Single Camera-based Algorithm for Depth Map Generation in UAV

In this paper, a method of depth map generating based on the single camera of the UAV is presented. First, feature point matching is performed on sequential image sequences. Then, the feature points are matched to the larger scale to find the current frame image and the previous larger scale of the frame image. Next, according to the UAV speed, image frame rate and the larger scale to calculate the depth information of these feature points. Finally, the deep map is generated according to the depth information. The experiment proves that the target which has obvious features can be detected by the method mentioned in this paper to generate the depth map, as a result, the path planning of the UAV can avoid obstacles to reach the destination.