Construction, compiling and reading methods of small-volume high-precision positioning layer

Abstract

The invention discloses construction, compiling and reading methods of a small-volume high-precision positioning layer, and the construction method comprises the steps: S1, respectively collecting three-dimensional marker information and lane line information contained in a region of the constructed positioning layer through a laser radar and a camera, and describing the information in a mode of feature points; S2, compressing the content described in the form of the feature points in the step S1 to form a binary data packet; s3, projecting the binary data packet to a vector map; and S4, comprehensively compiling the three-dimensional marker information and the lane line information described in the form of feature points with the vector map. A map format which is clear in specification and gives consideration to map data volume and reading efficiency is realized.

Description

technical field [0001] The invention relates to the technical field of high-precision maps applied to automatic driving, in particular to a method for constructing, compiling and reading a small-volume high-precision positioning layer. Background technique [0002] Autonomous driving refers to technologies that assist or replace human-driven cars. With the development of this technology, people's travel will be more convenient, and the influence of human factors in manual driving will be reduced, which can further improve driving safety to some extent. Among the technologies of autonomous driving, high-precision positioning is the most important, because it directly affects the input of other autonomous driving modules. Precise localization is a prerequisite for performing other autonomous driving functions such as perception and decision-making control. The positioning of autonomous driving relies on the combination of vision or lidar and high-definition maps, that is, co...

AI Technical Summary

Problems solved by technology

[0003] Existing high-precision maps for positioning are mainly formatted in two ways: a topological vector map, which is similar to the format of traditional navigation maps, with a small volume but not enough information for positioning , it is impossible to form a positioning method through feature point matching; the other is a panoramic point cloud or feature point layer. Although this kind of map directly describes the environmental information, it can realize a high-precision positioning algorithm through feature point matching. The amount is large, takes up a lot of storage space, and the retrieval speed is relatively limited