利用模糊综合评判进行面向对象的遥感影像变化检测  

Object-oriented Change Detection for Remote Sensing Images Based on Fuzzy Comprehensive Evaluation

作　　者：冯文卿[1] 张永军[1] 

机构地区：[1]武汉大学遥感信息工程学院,湖北武汉430079

出　　处：《武汉大学学报（信息科学版）》2016年第7期875-881,共7页Geomatics and Information Science of Wuhan University

基　　金：国家自然科学基金(41322010);国家863计划(2013AA12A401)~~

摘　　要：提出了利用模糊综合评判进行面向对象的遥感影像变化检测方法。首先对遥感影像进行多尺度分割来获取对象;然后,进行对象的特征优选,利用变化向量模法、χ2变换法、向量相似度法、相关系数法来构造"综合层间逻辑值",并作为因素集建立模糊综合评判模型,对目标对象是否发生变化做出判决。最后与单个"层间逻辑值"进行OTSU阈值分割的结果进行对比,证明了该方法的可行性。In the process of object-oriented change detection,the accuracy of the final result is directly related to the change threshold.Aiming at this problem,this paper presents a novel object-oriented change detection method using fuzzy comprehensive evaluation.Firstly,multi-scale segmentation is used to obtain initial objects;then,optional features for each object are chosen.Several criteria,such as objects change vector analysis,Chi-square transformation,the similarity of vector,and correlation coefficient,are treated as factors to get the＂synthetic inter-layer logical values＂of the fuzzy comprehensive evaluation model.The fuzzy comprehensive evaluation model is used to decide whether the target object has changed or not.Finally,the result of fuzzy comprehensive evaluation model is compared with the result of each single＂inter-layer logical value＂that using OTSU threshold segmentation.Based on this theory,experiments are done with SPOT5multi-spectral remote sensing imagery.The experimental results illustrate that the model proposed can integrate the spectral and texture features and also overcome the defects caused by using single criteria.The fuzzy comprehensive evaluation model is proved to outperform other methods.

关 键 词：像变化检测 面向对象 特征 层间逻辑值 模糊综合评判 

分 类 号：P751[交通运输工程—港口、海岸及近海工程]