Research on Information Acquisition System of Crossroads Based on Images |
|
Author | NiuYanMing |
Tutor | ShiZhongKe |
School | Northwestern Polytechnical University |
Course | Traffic Information Engineering \u0026 Control |
Keywords | Transport the amount of information Adaptive sensor control Background subtraction Image Segmentation Vehicle queue length |
CLC | U491 |
Type | Master's thesis |
Year | 2007 |
Downloads | 105 |
Quotes | 0 |
Real-time access to the the intersection traffic urban traffic adaptive sensor control the amount of information , has a very important significance to solve the traffic problems of the modern city . However, it is now widely used in traffic adaptive sensor control traffic information detection equipment both in the installation and maintenance aspects , or there is a big defect detection technology . Papers designed to solve this problem , a parallel image of a multi - phase traffic information acquisition , sharing processing detection system ; according to the characteristics of the vehicle queue image from the adaptability angle analysis and comparison of two typical approach in this on the basis of simulation tests to verify the effectiveness of the algorithm . Specifically, the main content of the dissertation : (1) detection for the intersection of information needs , system hardware DSP FPGA - based system architecture design of the multi-phase traffic information detection program ; combination of anti-jamming design of the high-frequency circuit , completed a board and debugging . (2) for the system functional requirements , the design of the basic module of DSPs , external memory and video decoder chip configuration program , based on multi- phase image acquisition and access need to be completed in the FPGA parallel acquisition and control logic and the synchronization signal VREF analog ; line , point to optimize the combination of acquisition mode , to solve a single dual-frame image data storage to the progressive storage, and the storage address and the read-write signal generation interlaced acquisition by optimizing the design of the acquisition and storage procedures . (3 ) completed by the simulation of a single traffic information acquisition adaptability discussion of dual - frame image processing algorithms to determine the processing algorithm based on background subtraction ; by using iterative threshold segmentation method and maximum between variance method (Otsu) strike threshold fleet image extraction , confirms the effectiveness of the algorithm ; design lanes on both sides of the background removed , horizontal projection and front rear boundary search procedures to ensure accurate extraction of the vehicle queue length . Finally, the design experience , further system upgrades outlook .