Research And Implementation Of Multi-Camera Scheduling System Based On Face Tracking And Recognition

With the continuous progress of scientific and technological civilization,the society nowadays has put forward higher requirements for security.As an important part of security engineering,video surveillance has attracted the attention of more and more researchers.However,most of the traditional video surveillance systems are still limited to the monitoring of specific areas in a single camera environment.Due to the limited field of view and the lack of target recognition capability,it is difficult to achieve long-term stable tracking of the moving target.Aiming at the shortcomings of traditional video surveillance,this thesis analyzes the actual scenes of face tracking and multi-camera scheduling,and proposes corresponding improved algorithms to solve the difficult problem of face tracking using multi-camera in complex environment.Futhermore,this thesis combines face detection and face recognition technology to realize a multi-camera scheduling system based on face tracking and recognition.The system firstly uses AdaBoost algorithm to detect the face in the video image,and then compares the detected face with the existing face in the face database by LBP histogram algorithm.If the similarity is less than the threshold,the recognition succeeds,then the face target begins to be tracked.In view of the defect that the traditional CamShift algorithm has low tracking accuracy under the complex background,the system introduces the FAST corner detection algorithm and the SIFT feature description algorithm to correct the tracking result with deviations,and improves the real-time performance of the tracking algorithm.Finally,through the priority-based multi-camera scheduling algorithm,the tracking effects of the cameras are quantified,and the camera with the best tracking effect is selected,whose captured video images are taken as the final output images of the system.Experimental results show that the system can achieve continuous and stable tracking of the face target,which has high practicability.