计算机应用 ›› 2013, Vol. 33 ›› Issue (06): 1662-1681.DOI: 10.3724/SP.J.1087.2013.01662

• 多媒体技术 • 上一篇    下一篇

散乱点云数据曲率估计方法

张帆1,康宝生1,赵建东1,李娟2   

  1. 1. 西北大学 信息科学与技术学院,西安 710127
    2. 陕西省科技资源统筹中心 共性技术推广部,西安 710061
  • 收稿日期:2013-01-29 修回日期:2013-03-04 出版日期:2013-06-01 发布日期:2013-06-05
  • 通讯作者: 张帆
  • 作者简介:张帆(1977-),男,陕西西安人,高级工程师,博士研究生,主要研究方向:计算机辅助几何设计、计算机图形学;康宝生(1961-),男,陕西咸阳人,教授,博士生导师,主要研究方向: 计算机图形学、图像处理;赵建东(1984-),男,陕西延安人,硕士,主要研究方向: 计算机图形学、图像处理;李娟(1981-),女,陕西西安人,工程师,硕士研究生,主要研究方向:图像处理。
  • 基金资助:

    国家自然科学基金资助项目(60873095)

Curvature estimation for scattered point cloud data

ZHANG Fan1,KANG Baosheng1,ZHAO Jiandong1,LI Juan2   

  1. 1. School of Information Science and Technology, Northwest University, Xian Shaanxi 710127, China
    2. Department of Generic Technology Promotion, Shaanxi Province Science and Technology Resource Center, Xi'an Shaanxi 710061, China
  • Received:2013-01-29 Revised:2013-03-04 Online:2013-06-05 Published:2013-06-01
  • Contact: ZHANG Fan

摘要: 针对带有强噪声离散点云数据曲率计算问题,提出一种基于稳健统计的曲率估计方法。首先,用一个二次曲面拟合三维空间采样点处的局部形状;其次,随机地选择该采样点邻域内的子集,多次执行这样的拟合过程,通过变窗宽的最大核密度估计,就得到了最优拟合曲面;最后,将采样点投影到该曲面上,计算投影点曲率信息,就得到采样点曲率。实验结果表明,所提方法对噪声和离群点是稳健的,特别是随着噪声方差的增大,要明显好于传统的抛物拟合方法。

关键词: 曲率估计, 稳健, 噪声, 点云

Abstract: For resolving the problem of curvature calculation for scattered point cloud data with strong noise, a robust statistics approach to curvature estimation was presented. Firstly the local shape at a sample point in 3D space was fitted by a quadratic surface. In addition,the fitting was performed at multiple times with randomly sampled subsets of points, and the best fitting result evaluated by variable-bandwidth maximum kernel density estimator was obtained. At last, the sample point was projected onto the best fitted surface and the curvatures of the projected point was estimated. The experimental results demonstrate that the proposed method is robust to noise and outliers. Especially with increasing noise variance, the proposed method is significantly better than the traditional parabolic fitting method.

Key words: curvature estimation, robust, noise, point cloud

中图分类号: