Application of Automatic Statistic Passenger Flow System Based on Depth Data
Xuebin Qin, Yizhe Zhang, Guo Lin, Mei Wang, Pai Wang · 2016
In the monitoring industry, such as shopping malls, bus stations, etc., it is necessary to analyze the passenger flow system. Traditional statistical method, the influence factors on the performance of the system, such as illumination, occlusion, body posture. This paper presents a new method based on kinect2.0 sensor depth data of passenger flow detection system, are determined by the shape of the depth data and the head of the human brain, area, then the system by tracking based on Kalman filter motion realization all traffic statistics. Experiments show that the processing speed of 30 frames per second depth image, the accuracy of the method is 96.7%. To solve the problem of occlusion and illumination in color image. The system has good practical value.