CO-REGISTRATION OF MULTI-TEMPORAL DEM BASED ON SIFT ALGORITHM FOR CHANGE DETECTION OF GLACIERS

Abstract. To detect the change of geographic objects by using multi-temporal DEM, the data must be co-registered firstly. In this paper, the Scale-Invariant Feature Transform (SIFT) algorithm is used to co-register multi-temporal DEM data and glacier change detection. Firstly, the DEM is converted into image space and extracts feature information, calculate multiple sets of match point coordinates, and achieve swift and accurate DEM data co-registration using SIFT algorithm. Secondly, the difference between co-registered DEM datasets are analysed. Total area change and average rate of change are calculated. Finally, the change of multi-temporal DEM data of glaciers in Langkazi County, Tibet from 2004 to 2014 is detected using the method proposed in this paper. The results indicate that the proposed method is able to detect change of the glaciers and the overall accuracy is higher than 85 %.