A Summary on Current Techniques and Prospects of Remote Sensing Change Detection

Change detection is one of the important topics in remote sensing application field.In this paper,the main methods applied in change detection were firstly introduced and evaluated.The problems existing in the currents techniques were analyzed,and then a novel image segmentation based change detection approach was proposed.In this method,one of the two images is firstly segmented into homogeneous regions,and then the second image is divided into regions using the same polygons generated in the first image.Finally,the two images are compared by region to region using the established detection rules.This method is expected to improve the accuracy and the speed of change detection to some extent.At last,the advantage of the image segmentation based change detection approach was presented by being compared with the pixel-based change detection method.