Research And Design Of Embedded Traffic Signal Control System Based On Image Processing Algorithm

In the field of intelligent transportation system, it is an important task to adjust the traffic lights display time automatically according to traffic jam. But because of the complexity of the road scene detection and the influence of vehicle’s inter-occlusion, shadows, road waterlogging or circadian conversion factors, the design of intelligent traffic system based on the embedded equipment will become a significant practical problem through the way of obtaining vehicle amount of queueing on road by image processing. The intelligent transportation system based on embedded system has incomparable advantages on function extension, system volume reduction, power consumption and cost reduction, real-time and stability etc. So the research on embedded intelligent transportation system based on digital image processing algorithms is of great practical significance.The subject "research and design of embedded traffic signal control system based on image processing algorithm" aims to design an embedded transportation signal control system composed by embedded S3C2440 chip and related peripheral hardware. The vehicles queuing images taken from a camera are processed through special algorithm to obtain some queuing contour features information and the number of vehicles, which will be as the basis for judgments of controlling embedded traffic system. The main work of this thesis is about realizing Linux3.10.42 kernel compile and transplantation, the file system, the compilation and transplantation of the driver, the realization of the software layer interface function and the main function. The system could adjust the traffic lights display time depending on the result of the image processing. So it is possible to regulate traffic order dynamically and efficiently by traffic signal for improving the traffic efficiency.