测绘通报 ›› 2018, Vol. 0 ›› Issue (12): 101-104,113.doi: 10.13474/j.cnki.11-2246.2018.0392

• 技术交流 • 上一篇    下一篇

最优邻域二次误差曲面的点云简化算法

陆建华, 刘闯, 吕志才   

  1. 苏州市测绘院有限责任公司, 江苏 苏州 215000
  • 收稿日期:2018-03-21 修回日期:2018-04-25 出版日期:2018-12-25 发布日期:2019-01-03
  • 通讯作者: 刘闯。E-mail:1137157737@qq.com E-mail:1137157737@qq.com
  • 作者简介:陆建华(1974-),男,硕士,教授级高级工程师,主要从事工程测量及三维激光数据处理方面的研究。E-mail:475375848@qq.com
  • 基金资助:
    江苏省第五期“333工程”科研资助立项项目(BRA2016069);苏州市2017年度产业技术创新专项(民生科技)项目(SS201740);国家自然青年科学基金(41501502)

Point Cloud Simplification Algorithm of Quadric Error Based on the Optimal Neighborhood

LU Jianhua, LIU Chuang, LÜ Zhicai   

  1. Suzhou Surveying Institute Company, Suzhou 215000, China
  • Received:2018-03-21 Revised:2018-04-25 Online:2018-12-25 Published:2019-01-03

摘要: 针对点云简化算法中普遍存在的误差累积与传递问题,提出基于最优邻域二次误差曲面的点云简化算法。该算法依据点云邻域的几何特性提取特征点;构建基于维度特征的熵函数,以最小熵函数原则确定最优邻域,结合二次曲面拟合误差控制理论,实现非特征点云的简化;构建基于二次误差曲面的信息熵评价指标对简化结果进行评价。试验表明,该算法能够有效减弱误差影响,更好地服务于后期的曲面重建。

关键词: 维度特征, 最优邻域, 二次曲面误差, 点云简化, 信息熵

Abstract: Point cloud simplification algorithm of quadric error based on the optimal neighborhood is proposed,aiming at solving the problem of error accumulated and error transferred.The feature points are extracted based on the geometric properties of point cloud neighborhood.The entropy function is built based on dimension feature,the optimal neighborhood is determined with the minimum entropy function principle,the non-feature of point cloud simplification is realized with the theory of quadric surface fitting error control combined.The result of the simplified is evaluated by the information entropy function based on quadric error evaluation index.Experiment shows that the algorithm can effectively decrease error influence,better service in the late of surface reconstruction.

Key words: dimension feature, the optimal neighborhood, quadric error, point cloud simplification, the information entropy

中图分类号: