Development of Double Iteration Sobel (DIS) for edge identification in surface features segmentation.

Measuring the surface feature falls within the broad spectrum of Quality Control and Inspection. Since most processes in the manufacturing industry that are used to prepare surfaces often induce defects and regularities, surface metrology is required to analyse surface topography. The study of surface topography analysis focuses on surface features segmentation procedure which is crucial since it will help to enhance the essential features found on the surface. Although there are many features segmentation procedures that can help in identifying the edge on the surface as presented in previous studies, problem arises when most of them are unable to fulfill the requirements when being implemented on a particular surface topography. Thus, an approach of Double Iteration Sobel has been introduced and highlighted by considering the effects of existing surface features segmentation procedures. The aim of this development is to increase the accuracy and stability of the feature edge detection to be close to human perception on the actual edge location. The process of development and implementation on simulated 3D sloped bumps surface topography data is carried out using MATLAB software. Overall, the findings of this study showed that with the help of DIS, the effectiveness of marker-controlled watershed segmentation has been successfully improved in terms of the edge visualization and marking process. [ABSTRACT FROM AUTHOR]