An automated dual-module AI-based solution for early detection and classification of crop diseases and stress conditions

Abstract

Abstract One of the most important challenges faced by smallholder farmers in the agricultural industry is the lack of accurate, timely knowledge to predict and detect crop health issues. Crop productivity is often threatened not only by diseases but also by environmental and physiological stresses, which contribute to significant yield losses and negatively impact the national economy. Traditional detection methods are time-consuming, costly, and require expert knowledge, creating a need for automated and intelligent systems. This work proposes a dual-functional framework that combines crop disease and stress detection using advanced machine learning and deep learning techniques to accurately classify healthy and diseased leaves. This integrated system ensures early detection, reduces crop loss, improves productivity, and provides a scalable, farmer-friendly solution for sustainable agriculture.

Plant phenotyping relevance

葉画像から健康・病害状態を機械学習で分類する手法が研究の中心であり、植物の病害・ストレス状態を直接推定するため、植物フェノタイピング手法として採用。

Code and data availability

The paper uses two public Kaggle datasets (the New Plant Diseases/PlantVillage dataset for disease detection and a crop-health sensor dataset for stress detection), but the authors' merged/refined dataset is explicitly not publicly hosted and is available only upon reasonable request from the corresponding author. No作者

No evidence-backed public reproduction asset is currently recorded.