Moving target detection algorithm combined background compensation with optical flow

In order to solve the problem of moving target detection caused by the ineffective calculation of the optical flow and the insufficient background compensation, an algorithm combined dynamic background compensation and optical flow is proposed. SURF (Speeded Up Robust Features) algorithm is adopted to extract the matching points and the iterative threshold segmentation algorithm is used to filter the outside point to improve matching accuracy. The motion estimation parameters are estimated by using the least-square theory. On the basis of accurate background compensation, the LK (Lucas-Kanade) optical flow algorithm is used to detect moving targets, which effectively reduces invalid background light flow calculation as well as the effect to target recognition and improves the motion target detection accuracy. Finally, VC++ and OpenCV software platform is used to design the system environment and realize the detection of moving objects in the scene of moving background. The simulation experiment results verified the feasibility of the proposed algorithm.