Abstract:

In UAV aerial triangulation,feature point matching is mostly processed by matching algorithm based on computer vision,to supplement the number of ground control points.Because of the high resolution and rich details of the UAV image,the number of matching feature points is huge,which brings difficulties to the subsequent block adjustment.The points thinning algorithm is provided in this paper,which greatly reduces the number of matching feature points,preserves the high quality points first,and ensures the uniform distribution of the reserved points.

Key words: UAV, aerial triangulation, image matching, points thinning algorithm