A digital image processing model for characteristics capture and analysis of irregular electronic components
Amid diversification, high performance, high integration, and low-cost development in consumer electronic products, the conventional system on board function integration model is giving way to system on chip (SoC). Layout of solder balls under ball grid array (BGA) is getting irregular as system in...
Gespeichert in:
Veröffentlicht in: | International journal of advanced manufacturing technology 2019-06, Vol.102 (9-12), p.4309-4318 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Amid diversification, high performance, high integration, and low-cost development in consumer electronic products, the conventional system on board function integration model is giving way to system on chip (SoC). Layout of solder balls under ball grid array (BGA) is getting irregular as system in package (SiP) is integrating multi-chip module and couple of active and passive components along with more complex circuit design. Due to the shortening life cycles of consumer electronic products and frequent new product launches, the incoming inspection operation of electronic components is getting tedious, let alone the component database updates required by mounter are also getting more frequent and time consuming due to increasingly complex and diversified components and soaring demands in characteristic measurement and identification difficulties. This study is aimed at building automated characteristic capture and analysis image processing model for BGA packages of irregular solder ball layout. It employs edge detection to get component body looks and dimensions by enhanced component edge characteristics. A image pre-processing model is used to identify consistency of individual solder ball characteristics with median filter, highlight image characteristics with binary conversion, remove noise with morphology, and connected component labeling to segment area covered by individual characteristic. In addition, a roundness and aspect ratio of the least rectangle-based solder ball characteristics identification method is proposed to determine number of, diameter of, and coordinates of center of border solder balls. The component characteristics acquired by the model comes in ± 3% difference against the actual one in terms of dimension. The model appears a good reference for system packaging operator in component database creation for the launch of new products in terms of reduced time for manual measurement and error as well as loss in mounter capacity. It also serves for electronic components incoming inspection to reduce inspection operation time and simplify the process. |
---|---|
ISSN: | 0268-3768 1433-3015 |
DOI: | 10.1007/s00170-019-03451-5 |