基于多特征融合的遥感影像变化检测研究OA

Change detection in remote sensing imagery has been widely applied in urban plan-ning,environmental monitoring,natural disaster prevention,resource exploration,and land-use investigation.Multi-feature fusion enables the comprehensive extraction of spectral and spatial information,leveraging the strengths of different features to enhance detection accu-racy.This study proposes a hyperspectral change detection method based on multiscale mor-phological attribute filtering.Spectral features(SALA)and morphological features(AP)were extracted separately from bitemporal image pairs,and a guided-filtering strategy was employed to fuse the features and produce the final change map.Experimental results dem-onstrate that when the false-alarm rate is fixed at 0.15,the proposed morphological-based approach achieves detection rates greater than 0.89.Compared with the traditional AD algo-rithm,the proposed method exhibits superior detection performance.