Deep Learning Approaches for Plant Disease Diagnosis Systems: A Review and Future Research Agendas

Abstract

To identify novel advancements in plant diseases detection and classification systems employing Machine Learning (ML), Deep Learning (DL), and Transfer Learning (TL), this research compiled 111 peer-reviewed papers published between 2019 and early 2023. The literature was sourced from databases such as Scopus and Web of Science using keywords related to deep learning and leaf disease. A structured analysis of various plant disease classification models is presented through tables and graphics. This paper systematically reviews the model approaches employed, datasets utilized, countries involved, and the validation and evaluation methods applied in plant disease identification. Each algorithm is annotated with suitable processing techniques, such as image segmentation and feature extraction, along with standard experimental metrics, including the total number of training/testing datasets utilized, the quantity of disease images considered, and the classifier type employed. The findings of this study serve as a valuable resource for researchers seeking to identify specific plant diseases through a literature-based approach. Additionally, the implementation of mobile-based applications using the DL approach is expected to enhance agricultural productivity.

Plant phenotyping relevance

植物病害を画像から検出・分類する深層学習手法を体系的にレビューしており、植物の病害状態を推定する方法論が中心です。

Code and data availability

This is a systematic literature review of plant disease detection papers; it presents no original phenotyping measurements, images, code, or models of its own. The data availability statement only offers the authors' referenced articles or data upon request, with no public deposit or URL, and all cited datasets (e.g.,

No evidence-backed public reproduction asset is currently recorded.