Saline patch identification using a semi-supervised knowledge integration deep learning framework based solely on UAV RGB imagery

Zhen Li

doi:10.1016/j.compag.2025.111061

Saline patch identification using a semi-supervised knowledge integration deep learning framework based solely on UAV RGB imagery

2025-11-08

0

PRE

AI

DOI：10.1016/j.compag.2025.111061

Original

Original request for help

Save

Abstract

En 中文

• Low-cost UAV RGB imagery was solely used for soil saline patch mapping. • A semi-supervised framework using limited labeled samples was proposed. • Salinity-related knowledge and texture-semantic joint information were integrated. • Multi-year experiments (2022–2024) verified strong temporal generalization ability.

Saline patch identification using a semi-supervised knowledge integration deep learning framework based solely on UAV RGB imagery

Abstract

Journal

Organization