A Method of Part Pose Acquisition Based on Geometric Primitives

Abstract

The invention discloses a part pose acquisition method based on geometric primitives, which belongs to the field of visual positioning; the part pose information is usually obtained through the point cloud registration method, and the existing feature descriptor registration method has characteristics Point extraction takes a long time, and fine registration through ICP is required after rough registration to meet the accuracy requirements. However, the present invention proposes a method for obtaining the pose of a part by using geometric primitives. The method of the present invention extracts the geometric primitive features (plane, cylindrical surface) in the point cloud of the part, and constructs the local coordinate system and the local coordinate features according to the positional relationship between the primitives, and directly calculates the relationship between the point cloud and the template through the matching of the local coordinate features. The local coordinate system transforms the relationship to obtain the point cloud pose of the part; the present invention can obtain accurate pose accuracy without fine registration, and has high calculation efficiency.

Description

technical field [0001] The invention relates to a method for obtaining a pose of a feature part with a geometric primitive such as a plane cylinder, belonging to the technical field of visual positioning, and in particular to a method for obtaining a pose of a part based on a geometric primitive. Background technique [0002] In the automatic assembly process, the robot first needs to complete the grasping of the assembly parts autonomously, and the three-dimensional pose information of the parts is a necessary condition for the robot to complete the grasping. The three-dimensional pose is the relationship between the current pose of the part to be grasped and the pose of the template. On the basis of using the 3D scanning device to obtain the point cloud data of the part to be captured, the rotation and translation matrix (R, T) between the part point cloud and the template point cloud are registered to obtain the two, and the obtained RT is the part's pose. [0003] Exis...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art, and propose a method for obtaining the pose of a part based on geometric primitives, so as to solve the problems of large amount of calculation and low efficiency of the existing three-dimensional pose calculation method

Images