Study on Brittle Graphite Surface Roughness Detection Based on Gray-Level Co-occurrence Matrix

Abstract

The brittle graphite machined surface quality cannot be completely evaluated only by surface roughness Ra measured by profilometer, owing to its 3D surface defects. A digital image detection method based on gray-level co-occurrence matrix was put forwards to detect the graphite surface quality. The results show that two feature parameters of gray-level co-occurrence matrix, including Secondary moment ASM and Entropy ENT, had good positive relevance with surface roughness Ra. The linear regression functions of ASM and ENT were established to fit Ra, which could be used to predict Ra by image processing method. This method also gives a feasible way to develop a machine vision detection system to automatically detect the graphite surface quality by digital image processing.