Comparative Study of Photogrammetric Image Matching Algorithms of Low Altitude UAV
The images obtained by UAV photogrammetry have advantages of high resolution and obvious texture characteristics,which are widely used in various fields of surveying and mapping industry. In view of the areas where the low altitude UAV acquires inconsis-tent features of ground objects,four image matching algorithms,namely SIFT,SURF,ORB and AKAZE,are used to match four types of different feature images of ground objects. The matching performance and characteristics of different algorithms for different ground objects are analyzed in depth. The stability of the four algorithms is evaluated by changing the image rotation angle,brightness and adding noise. The results show that SIFT and SURF are more stable in the process of matching images.
万方数据
维普
