期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B7
页码：1569-1574
出版社：Copernicus Publications
摘要：The unsupervised change detection of multitemporal SAR usually can get different results through different threshold selection algorithms. It is hard to determine what kinds of results are the best. In this paper, a novel automatic approach to the unsupervised identification of changes in multitemporal SAR image is proposed. This approach, unlike traditional ones, fuses many kinds of change detection results based on fuzzy logic theory. There are two steps in the proposed approach. In the first step, multitemporal SAR images are processed by many threshold selection algorithms separately, and several kinds of change detection maps are generated. In a second step, a framework for combing information from several individual threshold selection algorithms is proposed based on fuzzy logic theory. The robustness of the proposed approach is tested and validated with four threshold selection algorithms on SAR images in two regions. Experimental results, obtained on two sets of multitemporal SAR images, prove the validity and robustness of the proposed approach compared to each threshold selection algorithm
关键词：SAR; Unsupervised change detection; Information fusion; Fuzzy sets theory
