Lane line identification method and system

Abstract

The invention discloses a lane line identification method and system. The method comprises the steps of obtaining a lane line information graph; adopting a deep learning method to classify each pixelin the lane line information graph, wherein the types of the pixels comprise roadsides, yellow solid lines, yellow dotted lines, white solid lines and white dotted lines; and according to the classified images, extracting a lane line model to realize lane line identification. According to the invention, the lane line categories can be identified, including five categories of roadside, yellow solidline, yellow dotted line, white solid line and white dotted line, so that the information required by vehicle keeping, lane changing and accurate positioning can be provided, and the stable and safeoperation of vehicles can be realized. Similarly, on the basis of having enough labeled samples, the method can be extended to more categories of lane line recognition by increasing the categories ofnetwork output.

Description

technical field [0001] The invention relates to the technical field of unmanned driving and vehicle-mounted assisted driving, in particular to a lane line recognition method and system. Background technique [0002] With the rapid development of the economy and the continuous improvement of people's living standards, the number of automobiles in my country has increased rapidly. This has led to a high incidence of road traffic accidents, and traffic safety issues have become increasingly prominent. Vehicle safety driving assistance system has become one of the hot issues in the field of traffic safety research in the world. In the past few decades, experts and scholars at home and abroad have made a lot of contributions to the research of intelligent driving assistance systems. At this stage, relatively mature advanced driver assistance systems (ADAS) have appeared abroad and have been put into commercial applications. Compared with foreign technologies, domestic research...

AI Technical Summary

Problems solved by technology

However, due to the drawbacks in the network design, the image segmentation effect of this method is not ideal, which increases the error rate of lane line extraction in post-processing

Moreover, IPM mapping requires more precise calibration of camera parameters, which limits the generalization and promotion of the lane line recognition algorithm in actual safe driving.

In addition, the use of the DBSAN clustering algorithm will also reduce the running speed of the entire algorithm, which cannot meet the real-time requirements of lane line recognition on complex roads

More importantly, this method does not identify the roadside, which is difficult to meet the actual application requirements of the safety assisted driving system

Images