Journal of Theoretical and Applied Information Technology 10 th February 2016. Vol.84. No.1 © 2005 - 2015 JATIT & LLS. All rights reserved . ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195 10 VISION BASED INSPECTION SYSTEM FOR MISSING FOOTPRINT DETECTION ON PRINTED CIRCUIT BOARDS 1 MARZIEH MOGHARREBI, 2 SHAHNORBANUN SAHRAN, 3 MEI CHOO ANG, 4 ANTON SATRIA PRABUWONO 1 Institute of Visual Informatics (IVI), Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia 2 Center for Artificial Intelligence Technology(CAIT), Faculty of Information Science and Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia 3 Department of Information Technology, Faculty of Computing and Information Technology in Rabigh King Abdulaziz University, Rabigh 21911, Saudi Arabia E-mail: 1 Mogharrebi.marzieh@gmail.com, 2 shahnorbanun@ukm.edu.my, 3 amc@ukm.edu.my 4 aprabuwono@kau.edu.sa ABSTRACT An automated visual inspection system is needed to inspect missing component footprints on bare Printed Circuit Board (PCB). Missing component footprints on the PCB will result in lack of electronic components. Since the surface of electronic products is more compact and complex, the inspection task has become more difficult in order to achieve the required quality of finished products. The general aim of this research is to develop a real time automated visual inspection system to identify missing footprints on PCB. The specific objectives are: to implement production rules and Fuzzy logic techniques for quality classification of footprints. An experimental setup for a real time AVIS which consists of hardware and software frameworks has been developed. The hardware framework included a web camera, four florescent lamps, a conveyor belt, and the software framework included Image Processing System (IPS). The Region of Interest (ROI) of the PCB would be transferred to IPS under suitable light and specific speed of conveyor belt. The footprints are segmented based on their coordination. Feature extraction will be accomplished to classify the uncompleted footprints on PCBs. Ultimately, the production rule, Mamdani, and Sugeno fuzzy classifiers are implemented to classify the footprints. The experimental result confirmed that the proposed AVIS has a high accuracy rate. Keywords: Printed Circuit Board (PCB); Image Segmentation; Feature Extraction; Fuzzy Logic; Automated Visual Inspection 1. INTRODUCTION Traditionally, the quality inspection task in manufacturing sectors was performed by human [1] and it was dependent on experience and the utilisation of traditional equipment. It was dull, slow and had low inspection accuracy [2]. Hence, the automated visual inspection system (AVIS) was introduced to replace the human inspectors and AVIS was utilized in quality control of PCB in manufacturing [3]. AVIS was also applied in a number of inspection applications such as ship maintenance [4], metal parts classification and defect detection [5], solar cell panel crack detection [6] and ceramic textile [7]. Printed Circuit Board (PCB) is a board to hold and connect electronic components such as resistor, capacitor, diode and others. It is a thin board made out of fiberglass. Conductive pathways are etched or "printed" onto the board, connecting different components on the PCB, such as resistors, transistors and integrated circuits. They are used in many other electronic devices besides PCs. Most TVs, radios, digital cameras, cell phones, and tablets include one or more printed circuit boards [8]. The assembly of components on PCB is considered to be one of the most time intensive stages in PCB manufacturing [9, 10] A bare PCB involves footprints or shapes of the component. The footprints are designed to inform PCB operators the