Research And Implementation Of Vehicle Lane Change Detection Algorithm

Vehicle assisted driving is an important field of research in the global automotive industry,and lane line recognition and vehicle lane change detection and early warning are the key technologies.This technology can reduce traffic accidents caused by drivers changing lanes without reason,and has very important research significance and value.The thesis conducts in-depth research on vehicle lane change detection and its early warning technology,aiming to integrate the vehicle lane change detection algorithm into ordinary vehicle-mounted auxiliary equipment and achieve excellent practical application.The main research contents of the thesis are as follows:1.Aiming at the uncertainty of the installation angle of the camera of the vehicle auxiliary equipment,a method of setting the region of interest based on the vanishing point of the road is proposed to accurately locate the lane line area of the road image.First,a series of preprocessing operations are performed on the image,and the straight line segments are extracted from the whole road image,then the extracted straight line segments are filtered,and the weighted least squares algorithm is used for vanishing point fitting.Finally,the region of interest is set according to the determined vanishing point position.2.In order to solve the problem that the traditional algorithm has low accuracy of lane line recognition in complex lane line scenes,a lane line detection algorithm combining improved Hough transform and clustering algorithm is proposed.The line segment detected in the image is reclassified,and the inner edge lines of the two lane lines of the current lane are precisely located and used to represent the left and right lane lines.Experiments show that the overall accuracy of the algorithm in various scenarios is about 95%,and the average processing time per frame is 25.6ms.3.According to the characteristics of on-board auxiliary equipment,a vehicle lane change judgment model with multiple constraints is proposed.Firstly,the advantages and disadvantages of common models are analyzed,and it is concluded that these models are not suitable for ordinary on-board auxiliary equipment.Therefore,a lane change judgment model based on the combination of vehicle yaw angle and its change rate and the relative position of lane line in the image is proposed to ensure the real-time and accuracy of the system.Finally,the vehicle lane change detection algorithm is integrated into the vehicle auxiliary equipment to realize a complete vehicle lane change warning system,and multiple groups of real vehicle test experiments are carried out for the system.The experimental results show that the system has an accuracy of 95.5% and an average false alarm rate of 1.8% for vehicle lane change warning under the conditions of standard structured straight roads and vehicle speeds above 40km/h,and it processes about 26 frames of images per second on average.The system basically meets the needs of actual use in terms of accuracy and real-time performance.