基于曲率弦高法的海量测量数据精简

Massive Cloud Data Simplification Based on Curvature Hypotenuse-height Method

针对视觉测量的点云数据过大而不利于计算和重构的问题，在分析视觉测量点云数据特征的基础上，将曲率原则和弦高法相结合，提出基于曲率弦高法的海量测量数据精简方法。该方法在考虑曲线曲率变化的基础上构建弦高函数，并通过迭代得到各测点变化的弦高值，再根据弦高法的数据精简原则确定需要保留的测量点。仿真实验表明，该方法在平均误差小于0．2mm时，精简率为89．8％，能够有效地对海量点云数据进行精简，并实现精简后测点按曲面曲率的合理分布。

In the field of vision measurement, the size of clouds data of the point which obtained in vision measurement process is so large that it is inconvenient to be calculated, stored and reconstructed. Based on the analysis of cloud data characteristics of vision measurement point, combined curvature principle with the chord height algorithm, a mass cloud data reducing method based on curvature chord height algorithm is presented. With this method, the function of the chord height was built considering the curvature variation of the curves, and the value of changing chord height of each point is obtained by iteration, and then the measurement points to be retained were determined based on the data reducing method of chord height algorithm. Simulation results showed that the cloud data reduced 89. 8% through this method when the average error was less than 0. 2mm. So the cloud data of mass points were reduced directly and effectively, and the reasonable distribution of measurement point according to the curvature of the surface was realized.