Method and device for lane line detection

Abstract

This specification discloses a method and device for lane line detection. It obtains road images collected by unmanned driving equipment, and determines the top-view image corresponding to the road image according to the road image. Secondly, the road image and the top-view image are input to the pre-trained In the lane line recognition model, the recognition result of the lane line contained in the road image is obtained with reference to the feature of the lane line contained in the overhead image. Finally, according to the recognition result, lane line detection is performed. This method refers to the lane line features contained in the top view image. In the process of determining the recognition result of the lane line contained in the road image, the lane line feature that the lane lines are parallel in the top view image is fused, so that the identified lane line recognition The result is more accurate, which effectively improves the accuracy of lane line detection.

Description

technical field [0001] This specification relates to the field of unmanned driving technology, and in particular to a method and device for lane line detection. Background technique [0002] At present, in the technical field of unmanned driving, the images collected by the unmanned equipment are usually used to determine the lane line information of the current road during the driving process of the unmanned equipment, and then through the determined lane line information, the unmanned Human-driven equipment performs path planning and environment perception. [0003] In the prior art, the images collected by the unmanned equipment are often the images taken by the camera in the horizontal direction. The lane lines in this image generally reflect the convergence from far to near, while the lane lines in reality It is parallel, and the image captured by the camera does not reflect this feature. Therefore, if the lane line is recognized only based on the image data, the final...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, the images collected by the unmanned equipment are often the images taken by the camera in the horizontal direction. The lane lines in this image generally reflect the convergence from far to near, while the lane lines in reality are parallel, and the image captured by the camera does not reflect this feature. Therefore, the lane line is recognized only based on the image data, and the final recognition result is often inaccurate.

Images