SOBEL OPERATOR FOR EDGES DETECTION IN SURFACE TEXTURE ANALYSIS

Abstract

The classification of surface textures has been changed for the convenience of design, manufacture and metrology. Surface texture can be divided into two types which are engineered and structured surface. One examples of structured surface is dimple edge type of surface. This kind of surface usually has a deterministic structure, with high aspect ratio geometric features. Modern dimple golf ball is the example of structured surface application witches its functional requirement, was to reduce air drag through better hydrodynamic flow around the ball. From the analysis of dimpled surface, process variation can be enhanced and the uniformity of the surface features can be improved. Hence, the work presented in this paper is intended to improve the quality of segmentation method for the dimple edges type of surface and Sobel Operator algorithm is proposed to be used. In Sobel Operator algorithm, two kernels is applied which one kernel identifies the horizontal edges and the other kernel identifies the vertical edges. By passing these two kernels, gradient along x direction and gradient along the y direction can be computed at the different location and finally determined the edges of the surface. Sobel operator application shows that the location of the edges becomes clearer and topography segmentation based on features edges is possible.