PLANT DISEASE IDENTIFICATION AND PESTICIDES RECOMMENDATION SYSTEM USING CONVOLUTIONAL NEURAL NETWORK FOR PROTECTION

Abstract

Agriculture is one of the most important sectors contributing to economic development and global food security. However, plant diseases caused by fungi, bacteria, viruses, and other pathogens significantly reduce crop yield and quality, leading to substantial economic losses for farmers. Early and accurate identification of plant diseases is essential for effective crop management and timely application of suitable pesticides. Conventional methods of disease diagnosis rely on manual inspection by agricultural experts, which is time-consuming, labor-intensive, expensive, and often inaccessible to farmers in remote areas. Recent advancements in Artificial Intelligence (AI) and Deep Learning have provided efficient solutions for automating plant disease detection through image analysis. This project, "Plant Disease Identification and Pesticides Recommendation System Using Convolutional Neural Network (CNN) for Crop Protection," presents an intelligent system that automatically identifies plant diseases from leaf images and recommends appropriate pesticides for effective crop protection. The proposed system utilizes a Convolutional Neural Network (CNN), a deep learning model specifically designed for image classification tasks. The CNN model is trained using a large dataset of healthy and diseased plant leaf images collected from publicly available agricultural datasets. During training, the model learns to recognize disease-specific visual features such as color variations, lesion patterns, texture changes, and leaf deformities. Image preprocessing techniques, including resizing, normalization, and data augmentation, are employed to improve the quality of the input images and enhance the overall performance of the model. When a farmer uploads an image of a plant leaf through the system, the trained CNN model analyzes the image and accurately classifies it as either healthy or affected by a specific disease. After identifying the disease, the system recommends suitable pesticides, fungicides, insecticides, or biological treatments based on an agricultural knowledge database. It also provides additional information such as recommended dosage, application method, spraying schedule, safety precautions, and preventive measures to ensure responsible pesticide usage and minimize environmental impact. The proposed system offers several advantages, including rapid disease detection, high classification accuracy, reduced dependence on agricultural experts, optimized pesticide application, lower crop losses, improved productivity, and support for sustainable farming practices. Furthermore, the system can be deployed as a web or mobile application, enabling farmers to access disease diagnosis and treatment recommendations anytime and anywhere using smartphones or other digital devices. Overall, the proposed CNN-based plant disease identification and pesticide recommendation system provides a reliable, cost-effective, and intelligent solution for modern agriculture. By combining image processing, deep learning, and agricultural expertise, the system supports precision farming, enhances decision-making, reduces unnecessary pesticide usage, and contributes to increased crop productivity, environmental sustainability, and long-term food security.

Plant phenotyping relevance

葉画像から植物病害の状態をCNNで直接推定する手法が研究の中心であり、病害症状の画像ベース表現型計測に該当する。農薬推薦も含むが、植物病害識別というフェノタイピング要素が明示的である。

Code and data availability

The paper describes a CNN-based plant disease identification and pesticide recommendation system but provides no public dataset URL, no author code/model deposit, and no supplement with phenotyping data. It only vaguely references 'publicly available agricultural datasets' without naming or linking any, and the listed/

No evidence-backed public reproduction asset is currently recorded.