Automated Defects Detection System

Tan, Jun Ning (2017) Automated Defects Detection System. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Tang Jun Ning.pdf Restricted to Registered users only Download (2MB)

Abstract

A study has been conducted to investigate detection system for defect samples. This paper reviews the defective samples are used as input digital images for the system which include image acquisition; pixels and local pre-processing approaches for image pre-processing; region of interest, thresholding-based, size, shape, colour and texture features for image analysis. A different technique is reported to increase accuracy on detection of defects in the sample. This project is a combination of hardware part and software part. The hardware of the system includes a camera and a computer. The camera of this project is used a Canon EOS 1100D to capture images and loaded into the LabVIEW platform. The wooden box also called as black box is used to tolerate the uncertainty such as exposure of light and position of the sample. Most of the uncertainty is avoided in this project. The software part is written in LabVIEW with Graphical-based programming language to create programs called Virtual Instruments (VI) in a pictorial form called a block diagram. LabVIEW provides many methods of image acquisition. It is observed that techniques which follow the stage process of detection of defects achieved accuracy of 73.33% for 50% similarity percentage on Sample C.