A Method of Real Time and Fast Lane Line Detection

In order to effectively realize the feature extraction of road image, reduce the complexity of calculation and the complexity of algorithm and improve the detection accuracy of lane mark, combined with the characteristics of lane line, in this paper, an effective image preprocessing scheme was selected and designed, and a detection method based on segmented ROI model which included close region, sub-close region and respective region was proposed. In the different segmentation region, the SHT method was used to detect and carry out the segmentation method to detect the lane line. The experiment results showed that the method could improve the detection accuracy of the lane line, greatly reduce the amount of data processed and improve the efficiency of the detection time. The real-time performance for the lane line detection could be improved obviously by segmented ROI model.