Line segment matching of space target image sequence based on optical flow prediction

On the purpose to access the motion and structure information of space targets, a line segment matching method of space target sequence based on optical flow prediction was proposed. Firstly, the Edge Draw Lines (EDLines) algorithm was adopted to extract line segments, and then the optical flows of pixels on the line segments and Line Band Descriptor (LBD) of the line segments were calculated. Optical flow assessment was applied to predict the positions of line segments in the next frame. The line segment matching of image sequence was achieved by choosing the distance between the descriptors as the similarity measure. Compared with existing algorithms, the simulation results show that the method can efficiently reduce the number of correspondent candidate matches of every line segment, decrease the computational complexity in the matching process, and accurately accomplish the task of line segment matching of space target image sequence.