Analysis Of Image's Texture Based On Autocorrelation Function For Non-Planar Surface Roughness

With the rapid growing of economic development, there are more requirement for the surface quality of work-piece. One of the most important parameter for judging the surface quality of work-piece is surface roughness. So the measurement of the surface roughness is essential. Due to the technology of laser measurement has the characteristic of fast, non-contact, and nondestructive, it becomes more and more popular. With in-depth understanding of speckle aroused by correlate laser, laser speckle techniques, as a statistical measurement technology, have been used in the field of measuring the surface roughness by more and more people. This paper adopts the laser speckle method to research the surface roughness of non-planar rough block sample.This paper adopts the auto-correlative function to research the surface roughness of non-planar work-pieces which are made by different processing procedure. First, the test device is established to acquire plenty of speckle images by CCD. Using the graphic processing software to process the laser speckle image is acquired by experiment. Discussing the relationship between contrast ratio and surface roughness and the relationship between auto-correlative function as well as its spreading parameter and image texture characteristic, and then required the variation of autocorrelation function along different directions, as well as the change curve of its spreading parameter with surface roughness.This research provide an approach of multipoint detecting, acquiring and analyzing the speckle image parameter such as contrast ratio, measurement parameter for analyzing the roughness of non-planar work-piece. This approach has academic significance and application value.