Method and device for recognizing lane lines, equipment and storage medium

Abstract

The embodiment of the invention provides a method and device for recognizing lane lines, equipment and a storage medium, and the method comprises the steps: obtaining a video, and determining attribute parameters of the lane lines according to video frame information in the video, wherein the attribute parameters of the lane lines comprise the type of the lane line, stability of the lane line andthe position of the lane line; determining attribute parameters of lane lines of historical frames and attribute parameters of lane lines of future frames according to the attribute parameters of thelane lines; and determining the attribute parameters of the lane lines of current frames according to the attribute parameters of the lane lines of the historical frames and the attribute parameters of the lane lines of the future frames. The invention relates to the technical field of image processing, and the attributes of the lane lines of the current frames is reasoned by using the time sequence attribute of the front and rear frames of the video and the output of the segmentation model, so that the judgment of the type of the lane lines is more accurate, the lane line output result is more stable, and sudden changes of types are reduced.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method, device, equipment and storage medium for recognizing lane lines. Background technique [0002] The detection and recognition of lane lines is an indispensable part of the fields of automatic driving and intelligent traffic violation capture. Accurate detection of lane lines can assist vehicles to automatically drive in legal lanes, and smart traffic violations such as real-line lane changes, etc., have very high requirements for accurate identification of lane line types. At present, the commonly used lane recognition method uses the semantic segmentation model to obtain the mask of the lane line, and determines the type of the lane line by the category of the pixel in the mask area. [0003] In the prior art, the number of pixels to be processed is reduced by extracting candidate feature points representing the centerline of the lane line. By analyzi...

AI Technical Summary

Problems solved by technology

[0004] At present, in the market, the reliability of the detection of lane lines is low, and irreparable errors often occur in the application.

In addition, in the real scene, the lane line will be blocked by obstacles, so the lane line information cannot be obtained accurately, and the lane line information is unstable during the recognition process.

Images