Point cloud automatic registration method based on a local feature descriptor

Abstract

The invention provides a point cloud automatic registration method based on a local feature descriptor. an initial matching point pair can be obtained based on a local fast point feature histogram descriptor and a sampling consistency algorithm; A coarse registration matrix is obtained through an error measurement loss function, and then a fine registration rigid body transformation matrix is obtained by combining an iteration nearest point algorithm, so that the problem that three-dimensional point clouds cannot be automatically registered under different visual angles can be effectively solved; Compared with an existing method, the method has the advantages that mark points do not need to be pasted, the method is not affected by a positioning device, an additional auxiliary device is notneeded, the requirement for the environment is not high, efficiency is high, and the method has high robustness in actual measurement.

Description

technical field [0001] The invention relates to the field of point cloud data registration, in particular to an automatic point cloud registration method based on local feature descriptors. Background technique [0002] With the development of three-dimensional measurement technology, surface structured light three-dimensional measurement method, as a non-contact optical measurement method with fast measurement speed and high precision, is widely used in industrial production detection parts, biomedical and historical relic protection, product quality Control, model digitalization and other fields. In the actual measurement process, due to the occlusion of the measured object itself and the environment, or some limitations in the measurement range, a single measurement can only obtain 3D data on a part of the surface, so it is necessary to measure more from different perspectives and poses. Measure the measured object once to obtain a complete 3D data model. Among them, th...

AI Technical Summary

Problems solved by technology

[0003] Although the predecessors have done more detailed research on the point cloud registration technology, the existing registration methods mainly include manual registration, registration based on landmark points, use of motion positioning devices to assist registration, and use of the topographical features of the measured object These methods have problems such as the need to manually select points, the need to paste marker points, the measurement range is limited by the movement range of the positioning device, the need for additional auxiliary devices, and high environmental requirements. At the same time, when performing point cloud registration, the measured object The shape features of different shapes, the point cloud data obtained under the same conditions are also quite different, and there may be wrong matches in the registration process.

Images