A method and system for normal scan registration based on combined clustering

Abstract

The invention discloses a normal scan registration method based on combined clustering, which includes: obtaining two continuous laser radar point cloud data from a mobile robot in real time, the former laser radar point cloud data as a reference frame, and the latter laser radar point cloud data The point cloud data is used as the current frame, and the reference frame is processed using a density-based spatial clustering algorithm to obtain the initial classification result of the reference frame. For each classification in the initial classification result of the reference frame, determine its two largest singular values Whether the ratio of is greater than the set threshold, if so, use the K-means algorithm to process the classification to obtain K subdivisions, and add these K subdivisions to the final classification set of the reference frame, for the current frame for each lidar point. The invention can solve the technical problem that the existing normal scan registration method adopts grid division, which causes errors in the registration results, and further leads to the poor precision of the finally obtained environment map.

Description

technical field [0001] The invention belongs to the technical field of mobile robots, and more particularly relates to a method and system for normal scan registration based on combined clustering. Background technique [0002] In an unknown environment, simultaneous positioning and map construction are the cornerstones for mobile robots to achieve autonomy, intelligence, and complete more complex tasks. Scanning registration is a prerequisite for the realization of these two functions. [0003] At present, the most common scan registration method is the Normal-distributions transform (NDT) method, which obtains a series of point cloud data in the environment through perceptual sensors (such as lidar and depth cameras). The registration relationship between the two point cloud data is optimized to solve the coordinate transformation relationship between frames, and finally the environment map of the mobile robot's operating area is obtained. [0004] However, for the above-...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the above defects or improvement needs of the prior art, the present invention provides a method and system for normal scan registration based on combined clustering. There are errors in the registration results, which leads to the technical problem of poor accuracy of the final environmental map

Images