Product Surface Roughness Measurement Based on the Fractal Feature of the Laser Speckle Image

摘要

The product surface roughness measurement occupies an important position in the manufacturing process of the industrial product. The laser speckle image can be used for the non-contacted measurement. The Speckle images are produced by the reflected and scattered light beams from rough surface through free-space to observing plane when laser illuminates the object surface. Statistical distribution of speckles depends on the microscopic structure of the rough surface and can be used to distinguish the surface roughness. Firstly, for the existence of the noise and redundancy in the laser speckle image, the PCA(principal component analysis) method is utilized in the image processing. After extracting the principal components in the original image matrix, the reconstruction image which removed noises and irrelevances was earned. Secondly, the fractal features of reconstruction images were extracted by using the Double Blanket Method. The fractal dimension of the reconstruction image was analyzed under the moving window with optimum size to obtain the fractal dimension histogram. By comparing the histogram with the surface roughness, the obvious correlations of the frequency point distributing of the fractal dimension histogram and the product surface roughness was shown. On these bases, the multi-scale fractal features were extracted for the single-scales limitation. So, the method of product surface roughness measurement based on the fractal feature of the laser speckle image was given by the research. The measure system set-up of the method is simple, fast, and not sensitive to change of circumstance and vibration. Hence, it has great potential for application to in-process measurement.