H-minima transform assisted feature characterization system in areal surface metrology / Nor Hidayawati Mohd Zaini

Feature parameter is used to characterize the surface texture according to international standard ISO 25178 part 2. The topography data must undergo segmentation prior to the characterization. However, Watershed segmentation could results in over-segmentation. A mechanism is required to eliminate ov...

Full description

Saved in:
Bibliographic Details
Main Author: Mohd Zaini, Nor Hidayawati
Format: Thesis
Language:English
Published: 2018
Online Access:https://ir.uitm.edu.my/id/eprint/89198/1/89198.pdf
https://ir.uitm.edu.my/id/eprint/89198/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Feature parameter is used to characterize the surface texture according to international standard ISO 25178 part 2. The topography data must undergo segmentation prior to the characterization. However, Watershed segmentation could results in over-segmentation. A mechanism is required to eliminate over-segmentation and retain the necessary information for characterization. Thus, this research introduced the use of H-minima transform prior the segmentation and eliminate the use of Wolf pruning. H-minima transform is introduced before Watershed segmentation carried out. It can suppresses all regional minima in topography surface, whose depth less than or equivalent to threshold value. Demonstration of system design was tested on simulated data. The validation of the system design was verified via feature counts and feature segmentation conditions. Threshold value of H-minima transform is depending on percentage of total surface height. It was different for various type of different surface. The range of the H- minima threshold value obtained to remove insignificant features between 5% and 20% of total surface height. The results showed that, feature characterization system is able to eliminate over-segmentation and extract significant features from the topography surface. This system design could be considered as a toolbox solution for feature characterization in surface metrology.