Moving Target Detection Algorithm Based on SIFT Feature Matching

After the computer technology is mature, the sequence images obtained from the camera are processed by the computer, and the images converted into digital signals are processed by the computer, so the new discipline of computer vision is born. The reason why we should pay attention to computer vision is to use computers to replace the human brain and human eyes to extract, identify, track and other series of understanding and analysis of moving targets in specific scenes. The purpose of this paper is to research the moving target detection algorithm based on SIFT feature matching. According to the current video surveillance requirements for moving target detection, reduce the number of candidate samples in the input detection module in the TLD algorithm, so as to optimize the TLD algorithm and make it meet the real-time requirements. A tracking module that uses SIFT algorithm to optimize the TLD algorithm is proposed, and the TLD algorithm based on SIFT feature matching is tested. The experimental results show that the tracking accuracy of the algorithm reaches more than 90%, and the algorithm has better robustness to the rotation of moving objects.