基于OPTICS聚类和改进双边滤波的点云去噪方法  

Point Cloud Denoising Based on OPTICS Clustering and Improved Bilateral Filtering

作　　者：钟志鹏 张建州[1] 梁彪 ZHONG Zhi-peng;ZHANG Jian-zhou;LIANG Biao(College of Computer Science,Sichuan University,Chengdu 610065;Welding Plant,Dongfang Electric Machinery Co.,Ltd.,Deyang 618000)

机构地区：[1]四川大学计算机学院,成都610065 [2]东方电机有限公司焊接分厂,德阳618000

出　　处：《现代计算机》2020年第10期76-80,共5页Modern Computer

基　　金：智能机器人视觉产学研联合实验室(No.17H0938);四川大学德阳校市科技合作专项基金资助项目(No.HZYF201501)。

摘　　要：点云去噪是点云预处理中十分重要的一个步骤,经过预处理的点云可用于后续的点云配准、重建等工作。通过三维扫描仪获取的点云往往含有噪声,为了有效地去除这些噪声,提出了基于OPTICS聚类和改进双边滤波的点云去噪算法,首先利用基于OPTICS聚类的去噪算法去除大尺度噪声,然后再利用改进的双边滤波算法去除小尺度噪声。实验结果证明,提出的方法能有效地去除噪声且能保持点云模型特征,相比于常用的统计滤波、半径滤波和二者结合滤波,去噪效率均有提升。Point cloud denoising is an important step in point cloud preprocessing.The point cloud after preprocessing can be used for point cloud reg istration,reconstruction and other work.Point clouds acquired by three-dimensional scanners often contain noise.In order to remove these noises,a point cloud denoising algorithm based on OPTICS clustering and improved bilateral filtering is proposed.Firstly,the denoising al gorithm based on OPTICS clustering is used to remove large-scale noise,and then the improved bilateral filtering algorithm is used to re move small-scale noise.The experimental results show that the proposed method can effectively remove noise and preserve the feature of point cloud model.Compared with the commonly used statistical filtering and radius filtering,the denoising efficiency is improved.

关 键 词：点云去噪 OPTICS聚类 双边滤波 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN713[自动化与计算机技术—计算机科学与技术]