Design And Implementation Of Personnel Detection Subsystem In A Public Transportation Complex

The video surveillance system has become an extremely effective tool in the modern management of the complex because it can reflect the situation in real time,truthfully and vividly.At present,video target recognition and tracking have been widely used in many industries such as national defense,science and engineering.If it can also be used in the public transportation complex,we can identify and track the flow of people in the complex through moving target detection and tracking technology,and at the same time obtain the moving speed and trajectory,further improve the management level and reduce the incidence of safety accidents.In order to realize the detection and tracking of people in the complex,this paper uses software to simulate the video image preprocessing,the detection of moving targets,and the detection and tracking of people.The basic concepts and related knowledge points of digital image processing are introduced to achieve low illumination The image enhancement method for specular reflection points uses the inter-frame difference method to achieve moving target detection,and then uses HOG image features,cascaded classifiers,Mean Shift algorithm and Kalman filter algorithm to achieve human detection and tracking.Experiments prove that the method used can improve the effect of detection and tracking.