基于法矢量修正和参数修正的点云双边滤波改进算法

Improved Bilateral Filtering Algorithm for Point Cloud Basedon Normal Vector Correction and Parameter Correction

摘要:

针对现有点云滤波算法在实现过程中存在的目标特征保留能力差和时间复杂度高的问题,提出了基于法矢量修正和参数修正的点云双边滤波改进算法。介绍了点云双边滤波算法原理,在该算法的基础之上分别修正了法矢量、距离权重因子和特征保持因子,提出了一种改进的点云双边滤波算法,并通过仿真分析进行了算法验证。与传统的点云半径滤波算法和点云双边滤波算法相比,所提算法能够在时间复杂度较低的情况下保留更多的目标细节特征,提高滤波算法实时性。

Abstract:

Aiming at the problems of poor feature retention and high time complexity inthe implementation of existing filtering algorithms for point cloud, an improved bilateralfiltering algorithm for point cloud based on normal vector correction and parametercorrection was proposed. The principle of bilateral filtering algorithm for point cloud wasintroduced. On the basis of the algorithm, the normal vector, distance weight factor andfeature retention factor were modified respectively, and an improved filtering algorithm forpoint cloud was proposed. Through simulation analysis, the proposed algorithm is verified.Compared with the traditional radius filtering algorithm and bilateral filtering algorithm forpoint cloud, the proposed algorithm can retain more target details in the case of low timecomplexity, and improve the real-time performance of filtering algorithm.