Research of Embedded Real-Time Passenger Flow Detection Equipment

摘要

In order to further improve the accuracy and real-time performance of the passenger flow detection equipment, an improved passenger flow detection equipment based on embedded development platform was proposed. The equipment uses a new type of deep camera to gather the video image. We use the pedestrian detection technology based on the characteristics of head and shoulder on the embedded development platform to get real-time passenger flow information, and the technology based on the classical HOG algorithm, at the same time, the proposed technique can effectively improve the efficiency of the algorithm based on the depth information. Using network transport protocol implementation Ethernet high-speed transmission of data and video. The experiment shows that the equipment can achieve the real-time and accurate detection of the passenger flow in the complex environment. The average precision is 93.85% and the running time is 21 fps. Using the Ethernet interface, it has the advantages of easier development, lower cost, faster transmission speed, and remote control.