Lane Detection Under Complicated Environment

Vision-based lane line detection is one of the key technologies in intelligent driving systems.Due to the diversity and complexity of the actual road environment in real life,although many researchers have done a lot of research in this field,there are still many disturbances in achieving accurate lane line detection,such as lighting transformation,road shadow,vehicle driving traces and other traffic sign.Aiming at these problems,this article aims to achieve more accurate lane line detection in a complex driving environment,and conducts research on structured road lane line detection and unstructured road road boundary detection.Its main research work is as follows:1.Aiming at the problem that structured road lane line detection is susceptible to interference of road surface conditions and surrounding environment in complex driving environments,a method based on probability voting is proposed for estimating the location of vanishing points,and then on the basis of the lane line,ideally,passing the road vanishing point,a lane line extraction method based on the road vanishing point constraint is also proposed to realize lane line detection on various complicated driving roads.2.Aiming at the problem that the lane line detection algorithm may fail in a complex driving environment,a lane line tracking algorithm based on the inter-frame lane line relationship of continuous frame images is proposed,on the basis of a lane line detection results,this paper implement dynamically determination of regions of interest,this improve the real-time and robustness of the algorithm.3.Aiming at the problem that road area segmentation is easily affected by interference factor on the road surface in unstructured roads,this paper proposes an image segmentation method based on the color features of road images,on the basis of the one-time segmentation,an improved region growth algorithm is also proposed to extract road regions accurately.4.Aiming at the problem that road edge points cannot be extracted easily and accurately in unstructured roads,a road edge point extraction method based on image block algorithm is proposed,after the road edge points are obtained,a cubic polynomial curve will be fitted by using the least square method,and this will obtain an accurate boundaries of unstructured road.In this paper,experimental analysis of the proposed algorithm is performed on datasets of various driving environments.The experimental results show that the two algorithms proposed in this paper can effectively detect lane lines and road boundaries,respectively,under various complicated driving environments.