Plant Disease Prediction with the Help of Deep Learning based on the Analysis of Leaf Images

Abstract

The issue of plant diseases remains a significant threat to the worldwide food production and requires solutions to disease detection that are fast, accurate and scalable. Conventionally applied methods of diagnosis, including: hand field scouting and lab tests, are too slow, labor-intensive, and subject to human error, particularly during large scale cultivation. Recent technological progress in Artificial Intelligence (AI) and Machine Learning (ML) has brought in the groundbreaking solutions to early and accurate prediction of the disease. In this paper, the author will review how AI-based models, such as Convolutional Neural Networks (CNNs), Random Forest, Support Vector Machines (SVM), and hybrid deep learning models can be used to detect plant diseases on the basis of visual information, as well as environmental data. These models are very productive in recognizing patterns, automated feature extraction, and real-time prediction, and they are frequently much more precise than their human counterparts. Early warning systems and precision agriculture are further improved by the integration of the IoT sensors, drones, and satellite imagery. Regardless of the issues concerning the lack of data, model generalization, computational cost, and adoption of AI and ML on a farmer level, there are considerable prospects to enhance crop protection, minimizing yield losses, and sustainable agriculture. The paper identifies the current progress, limitations and future research in the development of strong, availability and scaleable AI-based plant disease prediction systems.

Plant phenotyping relevance

植物の葉画像から病徴・病害状態を推定するAI手法を中心に扱うレビューであり、植物フェノタイピング手法レビューに該当する。

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.