摘要
以激光-机器视觉测量方式得到的曲面数据云为基础,探讨了基于给定精度的曲面密集散乱数据点群的数据压缩以及几何建模方法.其中根据激光测量方式和三维点群分布的特点,提出一种在给定采样精度基于曲率的曲面自适应采样的方法;并通过激光扫描曲线间的采样点匹配为进一步数据压缩提供依据;并由度量曲面的初步网格点阵对测量点逼近程度好坏来进一步完成对曲面模型的修正.通过实例验证了这种方法的可行性.
Based on a laser-machine-vision measurement system, the curved surface data are first acquired. A data compression and geometric modeling method is investigated on a given precision mass data in the form of sparse point clouds. Subsequently, an adaptive curvature-oriented sampling technology is developed based on the distribution of three-dimensional point clouds. In this regard, data compression is facilitated through matching sampled points among laser scanned curves. Further studied, the surface model is revised according to the closeness between initial grid points and measured points. Consequently, this technology has been proven feasible upon completion of intensive testing.
出处
《中国工程机械学报》
2006年第4期474-477,482,共5页
Chinese Journal of Construction Machinery
基金
国家自然科学基金资助项目(59675075)
国家教育部高等学校博士学科点专项科研基金资助项目(97069835)
关键词
数据云
几何建模
自适应采样
data cloud
geometric modeling
adaptive sampling
