Spatial-Temporal Evolution Guided Change Detection Network for Remote Sensing Images
Qingwang Wang, Hong Zheng, Jiangbo Huang, Xiaobin Zhao, Jian Song, Kai Zeng, Jianwu Shi, Tao Shen
Abstract
With the rapid advancement of remote sensing technology, bi-temporal remote sensing change detection (CD) techniques have also seen significant progress. However, existing CD tasks still face two challenges: 1) Variations in lighting and seasonal factors complicate imaging conditions, causing pseudo-variation interference, and 2) The spatial distribution and shapes of building are diverse, leading to difficulties in extracting and utilizing effective change features. In this paper, we propose the spatial-temporal evolution guided change detection network (STEGNet) to capture and fully utilize rich spatial-temporal information. Specifically, we develop the chrono colorizer to mitigate pseudo-variant interference by standardizing color styles and enriching time series information. Additionally, we introduce the temporal-spatial guidance module (T-SGM), which combines spatial-temporal information to guide the decoding operation and mitigate information loss during spatial-temporal fusion, resulting in finer prediction results. Experimental results on three benchmark datasets demonstrate that STEGNet effectively suppresses pseudo-variation interference and significantly improves the integrity of detection boundaries.