Research On Algorithm Of Lane Detection And Departure Warning Of Intelligent Vehicle Based On Vision

The rapid development of economy and society has brought great changes to people’s life, and people’s demand for cars is also growing. In recent years, with rapid growth of the number of cars, traffic accidents happen frequently, and then the study of Intelligent Vehicle Safety Driving Assistant Technology, which can avoid traffic accidents and provide safe driving, has become a hot research topic.Firstly, in order to eliminate the irrelevant information and enhance the testability of useful information, the driveway image sequences collected by camera were preprocessed. Then a division of the preprocessed image was made to distinguish the Lane area and an image only contains driveway mark line was obtained. Secondly, the driveway mark line is refined to reduce the amount of data. The principles of extraction straight line are analyzed by the least squares and Hough transform. The optimal algorithm of extraction is then selected by experiments in this dissertation. At last, by introducing the principles of the most common current model of lane departure warning and analyzing its advantages and disadvantages, the driveway mark line equation and slope parameters are used to confirm the position and direction of the vehicles in the driveway, and a kind of lane departure warning model without calibration is established.This paper has studied the algorithm of Lane recognition and detection, feature extraction of lane markings and lane departure warning system. The simulation results show that the algorithm is feasible and effective, and can achieve real-time requirements of embedded platforms.