Surface roughness measurement by using image processing based techniques

Wan Jusoh, Wan Farhan (2015) Surface roughness measurement by using image processing based techniques. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 pages) Surface Roughness Measurement By Using Image Processing Based Techniques 24 Pages.pdf - Submitted Version Download (316kB)
	Text (Full Text) Surface Roughness Measurement by Using Image Processing Based Techniques.pdf - Submitted Version Restricted to Registered users only Download (1MB)

Abstract

Machine vision is a technology and method used in specimen preparation based on imaging that used to analyze image for an application like automatic inspection, process control and guidance for robot in the industry. Machine vision covers a wider area which based on what application it’s being used for. By using machine vision, image processing technique is widely used these days. In this work, the image processing technique is used to predict values of surface roughness (Ra) by using the mean gray level value calculated from an image captured using a vision system. The image is processed using MATLAB software as to find the mean gray level value of an image. The GUI also can be developed using the MATLAB software in order to ease the calculation of mean gray level value on an image. Surface roughness of specimen are being measured using the stylus profile. Using the surface roughness data and mean gray level value, it is possible to find the relationship between those two data and generate a regression line. By using the regression line equation, new surface roughness based on the linear equation line can be calculated. The new Ra based on the linear equation line will be compared with the value measured using stylus profile as to find the percentage of error for each data collected. In order to validate this method realibility, the maximum allowable error are set to 10% and lower. This method shows that by using the mean gray level value, it is possible to predict the surface roughness of specimen.

Actions (login required)

View Item