Plant Leaf Disease Detection Using Multiple CNN Models

Abstract

The earlier detection of diseases in plant leaves is very important in agriculture. This task requires huge time and someone with great knowledge about the subject. This project aims to develop a robust Web App capable of taking the input of an image and accurately identify plant diseases. A dataset with photographs of multiple types of plant leaves in both healthy and ill situations is selected. Multiple CNN models, including ResNet50 and a Custom CNN, were trained, achieving validation accuracies of 93.96% and 95.58%, respectively. The proposed system provides a scalable and efficient tool for precision agriculture, enabling farmers to address diseases proactively.

Plant phenotyping relevance

植物葉の画像から病害状態を推定するCNNモデルとWebアプリを開発しており、植物の病徴・病害状態の画像ベース推定が中心的な方法貢献です。

Code and data availability

公開本文の所在を確認できませんでした。非公開または購読が必要な可能性があります。

No evidence-backed public reproduction asset is currently recorded.