AI-Enabled Crop Disease Detection

Abstract

Agriculture plays a crucial role in sustaining human life, and plant health directly impacts food security and economic stability. However, plant diseases remain a persistent challenge, leading to significant crop losses and reduced yields. Traditional methods of disease detection, which rely heavily on visual inspection by experts, are often time-consuming, subjective, and inaccessible to many farmers. To address this challenge, this project presents an intelligent, automated system for plant disease identificationandpesticiderecommendationusing ConvolutionalNeuralNetworks(CNNs). Theproposed systemleveragesadeep learning-based CNN model trained on a comprehensive dataset of plant leaf images to accurately classify various plant diseases. Upon identification, the system provides targeted recommendations for organic pesticides to manage and mitigate the diagnosed disease effectively. The application is deployed as a user-friendly web platform, enabling users to upload plant images, receive instant diagnosis, and access curated pesticide suggestions. Through extensive testing, the CNN model achieved 95% accuracy, demonstrating its effectiveness in recognizing diverse plant diseases. The integration of organic pesticide data supports environmentallysustainablefarmingpractices. Usabilitytestswithrealusers, includingfarmersand agriculturalstudents, validated the system\'s ease of use and practical value in real-world scenarios.

Introduction

This paper presents an AI-based plant disease detection system that combines Convolutional Neural Networks (CNNs) with an organic pesticide recommendation module to support sustainable agriculture. Plant diseases caused by fungi, bacteria, viruses, and other pathogens significantly reduce crop yield and quality. Traditional disease identification methods depend on expert inspection, which is time-consuming, costly, and often inaccessible to farmers in rural areas.

The proposed system uses deep learning techniques, particularly CNNs, to automatically detect plant diseases from leaf images without manual feature extraction. Images undergo preprocessing steps such as resizing, normalization, and data augmentation to improve model performance and robustness. The CNN extracts visual features like textures, patterns, and edges and classifies plants into healthy or diseased categories.

The system is trained on a large dataset of labeled plant leaf images, divided into training, validation, and testing sets. After disease identification, an integrated recommendation module suggests organic and eco-friendly pesticides, including usage guidelines, helping farmers take immediate corrective action while promoting environmentally sustainable farming practices.

A Flask-based web application allows users to upload plant images and receive instant disease diagnosis along with treatment recommendations. Experimental results show strong performance, with an overall accuracy of 95%, precision of 92%, recall of 90%, and F1-score of 91%. Confusion matrix analysis indicates high reliability, with low false-positive and false-negative rates.

The study concludes that the proposed system is effective, user-friendly, and practical for real-world agricultural applications. By integrating disease detection with actionable recommendations, it provides a complete decision-support tool for farmers. Future improvements include expanding datasets, developing a mobile application, integrating real-time environmental data, and adopting more advanced deep learning models to further enhance accuracy and robustness.

Conclusion

Thispaperpresentsasmartplantdiseaseidentificationsystem using Convolutional Neural Networks, combined with an organic pesticide recommendation module. The proposed system achieves high accuracy and provides reliable results. Theweb-basedimplementationensuresaccessibilityandease of use, making it suitable for farmers and agricultural professionals. The system contributes to smart agriculture by integrating AI with sustainable farming practices.

References

[1] S. P. Mohanty, D. P. Hughes, and M. Salathé, \"Using Deep LearningforImage-BasedPlantDiseaseDetection,\"Frontiersin Plant Science, vol. 7, pp. 1419, 2016. [2] K.P.Ferentinos,\"Deeplearningmodelsforplantdiseasedetection and diagnosis,\" Computers and Electronics in Agriculture, vol. 145, pp. 311–318, 2018. [3] Z. Zhang, Y. Wang, and Y. Chen, \"Transfer Learning for Plant DiseaseClassification,\"JournalofPlantPathology,vol.101,no.1, pp.1–10,2019. [4] S.Sladojevicetal.,\"DeepNeuralNetworksBasedRecognitionof Plant Diseases by Leaf Image Classification,\" Computational Intelligence and Neuroscience, 2016. [5] R.M.Haralick,K.Shanmugam,andI.Dinstein,\"TexturalFeatures for Image Classification,\" IEEE Transactions on Systems, Man, and Cybernetics, vol. 3, no. 6, pp. 610–621, 1973. [6] O. Koul, S. Walia, and G. S. Dhaliwal, \"Essential Oils as Green Pesticides:PotentialandConstraints,\"JournalofPesticideScience, vol. 33, no. 1, pp. 1–12, 2008. [7] A. Krizhevsky, I. Sutskever, and G. E. Hinton, \"ImageNet Classificationwith Deep Convolutional Neural Networks,\" AdvancesinNeuralInformationProcessingSystems,2012.

Copyright

Copyright © 2026 Bora Harikrishna, Totakura Rohit Royal. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.