Based On The Kinect People Tracking Algorithm

In the dail y management of large supermarkets,market survey done by manpower was often used to collect some data information of customers ' activities in the early stage,but it was time-consuming and inefficient.However,it has become more convenient and flexible because guidance is provided for the management after the People Target Detection and People Motion Tracking Algorithm are adopted.With the computer software processing,these two methods can anal yze the number of customers in a particular area of the supermarket,their coordinates and the time when they come and leave,and thus figure out the current best-selling commodities and their reasonable placement,which greatl y increases the sales.Through the anal ysis and research on the People Target Detection Algorithm,this essay tries to find out a solution to reducing the change of backgroun d and the influence of noise interference in the video images,analyzes and optimizes the Difference Background Subtraction Algorithm,and compares the two characters ' tracking localization algorithms – the CAMShift Algorithm and the Kalman Filtering Algorithm.It also focuses on how to build a character contour model to analyze character actions and use the Kinect technology to build a system model.With the above methods,customers ' behavior will be anal yzed to collect the data of their activities in a pa rticular area,and their characteristics will be also analyzed b y computer technology to produce valid data that can be used.The valid data from this experiment will finally be analyzed in the essay,which shows that this system model has certain robustn ess in some aspects,such as detecting and tracking people targets,and analyzing their activities.