摘要
为了压缩大规模激光扫描测点,提出了基于三角面片的自适应数据压缩方法.采用环形数据结构有效存储大规模数据点,通过计算点到平均平面的有限误差距离进行压缩点判别.在局部三角化中考虑了三角形内角和顶点距离,采用规格化最小顶点距离法实现局部三角化,避免了狭长三角形出现.实验表明,该数据压缩方法具有很高的效率,在较大的压缩比下仍可获得可靠的表面精度.数据压缩后的顶点是原来大规模测量点的优化子集,所生成的优化三角面片可以直接生成STL(stereolithography)文件或者数控加工路径,避免了手工建模.
To compress massive data from laser scanner, an adaptive data compression approach based on triangulated surfaces was proposed. It used ring data structure to save massive data effectively. Bounded error-distances from vertices to average plane were computed to find prime vertices for removal. A new re-triangulation method based on normalized minimum-vertex-distance was presented to avoid long and thin triangles, and it considered the vertex distance and the interior angle of triangle. The experimental results indicate that the data compression approach has high efficiency and reliable surface precision with a high compression ratio. The vertices after compression constitute an optimal subset of the original vertices. The optimal triangulated-surfaces created after compression can be used to generate STL (stereolithography) file or NC (numeral control) code directly, which avoids manual modeling.
出处
《浙江大学学报（工学版）》
EI
CAS
CSCD
北大核心
2004年第9期1200-1203,共4页
Journal of Zhejiang University：Engineering Science
基金
浙江省科技计划重点资助项目(2003C21031).
关键词
数据压缩
反求工程
三角面片
最小顶点距离法
环形数据结构
Data structures
Iterative methods
Machining
Numerical control systems
Rapid prototyping
Reverse engineering
