利用指数熵的多时相遥感影像变化检测方法

摘要

针对像素级变化检测法中变化阈值的提取不够自动化和准确化,导致变化检测结果精度不高的问题,提出一种利用双阈值指数熵的多时相遥感影像变化检测方法.该方法首先采用差值法构造2个时相遥感影像的差异影像;其次采用双阈值指数熵的方法确定差异影像的最佳变化阈值,并将其用于分割差异影像,得到变化区域.采用客观评价法对变化检测结果进行精度评定.选择我国鄱阳湖局部区域2个时相的遥感影像进行试验,并与基于模糊C均值的变化检测方法进行对比.通过试验,所提出方法变化检测精度达94.22％,是一种有效、可行的变化检测方法.%The change threshold of pixel level change detection method is not automatic and accurate,which leads to the problem of low accuracy of detection results.Aiming at this problem,a multi-temporal remote sensing image change detection algorithm based on dual-threshold exponent information entropy is proposed in this paper.It firstly obtains difference image of multi-temporal remote sensing image by the difference method.Then,the best threshold of change detection in difference image is deterred by dual-threshold exponent information entropy method,the different image is segmented by dual-threshold exponent information entropy thresholding segmentation method,and the change regions are obtained.The accuracy of change detection is then analyzed by objective evaluation methods.The bi-temporal remote sensing images in local regions of China's Poyang Lake are selected to verify the proposed method.And the proposed method is compared with change detection method based on fuzzy C-means.Experimental results show that this new algorithm can achieve higher accuracy about 94.22％,also,the poposed method is an effective method in change detection.