GCOANet: A gradient consistency constraints semi-supervised network for optical image-assisted SAR despeckling

• Published in: International journal of applied earth observation and geoinformation Vol. 142; p. 104677
• Main Authors: Yang, Yang, Pan, Jun, Xu, Jiangong, Fan, Zhongli, Geng, Zeming, Li, Junli
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2025; Elsevier
• Subjects: Blind spot network; disaster preparedness; image analysis; Image despeckling; Multi-source image fusion; Semi-supervised method; signal-to-noise ratio; spatial data; synthetic aperture radar; Synthetic aperture radar (SAR); texture; Synthetic aperture radar (SAR); Blind spot network; Semi-supervised method; Multi-source image fusion; Image despeckling

•Optically-guided SAR despeckling with semi-supervised learning.•Rapid cross-domain feature generation with non-local filtering and conditional diffusion model.•Multi-scale blind spot network and interval downsampling strategy.•All-directional gradient loss for homogeneous and heterogeneous supervis...