The comparative study of three methods of remote sensing image change detection

This paper discusses three main methods of change detection: 1) image subtraction method; 2) image ratio method; 3) the method of change detection after classification. Firstly, the elimination method of influence factors of change detection is discussed. Then the basic principle of the three main methods is introduced and the experiments of the methods are carried on ERDAS software. At last, the analysis comparison is carried on and the relative merits and the applicable scope of the three methods are pointed out. Image subtraction method is a simple concept easy to understand and easy to use, a background value usually be repressed and a subtraction value often be enhanced in the result image. It is beneficial to information extraction, which value of the target and background is smaller, such as the beach zone, the ditch of estuaries. The main disadvantage is that it can not reflect which category is changed. Image ratio method is applicable to be used in change detection of city. Its disadvantage is also that it can not reflect which category is changed. The information of change property is provided in this method. The disadvantage is that the accuracy depends on the classification accuracy; it can not be used for the detail change detection of city.