全文下载： 
20240503.pdf
文章编号： 1672-6987（2024）05-0135-07； DOI： 10.16351/j.1672-6987.2024.05.018
崔凤英， 李佩佩， 曹梦龙（青岛科技大学 自动化与电子工程学院， 山东 青岛 266061）
摘要： 针对经典统计滤波算法无法自适应选取参数以及传统双边滤波算法难以兼顾保特征和光顺性的问题，提出一种自适应组合滤波算法。首先引入基于局部点云体积的自适应标准差倍数以灵活滤除假肢点云大尺度噪声；在滤除大尺度噪声的基础上，引入一种新的协方差矩阵加权方式，提高估计点云法向的准确性，并通过法向夹角变化程度的均值对特征权重因子进行改进，增强双边滤波因子的保特征性，旨在光顺三维假肢模型小尺度噪声。与单独使用统计滤波、双边滤波相比，所提算法在3个假肢模型的最大误差Emax至少降低了5%；平均误差Eave至少降低了69%。仿真结果表明，该改进算法在有效剔除假肢模型大尺度噪声的同时又避免了过光顺和去噪不彻底，可以较好地保持模型中的几何特征。
关键词： 自适应策略； 滤波算法； 3D假肢； 模型重建
中图分类号： TP 391文献标志码： A
引用格式： 崔凤英， 李佩佩， 曹梦龙. 自适应组合滤波算法在3D假肢模型中的应用［J］. 青岛科技大学学报(自然科学版), 2024, 45(5): 135-141.
CUI Fengying, LI Peipei, CAO Menglong. Application of adaptive combined filtering algorithm in 3D artificial limb model［J］. Journal of Qingdao University of Science and Technology（Natural Science Edition), 2024， 45（5）： 135-141.
Application of Adaptive Combined Filtering Algorithm
in 3D Artificial Limb Model
CUI Fengying, LI Peipei, CAO Menglong
(College of Automation and Electronic Engineering, Qingdao University of Science and Technology, Qingdao 266061, China)
Abstract: Aiming at the problems that the classical statistical filtering algorithm can′t adaptively select parameters and the traditional bilateral filtering algorithm can′t consider both feature preservation and smoothness, an adaptive combined filtering algorithm is proposed. Firstly, the adaptive standard deviation multiple based on local point cloud volume is introduced to flexibly filter out the large-scale noise of prosthetic point cloud; On the basis of filtering out large-scale noise, a new covariance matrix weighting method is introduced to improve the accuracy of estimating the normal direction of point cloud, and the feature weight factor is improved by the mean value of the degree of change of the included angle in the normal direction, so as to enhance the conservation characteristics of the bilateral filter factor, aiming at smoothing the small-scale noise of the three-dimensional prosthetic model. Compared with statistical filtering and bilateral filtering alone, the maximum error of the proposed algorithm in the three prosthetic models is reduced by at least 5%; The average error is reduced by at least 69%. The simulation results show that the improved algorithm can effectively eliminate the large-scale noise of the prosthetic model while avoiding the over smoothness and incomplete noise removal, and can better maintain the geometric features in the model.
Key words: adaptive strategy; filtering algorithm; 3D prosthesis; model reconstruction
收稿日期： 2023-11-19
基金项目： 山东省研究生教育质量提升计划项目(SDYJD18029).
作者简介： 崔凤英(1972—)，女，副教授.
