Method for moving point target detection in image sequences based on directional cumulation
Shan Gao, Peng‐Lang Shui · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2007
A new algorithm based on directional cumulation is presented to detect the trajectory of moving point target from image sequences in low SNR environments. First, single frame detection is performed, and then the logical "or" operation of multiple frames are used to obtain a binary planar projective image. With that based on the continuity of the moving trajectory of the point target, the second detection using directional cumulation is performed to acquire the trajectory of the target and its parameters. The result of experiments shows that the proposed method is effective to detect and track targets moving along linear or curvilinear trajectory with unrestricted velocity in low SNR, and it is of low complexity.