Three-dimensional point cloud high-precision registration method based on improved PPF algorithm

Abstract

The invention relates to the technical field of three-dimensional point cloud registration, in particular to a three-dimensional point cloud high-precision registration method based on an improved PPF algorithm, and the method comprises the steps: obtaining a three-dimensional scanning point cloud and a three-dimensional standard point cloud, obtaining a three-dimensional scanning key point cloud and a three-dimensional standard key point cloud, obtaining a matching point pair of the three-dimensional scanning key point cloud and the three-dimensional standard key point cloud, and obtaining a matching point pair of the three-dimensional scanning key point cloud and the three-dimensional standard key point cloud; and solving coarse registration poses of the real object and the three-dimensional model, and carrying out registration on the three-dimensional scanning point cloud and the three-dimensional standard point cloud. According to the method, key point extraction is carried out on the preprocessed point cloud by adopting a key extraction strategy fusing an ISS algorithm and a 3D Harris corner detection algorithm, so that the extracted key points are more accurate, and on the basis of completing pose estimation and realizing coarse registration by using a weighted Hough voting method, a coarse registration result is optimized by using an ICP algorithm, so that the registration precision is further improved.

Images