关键词: 表面粗糙度, 高斯回归滤波, 稳健滤波, 快速算法

Abstract:

To solve the time consuming problem in implementation of the robust Gaussian regression filter, an algorithm of fast convolution using truncated Gaussian weighting function was presented. On the basis of robust Gaussian regression filter and convolution function, the truncated Gaussian weighting function was derived, and its appropriate parameter was
obtained through experiments. The experimental results show that the running time of robust Gaussian regression filter using truncated Gaussian weighting function is shortened obviously while maintaining the precision.

Key words: surface roughness, Gaussian regression filter, robust filter, fast algorithm