Adaptive hybrid harmony search optimization algorithm for point cloud fine registration

The fine registration of point clouds is a key step in point cloud pre-processing. However, the point cloud fine registration algorithm still has problems, such as low iteration accuracy and slower iteration speed. To further improve the efficiency of the point cloud fine registration algorithm, a point cloud fine registration method based on an adaptive hybrid harmony search algorithm was designed. First, the fine registration mathematical model of the point cloud was established according to the iterative closest point algorithm. Second, regarding the basic harmony search algorithm, the disadvantages were described through algorithm mechanism analysis, and four strategies—cube chaotic mapping, the shuffled frog leaping algorithm, adaptive parameters, and quadratic interpolation—were introduced to improve the algorithm’s computational performance. Finally, the adaptive hybrid harmony search algorithm was validated by benchmark functions and point cloud registration data. Another five traditional algorithms were used for comparative analysis with adaptive hybrid harmony search. The results of the simulation experiment show the effectiveness of the adaptive hybrid harmony search algorithm in point cloud fine registration.