and performs better than traditional methods. It classifies healthy and diseased specimens by examining signs such as black dots, mosaics also greenish ventral patterns on the leaves. For comparison, the CNN model with 7 convolution layers was used, which showed an accuracy of 90.26% to 92.16% whereas, F1 score and recall less than the proposed method. The NOCL model is implemented through the Python language and is efficient in training and validation with a significance of p < 0.05. Conclusion: The recommended NOCL based strategy provides an efficient and reliable method for detection of plant diseases, which enables farmers to take measures for early control of diseases and ensure sustainable agriculture through robust crop production practices.

Paper Citation

in Harvard Style

V S., B L., S D., A C D., D S. and V U. (2025). NOCL Based Automatic Leaf Disease Detection with High Accuracy. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 38-44. DOI: 10.5220/0013876400004919

in Bibtex Style

@conference{icrdicct`2525,
author={Sindhuja V and Latha B and Dhanushiya S and Dharani A C and Swetha D and Udhika V},
title={NOCL Based Automatic Leaf Disease Detection with High Accuracy},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={38-44},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013876400004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - NOCL Based Automatic Leaf Disease Detection with High Accuracy
SN - 978-989-758-777-1
AU - V S.
AU - B L.
AU - S D.
AU - A C D.
AU - D S.
AU - V U.
PY - 2025
SP - 38
EP - 44
DO - 10.5220/0013876400004919
PB - SciTePress