A Point Cloud Registration Method Based on Differential Evolution Algorithm and TrimMedicp Algorithm

Abstract

The invention discloses a point cloud registration method based on a differential evolution algorithm and a TrimmedICP algorithm. The TrimmedICP algorithm can solve the problem of low overlapping rate in point cloud registration, but needs the setting of a proper initial value of a rotation matrix and a proper initial value of a translation matrix at first, otherwise it will fall into the local optimum; the differential evolution algorithm can generate an initial population through a random means and makes the population distributed in all ranges through using mutation, crossover and selection operation; and the differential evolution algorithm and the TrimmedICP algorithm are combined, and therefore, the problem of difficulty in selecting the initial values of the rotation matrix and the translation matrix can be solved, a globe optimal solution can be obtained, and thus, an accurate registration result can be obtained.

Description

technical field [0001] The invention belongs to the fields of computer graphics, surveying and mapping, reverse engineering and the like, and relates to a point cloud registration method based on a differential evolution algorithm and a TrimmedICP algorithm. Background technique [0002] In the field of 3D reconstruction of objects, there are two common ways to obtain the depth information of the target surface: one is to use a 3D laser scanner to scan in all directions, and the other is to use image matching to obtain depth information through binocular stereo vision. For the former, due to the complexity of the object, it is impossible to scan and obtain the overall point cloud data of the target at one time. For the latter, it is necessary to take binocular pictures from different angles to obtain point cloud information from different angles. Handle the problem of multi-site cloud registration. The problem of multi-site cloud registration is to align and merge the point...

AI Technical Summary

Problems solved by technology

The public point method requires at least 3 pairs of artificial marker points or feature points between the two point clouds, and the transformation matrix parameters can be obtained by the least square method, but this method needs to separately identify certain marker points and number the recording tape to a large data complexity; the ICP method iteratively assigns the corresponding relationship based on the nearest distance criterion, and obtains the least squares of the rigid transformation of the two point clouds, and then re-determines the corresponding relationship and continues to iterate until reaching the minimum value, This method is simple and has low computational complexity, but the premise of the ICP algorithm requires that the point cloud overlap rate is greater than 50%, and a good initial value must be set for it so that it will not fall into a local optimal solution

Images