Feature-Extraction-Based Inspection Algorithm for IC Solder Joints

• Published in: IEEE transactions on components, packaging, and manufacturing technology (2011) Vol. 1; no. 5; pp. 689 - 694
• Main Authors: Wu, Fupei, Zhang, Xianmin
• Format: Journal Article
• Language: English
• Published: Piscataway, NJ IEEE 01.05.2011; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Applied sciences; Automatic optical inspection; Design. Technologies. Operation analysis. Testing; digital feature; Electronic equipment and fabrication. Passive components, printed wiring boards, connectics; Electronics; Exact sciences and technology; Feature extraction; Inspection; integrated circuit component; Integrated circuits; Lead; logical feature; Pixel; Printed circuit boards; Semiconductor electronics. Microelectronics. Optoelectronics. Solid state devices; Shape; shape feature; solder joint; Soldering; On board equipment; Performance evaluation; logical feature; solder joint; Printed circuit board; Automatic optical inspection; Algorithm; shape feature; Integrated circuit bonding; integrated circuit component; digital feature; Soldered joint; Integrated circuit; Optical characteristic; Feature extraction; Printed circuit; Optical reflection; Color center
• ISSN: 2156-3950; 2156-3985
• DOI: 10.1109/TCPMT.2011.2118208

In order to inspect solder joint defects of integrated circuit (IC) components on printed circuit boards (PCBs), an automatic optical inspection (AOI) algorithm is developed. Firstly, considering the shape of a solder joint and its optical reflection property, the IC solder joint is divided into several subregions, which are called shape features. Secondly, the digital features that are used to evaluate the solder quality of the subregions are expressed by the color, area, mass center, and continuous pixels. Thirdly, the logical features are developed according to the color distributions of the solder joint image and the relationships between different subregions and the types of solder joints. Finally, to evaluate the performance of the proposed algorithm, 137 PCBs with defects were inspected by an AOI system that integrates with the proposed algorithm. Inspection results show that solder joint defects of IC such as surplus solder, lacking solder, no solder, lead lift, lead bend, shift, and bridged and pseudo joints can be identified effectively.