Surface Roughness Partition Based On Gaussian Filter Technology

Surface roughness of part is one of the most important precision index to evaluate machining parts quality, so how to accurately extract the roughness information is particularly important. At present, the two-dimensional roughness has been unable to meet the needs of the engineering that the experts and scholars have begun to look for new ways to obtain more comprehensive information about the surface roughness, the evaluation and the assessment of the roughness information begin to transform from 2D to 3D.The usual methods of filtering are compared to choose the Gaussian filter to extract the roughness information. The filtering principle and the filtering process are analyzed base on the traditional Gaussian filter method, the problem of the traditional Gaussian filter was been found. Robust estimation theory is discussed. For the boundary effect of the Gaussian filtering, the vertical function of the robust estimation is introduced to deal with the robustness of the Gaussian filtering. Under the environment of the MATLAB software programming, Gaussian filtering algorithm is programmed to simulate the two-dimensional original surface contour of the parts, and the roughness information of the two-dimensional surface is extracted by the robust Gaussian filter to analyze the roughness information. On the basis of the twodimensional extraction, the two-dimensional surface extraction was extended to the three-dimensional. According to the principle of the robust Gaussian filtering, twodimensional Gaussian filtering algorithm model was set up to implement Gaussian filtering process. The three-dimensional surface roughness information was extracted out.The roughness parameter evaluation method is analyzed to establish the roughness evaluation datum. According to the roughness parameter system, the appropriate roughness parameters are selected to be calculated. The roughness standard block and parts are measured to choose the appropriate sample points and sample area for the original surface topography. The two-dimensional and threedimensional roughness information was extracted by Gaussian filtering, and the twodimensional and three dimensional roughness parameter values are calculated by the principle of the roughness parameters algorithm. On the basis of the experiments and measurements, the roughness measurement and the roughness parameter values which are extracted by the Gaussian filter are compared to calculate the relative error and evaluate the error range, so we can verify the correctness and feasibility of the method in this paper.