Automatic lane line extraction method and system based on region of interest

Abstract

The invention discloses a lane line automatic extraction method and system based on a region of interest, and the method comprises the steps: light incoming amount control: collecting an image, and adjusting an adjustment parameter according to the illumination scene of the image and a mapping relation obtained through the pre-illumination applicability training; foreign matter detection and processing: selecting a region of interest from the image, establishing an energy function and performing loop iteration, solving to obtain pixel combination of the foreign matter, and zeroing pixels belonging to the foreign matter to obtain the region of interest after the foreign matter is removed; and lane line extraction: carrying out graying reading on the region of interest after foreign matter removal, carrying out dichotomy according to gray similarity, setting loss function convergence and termination conditions, stopping iteration after the termination conditions are met, completing pixel distinguishing of the lane line and a non-lane line, and finally extracting the lane line. According to the invention, the foreign matters in the lane line area are detected and the corresponding pixels are removed by controlling the light incoming amount, and the method has the effects of high recognition response speed and accurate recognition result.

Description

technical field [0001] The invention relates to the technical field of automatic driving environment perception, in particular to a method and system for automatically extracting lane lines based on an area of ​​interest. Background technique [0002] The automatic driving system is divided into: perception system and decision system. Among them, the perception system is composed of vehicle positioning, moving obstacle detection, traffic sign detection, and lane line extraction. Lane line extraction using visual sensors is mainly divided into deep learning methods and traditional image processing methods, but both have obvious shortcomings. In the process of extracting lane lines using deep learning methods, a large amount of prior information needs to be collected. Moreover, deep learning methods generally have high time complexity and consume more storage space. Delay requirements are violated. Although the traditional method does not require a large amount of prior inf...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that the two lane line extraction methods in the prior art have obvious shortcomings and cannot be well adapted to the application scenarios of automatic driving, the present invention provides a method and system for automatically extracting lane lines based on regions of interest. area, train the parameters of gray-level nonlinear transformation under different lighting scenarios, control the "light input" of the self-driving vehicle under different lighting conditions, detect foreign objects within the range of the lane line, and remove foreign objects within the range of the lane line. Avoid the impact on the lane line extraction, and then classify each pixel, so that the pixels belonging to the lane line and the non-lane line pixels can be accurately distinguished

Images