江西科学2026，Vol.44Issue(1)：69-75,7.DOI:10.13990/j.issn1001-3679.2026.01.010
基于多特征融合的遥感影像变化检测研究
Change Detection of Remote Sensing Imagery Based on Multi-feature Fusion
摘要
Abstract
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.关键词
变化检测/形态学属性剖面/多特征融合/PCA/引导滤波Key words
change detection/morphological attribute profile/multi-feature fusion/PCA/guided filtering分类
信息技术与安全科学引用本文复制引用
吴彤,赵明松,李军军,盛乐,陈信江,张道宇..基于多特征融合的遥感影像变化检测研究[J].江西科学,2026,44(1):69-75,7.基金项目
安徽省自然科学基金项目(2208085MD88) （2208085MD88）
安徽理工大学研究生创新基金项目(2024cx2151). （2024cx2151）
