Point cloud registration method and device based on multi-feature fusion and storage medium

Abstract

The invention discloses a point cloud registration method and device based on multi-feature fusion and a storage medium. The method comprises the following steps: respectively extracting a plurality of source point cloud feature points and a plurality of target point cloud feature points from a source point cloud and a target point cloud; extracting a local depth feature, a normal angle feature, apoint cloud density feature and a local color feature of each feature point, and then generating a feature descriptor corresponding to each feature point according to the local depth feature, the normal angle feature, the point cloud density feature and the local color feature of each feature point, wherein the feature points comprise source point cloud feature points and target point cloud feature points; pairing the source point cloud feature points with the target point cloud feature points according to the feature descriptors to generate feature point pairs; and generating a transformation matrix according to the feature point pairs, transforming the source point cloud according to the transformation matrix to generate a second source point cloud, and performing fine registration on the second source point cloud and the target point cloud. By implementing the embodiment of the invention, the expression capability of the feature descriptor can be improved, so that the overall registration precision is improved.

Description

technical field [0001] The present invention relates to the technical field of three-dimensional point cloud registration, in particular to a method, device and computer-readable storage medium for point cloud registration based on multi-feature fusion. Background technique [0002] The use of 3D scanning equipment to obtain object surface data is a hot issue that has developed rapidly in recent years. The most important technology is the processing of point cloud data. Due to the limitation of the scanning range of the equipment itself, in order to obtain high-precision omnidirectional point cloud data, It is necessary to register the point cloud data obtained from different perspectives. Convert the multi-view acquisition data into the same coordinate system to obtain the complete surface information of the object, and then realize the surface reconstruction of the object and target recognition and other applications. [0003] The iterative closest point algorithm and its...

AI Technical Summary

Problems solved by technology

[0005] The above-mentioned technology has a single feature extraction method for point clouds, uses global features as the description of feature points, has poor robustness, insufficient description of feature points, and cannot fully characterize the attributes of feature points, so it is easy to generate a large number of mismatches, which in turn affects Overall Registration Accuracy

Images