Intelligent Traffic Light Design Based On Sight

Nowadays, with the development of industrial automation and automotive industries, the most cities of our country have entered car-oriented society in nearly a dozen years. Although the road is improving day by day, it also cannot solve the underlying problem. The problem of traffic environment is going from bad to worse. Traffic breakdown and urban traffic congestion occur more and more frequently, causing a direct result of poor operating condition of the car, low traffic limitations, energy waste, excess of automobile exhaust gas emission and other transportation issues, and bring great losses to the economy. In order to solve this problem, this article proposes an intelligent traffic light based on Sight that can intelligent control the traffic to improve the efficiency of the scheduling of the traffic lights.This article described from the two aspects of hardware architecture and software design for intelligent traffic lights in detail. The core of the Vision-based intelligent traffic’s hardware platform is composed of FPGA and DSP. Function modules of the system can be divided into image collection module, image processing module and the power supply module. In the image collection module, through the image decoder chip ADV7181B decodes the CCD image sensors’ analog video into standard digital video signal CCIR656YCrCb4:2:2. FPGA is the main controller for image collection module, because of its advantages of interface design. At first, FPGA complete the configuration of ADV7181B, and then capture an image from the video stream and the use of ping-pong operation stored in the SRAM. DSP is the core of image processing module. DSP accesses real-time images from image collection module by its EMIF interface. Then, filter the image noise, extraction and update background, get out the image of the moving vehicle prospects using the background difference method. DSP uses threshold to do image binaryzation. Using the improving area weighted method to get traffic flow information, including the number of vehicles, and the presence or absence of vehicle. Comprehensive every approaches’traffic flow information of the intersection to optimal timing of traffic lights. Power supply module provides stable and reliable in a variety levels of voltage to system. Finally, this article to verification of functions based the physical model of the intelligent traffic light. Test results showed that the system can get the correct traffic information through image processing. Under the typical conditions, intelligent traffic light system can put out reasonable traffic light timing.Intelligent traffic light based on Sight can optimize the timing of traffic lights, reducing no reason stay of vehicle at traffic junction, improving movement efficiency in junction. And the intelligent traffic light has many significant advantages that miniaturization, easy to install, commission and update, and low cost so that it can be promoted easily, has good market prospects.