Surface roughness measurement based on image comparison

To solve the measure of the inner wall's surface roughness of slender hole and slit, a comparative measure method was proposed on the basis of extracting the surface image feature of parts and standard blocks by utilizing image processing. The surface image features were analyzed, and the surface image could be classified into two kinds, one had significant texture features and the other had no texture features. Taken texture feature as the template the surface images with significant texture features were classified according to the maximum degree of fuzzy classification. The images with no significant texture features could be expressed by pixel features. Compared the template with part surface to be measured by template matching method, the surface roughness of parts could be obtained by utilizing the maximal correlation coefficient. The results show that this image comparison measure method is convenient and fast, and it can measure the surface roughness of parts reasonably well.