Research On Traffic Statistics Algorithm Based On DSP Of Intensive Place

This topic is based on the study of human traffic statistics based on DSP intensive places.Firstly,the algorithm of human traffic statistics process is analyzed,and then the hardware design based on DSP is built.Accurately informed of the intensive places of the flow of people,can effectively carry out monitoring and management of intensive places to protect the public places pedestrians and property safety.This thesis mainly introduces four algorithms of image preprocessing,moving target detection,moving target recognition and moving target tracking in detail from four aspects.Image preprocessing is the image of the gray,histogram binarization and noise reduction processing.For the location of people with dense background,a more appropriate background difference method is used as the algorithm for the detection of moving objects.In-depth analysis of elliptic fitting and Hough variation.In many cases,there is a possibility that pedestrians or pedestrians are clogged in many cases,and Hough needs a degree of optimization to more effectively and accurately count the number of pedestrians.This will increase the amount of calculation,increase the computer and other hardware requirements.Compared with other algorithms,the elliptic fitting method described in this thesis is less complex and complex,and it can highlight its good effect.This thesis chooses the ellipse fitting method as the recognition method of the target.A tracking method suitable for this thesis is designed based on the trajectory Kalman filter tracking algorithm and Meat Shift tracking algorithm.In this thesis,the hardware structure of the most important to take three modules,using the United States Texas Instruments(TI)production TMS320DM642 DSP chip image processing module,Sony CCD camera image acquisition and coding TVP5150 chip image acquisition module,Using TCL T22M2 liquid crystal display and SAA712IH display module.Combined with the introduction of the flow of the algorithm and the proposed hardware platform to provide a verification and simulation.Through the statistics on the flow of people in intensive places,you can understand and arrange the reasonable density of these entrances and exits in real time.It can allocate management and maintenance personnel more effectively by observing the changes of people's traffic.Real-time understanding of the dynamics of the flow of people in intensive places,managers can effectively change the trend of these people to take measures to prevent emergencies,to protect pedestrians of personal and property safety,to maintain social stability.