Traffic Flow Detection Based On Video

Along with the growing popularity of automobiles , the road on use cannot keep up with the increasing number of cars. Then we should study on how to make the best use of the traffic network, which is also the aim of intelligent transportation system(ITS) based on video. ITS based on video can detect much traffic information, one of which is traffic flow.This article introduces the design of traffic flow detection system. First, it explains image pre-processing algorithm and image recognition algorithm about traffic flow detection. Second, after introducing traffic flow detection algorithm in existence, it brings forward an improved traffic flow detection algorithm based on edge information. Third, it simulates the algorithm on PC. Last, it considers system design and discusses several key problems of the design. This Paper is divided into four parts to discuss this task as below:In the first part, we introduce the development and application of ITS, Bring out a typical application of image recognition technology, Which is traffic detection. Then we introduce the system structure of simulating traffic flow detection algorithm.In the second part, we introduce image pre-processing algorithm, discuss the key problem of traffic flow detection. How to detect image moving information is an important problem. After comparing these algorithm, we put forward an improved traffic detection algorithm based on edge information. Relative to original algorithm, the improved algorithm have many renovations. In the third part, we discuss the total process of the simulation. That is to say, we use vidicon to record road status, then make it into VCD. In laboratory, we use image capture card V700 to capture traffic information output by VCD player, display and process the kinescope on PC, finish algorithm simulation. At the same time, we compare image smoothness algorithm, image enhancement algorithm and edge detection arithmetic operators, test detection veracity of the improved traffic flow detection algorithm based on edge information under all kinds of weather and road circumstances. The result shows this algorithm is better than frame difference method, background difference method and road mark method.In the fourth part, we discuss the design of traffic detection system, talk over some key questions. For example, the simulation of I2C bus with microchip unit to control video chip--SAA7110, The manipulation of flash chip --Am29LV200B, the control of the whole system with CPLD. In the end, we point out the limitation of this design and the way to improve. Certainly, we should continue to make research on this topic.