Disparity map-based road detection method and device

Abstract

The invention provides a disparity map-based road detection method and device, and relates to the technical field of image processing. The method comprises the following steps of: determining a sparsedisparity map and a V disparity map of a to-be-detected image; obtaining the number of effective pixel points in a column of the V disparity map and a row range corresponding to the effective pixel points according to a preset pixel threshold value, and determining a row to which a road surface in the V disparity map belongs; and determining a road surface parallax point in the sparse disparity map according to the row to which the road surface in the V disparity map belongs, wherein a parallax value of the road surface parallax point is a parallax value corresponding to the column. By applying the method, the correctness and efficiency for determining road surface parallax points in disparity map images can be improved; and the method is suitable for scenes with undulating practical roadsurfaces.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular, to a method and device for road detection based on a disparity map. Background technique [0002] In recent years, the obstacle detection method based on Binocular Stereo Vision technology has become a research hotspot in the field of assisted driving. Before detecting obstacles based on parallax images, road parallax points are accurately detected and the road parallax points are deleted. It is especially important for obstacle detection results. In the prior art, a sparse disparity map is first generated based on the road image collected by the binocular camera, then a V disparity map is generated based on the sparse disparity map, and one or more straight lines are detected in the V disparity map by using the Hough line detection method. To fit the straight line equation of the road surface in the V disparity map, then, find the disparity points that sati...

AI Technical Summary

Problems solved by technology

[0003] However, the above method of deleting the road surface is only suitable for ideally flat roads, and the roads in actual scenarios are likely to be undulating, such as from flat roads to downhill roads, from uphill roads to flat roads, and even other more complex roads. For this type of road conditions, it is difficult to use a straight line for fitting because the transition zone of different road sections usually manifests as a circular arc transition. Therefore, if a Hough straight line is used to detect the parallax point of the road surface, it is easy to cause The obstacle parallax is deleted by mistake in the road section, or the distant road surface parallax cannot be deleted, resulting in inaccurate obstacle detection results in the later stage

[0004] Based on this, it is proposed in the prior art to segment the road according to the distance and the undulations of the road surface, and each section corresponds to a Hough straight line to detect the road surface parallax point on the section of the road. However, the more times the Hough line fits, the longer it will take, resulting in the consumption of computing resources; at the same time, due to the complex and unpredictable road conditions in practical applications, it is inevitable to appear in the Undulations appear on a pre-divided section of the road; in addition, in the case of less parallax on the road surface, if the road is divided into multiple sections for detection, it may not be possible to fit the straight line equation of the road surface

It can be seen that the existing technology is not suitable for the actual scene of uneven roads, and it is easy to cause the accuracy of determining the parallax of the road surface to be low, which will affect the result of obstacle detection in the later stage.

Images