Fast Target Three-Dimensional Reconstruction Algorithm for Streak Tube Laser Imaging Radar

Real-time dynamic target tracking and recognition has become an urgent issue that should be resolved with respect to streak tube laser imaging radar. In this study, we propose a sampling interpolation algorithm for target three-dimensional (3D) reconstruction to solve the problem of long operation time associated with the peak detection 3D reconstruction algorithm. First, the streak image is sampled with a certain sampling rate at equal intervals and is interpolated; subsequently, the peak feature is extracted. When the sampling rate is 1%, the sampling bilinear interpolation reconstruction method is approximately 2.3 times faster than the original reconstruction method, and the correlation coefficient of the 3D reconstruction matrix is about 0. 99 for targets with a detection distance of 706 m; further, in case of the nearest neighbor sampling interpolation method, the reconstruction speed is 3. 5 times faster than that of the original reconstruction method, and the correlation coefficient of the 3D reconstruction matrix is about 0. 98. The experimental results denote that the proposed algorithm can improve the 3D object reconstruction speed; the larger the number of frames of the target streak image is, the better the effect of improving the 3D target reconstruction speed will be.