Lane Recognition And Lane Departure Warning System Based On Edline Algorithm

In recent years,although the number of motor vehicle traffic accidents in China is on the decline,the total amount is still very large.So at present,people need the assistant driving system which can improve the safety performance of vehicles.There are many causes of traffic accidents,among which the driver's unintentional lane departure ranks the fourth.Therefore,the research on lane departure warning system has become the focus of the research on safe driving assistance.With the continuous development of safe driving assistance technology,people's requirements for the warning system are also increasing.In this paper,early warning decision algorithm in lane departure early warning system are studied and implemented.The main research is as follows:(1)Firstly,for image preprocessing,taking the road environment of our country as the research object,through the experimental comparative analysis and the improvement of the existing algorithm,we get the optimal algorithm flow: weighted average method for image gray-scale processing;an improved adaptive Gauss filter algorithm for image smoothing;the maximum class difference method for image binarization;finally,the static division method is used to select the region of interest.These methods can effectively reduce the interference information in the image,and provide better input for the following lane line feature extraction.(2)Then,a novel lane recognition algorithm based on Edline features is used to recognize lane lines.The lane lines are extracted by edge detection and line segment detection,as well as filtering and clustering the detected line segments.Then Kalman filter algorithm is used to track and predict lane line,and the recognition effect of this algorithm under different road conditions is verified by experiments.(3)Finally,in order to realize the classified early warning of the road area,the road is divided into four areas,and longitudinal TLC algorithm is used to establish the early warning model.Through simulation and real vehicle test.The results show that the lane line detection based on Edline algorithm has good detection performance,and it is appliedto lane departure warning system.The validity of lane departure warning decision-making method is proved by multiple lane departure tests.