Inspection of surface defects on engine cylinder wall using segmentation image processing technique

Abstract

A regression formula has been developed for dividing the surface image obtained from an engine cylinder wall. By applying the formula, segmentation error can be minimized. The error, which results from the separation of the cylinder wall image, will lead to a potential loss of precision when the surface images of the cylinder wall are processed. By using the segmental image processing method, the experimental results show that the formula is correct and the technique is reliable.