LesionDiff: Synthetic data via lesion information transfer diffusion model facilitates plant disease diagnosis

The Crop Journal 2026 Open access

Read full text
Authors: Xingcai Wu, Qiaoling Wan, Ya Yu, Yujiao Dan, Hanying Xie, G.M.A.D Sirishantha, Qi Wang, Gefei Hao, Yongjin Liu
DOI: 10.1016/j.cj.2026.01.011
Cited by: 3
References: 39

Abstract

Training software models for crop disease diagnosis requires large image datasets to achieve high accuracy. We describe a lesion information transfer diffusion model, LesionDiff, for generating image data that augments a real-world disease lesion image dataset. An information preprocessing module identifies lesion areas on leaves, an enhancement module captures diverse visual and semantic lesion features, and a generation module fills missing regions in masked disease images by synthesizing lesion phenotypes. This augmentation increased the average diagnostic accuracy of a test dataset by more than 3%.

Fields of Study & Subjects

Smart Agriculture and AI Plant Virus Research Studies AI in cancer detection

Publication details

Journal / source
The Crop Journal
Publisher
Elsevier BV
Volume
14
Issue
3
Pages
1051–1063
Publication date
27 February 2026
Language
EN