基于鲁棒最小二乘法的B样条曲面重构技术

The surface reconstruction by robust least-squares regression which based on B-spline

利用非接触式长距离激光扫描测量技术实现对目标模型的精确测量,对测量获得的点云数据,应用鲁棒最小二乘法(Robust LSR)来拟合B样条曲线曲面,该方法对拟合曲线上不同距离的残差赋予不同权值,重复迭代进行运算。从而有效抑制噪声数据点对拟合曲线曲面的影响。

By using non contact type long distance laser scanning measurement technology to achieve the goal of accurate model- ling of the measurement,the measurement of the point cloud data,the application of Robust Least Squares（ Robust LSR ） to the fitting of B spline curves and surfaces, the method of fitting curve at different distances to give different weights of residuals, itera- tire algorithm. In order to restrain the noise data points on the curve surface effects.