PRODUCT NAME An Automated Visual Printed Circuit Board Inspection System With Defect Classification Capability
ABSTRACT Automated visual inspection of printed circuit board(PCB) is indispensible in electronic industries since manual inspection of bare PCB have been proposed in the literature. These approaches can be divided into three main categories : reference-based approach, design rule checking approach and hybrid approach, which involved a combination of reference comparison and design-rule approach. Note that these approaches emphasized more on the defect detection on printed circuit board (PCB). However, in order to identify the root cause of these defects, the defect classification must be performed after the detection. Based on the literature, the defect classification has been overlooked and not given sufficient attention. Therefore, the enhancement of PCB inspection algorithm and system by incorporating defect classification element is the primary concern of this research proposal. Even though a defect classification algorithm has been develop previously, the algorithm has been tested only tto computer generated PCB images. The performance of the defect classification algorithm may be degraded if the inspection is implemented on real PCB images. Hence, the first step of this research is to enhance the PCB defects classification algorithm by considering real PCB images. The illumination and alligment issues, which are common in real-time automated visual inspection, will be solved algorithmically and by properly investigating the best lighting set up for the system. The final up of this prototype will consists of high resolution cameras, LED light for lighting purpose, and a computer to process the images. The prototype will provide better assessment of PCB based on the abilityto classify the defects compared to the existing approaches or sytems.
FILING COUNTRY Malaysia
REG. NUMBER
INTELLECTUAL STATUS Novel
FILE DATE
IP TYPE Patent
YEAR APPLY 2012
DEPARTMENT PUSAT PENGAJIAN KEJURUTERAAN MEKATRONIK
©2014 UniMAP