Defect Detection On The Metal Surface By Using Machine Vision Sensor

Ahmad, Muhammad Ismail (2015) Defect Detection On The Metal Surface By Using Machine Vision Sensor. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Defect Detection On The Metal Surface By Using Machine Vision Sensor 24 Pages.pdf - Submitted Version Download (364kB)
	Text (Full Text) Defect Detection on The Metal Surface By Using Machine Vision Sensor.pdf - Submitted Version Restricted to Repository staff only Download (1MB)

Abstract

This report presents a defect detection system to replace the conventional method of inspection done by human operator. In using manual inspection, several problems were countered where by human inspector cannot maintain long-term consistency due lack of knowledge and experience. Standardization is impossible because depends on their understanding of the individual inspector. A defect detection is very important to control the quality of the products, as scratches are the most common defect in industry. The objectives of the current study is to develop an automation system by using vision sensor that can detect scratch defect on the metal surface and to develop the algorithm for defect detection system. The method involved five procedures : (1) obtaining desired image by using cropping, (2) adjusting the images to more suitable display for analysis using Image Enhancement, (3) converting image condition into binary, (4) removing unwanted point by shape filtering using morphological filter with new structuring elements defined based on the shape of characteristic of scratches and (5) analyzing the defect by calculating the area of scratch. The MATLAB software was used to build this system and GUI Interface used as user interface. From this experiment, 104 sample images have been inspected and the efficiency of system achieved is 99%. The environment for the inspection should be controlled by reducing the illumination of the area avoiding any glare or reflection from occurring to the product. Finally, the ability of this system needs to be upgraded to detect another defects so that the system can perform better to control the quality of the product.

Actions (login required)

View Item