Research On Vehicle Bus Lane Monitoring System Based On Video

With the development of social economics and the general improvement of human livingstandards, there is a growing demand of convenient transportation environment. IntelligentTransportation Systems focus on the monitoring and management of the urban transportation.In order to solve the traffic jam and traffic pollution, many cities establish the bus lane. Busin lane driving is a high efficiency, low energy, low pollution and low cost form of publictransportation and it also effectively improve the efficiency of urban public transportation.However, there is a general phenomenon that non-public delivery vehicles often occupy thebus lane, which weakens the value of the bus lanes. Therefore, It is very important toresearch an effective monitoring system for ensure the value of the bus lanes.At present, the most of bus lane monitoring method often mounts fixed camera in theparticular lanes to capture the lane data in china. And this need supplemented by field ofartificial traffic police on duty to monitor the lane. By watched the original image onbackground, staff can monitor the violating vehicle. This method suffers from manydisadvantages, such as blind spots in the monitoring area, huge monitoring information andbig investment of human resources. And it did not meet the needs of the rapid developmentof urban intelligent transportation system.In order to solve the phenomenon of bus lanes are occupied, this paper preciselyresearches video-based vehicle bus lane monitoring system, and use digital image processingtechnology to complete the designed of bus lanes monitoring algorithm. And used TI’s DSPchip TMS320DM642as core processor design to build the vehicle bus lane monitoringsystem hardware and software platform, it was called embedded real-time monitoringsystem which to meet bus lane monitoring system requirements. At the same time, theseresulting data and images are sent over a network connection, and it provided a low-cost andhigh-efficiency solution for the vehicle bus lane monitoring systems and intelligenttransportation. The main contributions of this paper are:（1）By digital image processing technology, which include image preprocessing, binary,mathematical morphology filtering algorithm and so on, combined with the Houghtransform to complete the lane detection and lane extraction; （2）Detected the vehicle in the region of lane line (ROI), judged suspected vehicle in thelane according to the processed image. And completed vehicle classification and didviolation vehicle license plate location.（3）Sent the detected and extracted data to the monitoring host through the wirelessnetwork GPRS, Completed Real-time processing for the detection results and local storagefor the important result and video.（4）Completed hardware and software design for the bus lanes monitoring system andmade a PCB board around DM642.