The Research Of Video Traffic Flow Detection Method Based On FPGA Technology

In recent years, with the rising number of cars on city roads and rural roads,the traffic problem has become one of the focus of people’s livelihood issues. How to monitor and manage the traffic efficiently has become a hot spot of people care about. Compared with the traditional ultrasonic testing and induction coil detection, digital image processing monitoring system and traffic management based on FPGA platform have many advantages, such as convenient installation and maintenance, low cost, wide scope of monitoring, fast processing speed and getting more kinds of types. This paper expounds the research status at home and abroad on FPGA platform of video vehicle detection firstly from the video vehicle detection technology, algorithm and the FPGA platform research. In order to provide a reference direction to traffic video detection technology development, this paper analyses the basic principle of digital image processing based on FPGA, counting from the video vehicle detection and tracking algorithm and the FPGA system building.Algorithm mainly includes detection algorithm and tracking algorithm counts. In view of the detection algorithm, this paper analyzes the three kinds of commonly used video vehicle detection algorithm and proposes a dynamic updating method which is based on the background difference method, using the background changed little in two adjacent frames, adopting continuous image frames as poor and then establish initial background by using the method of threshold value judgment and N consecutive frames without moving target area between updates as the background. In view of the counting tracking algorithm, this paper analyzes the principle of Kalman filtering and prediction model and puts forward a kind counting tracking algorithm which is given priority to overlapping area and is complementary location and color histogram.About the FPGA platform construction, the paper also complete a kind of video traffic detection system based on FPGA.First of all, this paper analyzes the FPGA system design principles and design methods. Then the FPGA platform of video vehicle detection system is completed by these design principles and methods The system consists of image acquisition module, image data storage control module, image pretreatment and analysis detection module, VGA display module, etc, which solve these problems such as the video image collection, storage, and video analysis and provide good condition for system function expansion. Also studies how to use vehicle detection algorithm on the FPGA system implementation through the hardware description language. Finally experiment contrast analysis shows that this system has high detection accuracy. Due to the FPGA processing speed and high code execution efficiency, the system improves the detection speed effectively and has good real-time performance.