Autonomous mobile robot, map splicing method and device thereof and readable storage medium

Abstract

The invention discloses a map splicing method. The method comprises the following steps of: constructing a first sub-map and a second sub-map with an overlapping area based on laser data detected by amobile laser radar; acquiring a plurality of first poses of the laser radar in the construction process of the first sub-map and laser data corresponding to each first pose as first laser data, and acquiring a plurality of second poses of the laser radar in the construction process of the second sub-map and laser data corresponding to each second pose as second laser data; generating a pose relationship according to the plurality of first poses and the corresponding first laser data, and the plurality of second poses and the corresponding second laser data; and splicing the first sub-map andthe second sub-map according to the pose relationship. The invention further discloses a map splicing device, an autonomous mobile robot and a readable storage medium and aims to improve the accuracyand success rate of map splicing.

Description

Technical field [0001] The invention relates to the technical field of map construction, in particular to a map splicing method, a map splicing device, an autonomous mobile robot and a readable storage medium. Background technique [0002] At present, autonomous mobile robots using laser navigation are widely used. In the process of its application, in some scenarios, it is necessary to fuse two or more laser maps constructed into a larger laser map to ensure the operation requirements of the robot. [0003] When performing map splicing, some use brute force to search two raster maps to find the optimal solution as the conversion relationship between the two raster maps, but this method is not accurate and may not find the optimal solution. It is easy to cause mosaic failure; in addition, some extract the SURF features of two raster maps separately, find the conversion relationship between the two raster maps through feature matching, and then cannot extract effective features in ...

AI Technical Summary

Problems solved by technology

[0003] When doing map splicing, some use brute force to search two raster maps to find the optimal solution as the conversion relationship between the two raster maps, but the accuracy of this method is not high, and the optimal solution may not be found. It is easy to cause splicing failure; in addition, some SURF features of two raster maps are extracted separately, and the transformation relationship of the two raster maps is found through feature matching, and then effective features cannot be extracted in scenarios such as long corridors and open areas. , resulting in the inability to achieve map stitching

It can be seen that the current map stitching methods need to determine the pose relationship between the two maps based on the grid map, which has the problem of low accuracy and success rate.

Images