Early Identification and Classification of High-Impact Cotton Plant Diseases through IoT

Abstract

Cotton, often referred to as “white gold,” is one of India's most critical cash crops, forming the backbone of both the agricultural and textile sectors. However, despite its economic significance, cotton cultivation is increasingly threatened by the emergence of severe plant diseases such as Bacterial Blight, Cotton Leaf Curl Virus (CLCuV), and Fusarium Wilt. These diseases not only diminish yield but also exacerbate rural distress, especially in drought-prone regions like Marathwada, Maharashtra. This research explores the biological characteristics and economic impact of these diseases, reviews conventional and advanced detection methods, and proposes a hybrid technological framework involving IoT sensors and Convolutional Neural Networks (CNN) for early disease diagnosis. With real-time monitoring, image-based classification, and predictive analytics, this model aims to empower farmers, reduce production losses, and promote sustainable cotton farming practices.

Plant phenotyping relevance

綿花の病害症状を画像・IoTセンサー・CNNで早期診断する枠組みを中心的に提案しており、植物の病害状態を直接推定するフェノタイピング手法に該当します。

Code and data availability

The paper describes a CNN (ResNet34/FastAI) + IoT cotton disease detection system, but contains no public dataset deposit, no author code/model release, and no availability statements. The PlantVillage dataset is cited prior work, not a paper-specific asset; the trained model (cotton_disease_model.pkl) is mentioned but

No evidence-backed public reproduction asset is currently recorded.