Research On Pedestrian Tracking Algorithm Based On TLD

The problem of tracking is the hotspot in the field of computer vision.With the gradual improvement of social and public security system,public places in the security intelligent monitoring system requirements are getting higher and higher,which makes many well-known companies and research institutions in this subject to invest a lot of manpower and resources.Detection and tracking of human moving targets is an important part of vision tracking.It will have broad prospects in the field of future research and can bring huge social benefits.Pedestrian tracking mainly includes three aspects of the study,the first is the image processing,and then the dynamic target detection,the third is the dynamic tracking algorithm.Image processing includes grayscale processing,image denoising and filtering,morphological processing and other basic methods.Target detection and identification use the concept of image window and the regional growth method to simplify the window,to achieve pedestrian target recognition.An improved dynamic pedestrian tracking algorithm based on TLD(tracking-learning-detection)is proposed to solve the problem of tracking drift in dynamic pedestrian tracking when the pedestrian is occluded.After tracking the pedestrian target successfully,an occlusion window method is incorporated to determine the degree of the pedestrian occlusion.When the pedestrian is partially occluded,the TLD learner is utilized to solve the problem.When the pedestrian is seriously covered,the Kalman filter is added into the TLD tracker to estimate the area in the current frame where the pedestrian may exist.Then the search range is reduced,which improves the processing speed of the tracker.Meanwhile,the direction predictor based on Markov model is added into the TLD detector to reduce the detection range of the detector and enhance the discrimination ability for the similar pedestrian trajectory.The experiment shows that the improved TLD algorithm has strong robustness in different environments and can quickly and accurately track the dynamic pedestrian trajectory.The proposed algorithm also improves the tracking drift problem when the pedestrian is occluded.The system efficiency is greatly improved through the algorithm optimization.