A method and device for extracting lane lines

Abstract

The invention discloses a lane line extraction method and device, belonging to the technical field of autonomous driving. The method includes: obtaining a color image for which lane lines need to be extracted; selecting a plurality of pixels on the color image at a predetermined interval of pixels, and taking each selected pixel as a center pixel, sequentially selecting an image area of ​​a predetermined size from the color image; obtaining each The Haar_like feature of each image area; use the preset classifier to classify the Haar_like feature of each image area respectively, and obtain the image area of ​​the lane line; the preset classifier is to train samples of predetermined size through the boosting training algorithm The cascade classifier obtained by the Haar_like feature of the image; according to the image area classified as lane line, extract the lane line on the color image that needs to extract the lane line. The device includes: a first acquisition module, a selection module, a second acquisition module, a classification module and an extraction module. The invention improves the stability and accuracy of the lane line extraction method.

Description

technical field [0001] The invention relates to the technical field of autonomous driving, in particular to a method and device for extracting lane lines. Background technique [0002] In recent years, autonomous driving technology has received widespread attention from automotive researchers. The extraction of lane lines is an important part of autonomous driving technology. The extraction of lane lines refers to the detection of lane lines and separation of the lane lines from the background from the road video image obtained by the vehicle camera. [0003] The existing lane line detection methods can be divided into three types: lane line area detection method, feature-driven method and model-driven method. The lane line area detection method is to identify the lane line according to the difference between the color and texture of the lane line and the color and texture of the background. The feature-driven method is to detect the lane line through the shape features o...

AI Technical Summary

Problems solved by technology

[0004] In the process of realizing the present invention, the inventors have found that the prior art has at least the following problems: the existing lane line detection method is mainly based on the inherent appearance characteristics of the lane line (including the true color and the inherent contour of the lane line), but in the vehicle During driving, the appearance of lane lines has many changes. For example, lighting and weather conditions change significantly at different times, resulting in different brightness and darkness of the collected lane line images; road conditions in different geographical environments are different, and lane lines may be blocked by vehicles or pedestrians. Occlusion, which affects the outline of lane lines

These changes degrade the stability and accuracy of existing lane detection methods

Images