End-to-end-based lane line detection method, device and equipment and storage medium

Abstract

The embodiment of the invention relates to the field of artificial intelligence, and discloses an end-to-end-based lane line detection method, device and equipment and a storage medium, and the method comprises the steps: obtaining a training sample set, and adding a lane line label to each sample image in the training sample set; inputting the sample image added with the lane line label into a specified depth polynomial regression model to obtain a prediction result; adjusting model parameters in a specified depth polynomial regression model according to a prediction result, and retraining the specified depth polynomial regression model to obtain a lane line detection model; and obtaining a to-be-detected image shot by a foresight shooting device of the target vehicle, and inputting the to-be-detected image into the lane line detection model to obtain a lane line of the target vehicle. In this way, the lane line can be determined more accurately and effectively, and the efficiency and the real-time performance of lane line detection are improved. The invention relates to a block chain technology, for example, image data can be written into a block chain so as to be used for scenes such as data evidence obtaining.

Description

technical field [0001] The present invention relates to the field of artificial intelligence, in particular to an end-to-end lane line detection method, device, equipment and storage medium. Background technique [0002] Autonomous driving is a challenging research field that has attracted much attention in recent years. Perception problems related to this field have been greatly influenced by deep learning. Autonomous vehicles should have the ability to estimate lane lines, since each lane provides specific visual cues to control driving, in addition to serving as a spatial constraint. In this case, the two most important lines of traffic (i.e., lane lines) are the lines that define the lanes of vehicles. These lane lines set limits on the driver's actions, and their type determines whether manoeuvring the vehicle (such as changing lanes) is allowed. In addition, detection of adjacent lanes is very useful, and a comprehensive understanding of the traffic scene can lead t...

AI Technical Summary

Problems solved by technology

Most of the research is to extract features through the model based on the idea of ​​segmentation, which is inefficient and poor in real-time for autonomous driving

Furthermore, since the segmentation map must be post-processed to output traffic lines, the segmentation step alone is not sufficient to provide lane marking estimates

Furthermore, this approach may ignore global information, which is especially important when visual cues are missing, such as in strong shadows and occlusions

Images