Research On Information Acquisition System Of Crossroads Based On Images

One of the primary methods to solve the urban traffic problems is to apply adaptive sensing control with the real time sampling information of the crossroads. However, there exist great drawbacks of the information detecting facilities in the aspect of either maintenance or detecting technology that are widely used in adaptive sensing controlling. Multi-phase information detecting and time-sharing methods based on image processing are presented and researched. Moreover, according to the image features of traffic queue, two classic processing methods are compared and analyzed on the view of adaptability, which proved the validity of the algorithms.The main work of the thesis is as follows:(1) Considering the demand of the crossroad information detection, present an implementation scheme of multi-channel parallel distributive processing based on DSP and FPGA. With the anti-jamming design of radio circuit, implement the debugging of the hardware board.(2) Considering the demand for the system functions, design the basic module, external memory and the configuration program of the video decoding chip. Implement the parallel acquiring control logic and synchronizing signal VREF simulation based on the need for multi-phase image acquiring and storage. The acquiring method with line-point optimization is realized with the design of optimizing acquiring and storage program, which solves the problems of one-way, double frames image storage, interleaving acquiring while line-by-line storing, and generating storing address and reading-writing signals.(3) Discussing the adaptability of the single and double frames image processing method with the simulation experiments. Thresholds are acquired by iterative threshold segmentation and OTSU method applied on the target images based on frame difference. Meanwhile, measures are taken to remove the background at the sides of the road. Horizontal projection and boundary searching methods realize the accurate attraction of traffic queue.Finally, the upgrade of the system is forecasted with the help of the experience of this design.