Title :
Development of Enhanced Active Shape Model (EASM) for in-situ Automatic Optical Inspection (AOI)
Author :
Chen, Liang-Chia ; Tapilouw, Abraham Mario ; Van Thai, Nguyen
Author_Institution :
Grad. Inst. of Autom. Technol., Nat. Taipei Univ. of Technol., Taipei, Taiwan
Abstract :
Conventional current active shape model (ASM) applied to in-situ Automatic Optical Inspection (AOI) has the drawback of inability to locate position, size or orientation of detected targets. This article presents an enhanced active shape model (EASM), which evaluates both offset and orientation during component detection and positioning. The developed method is employed to inspect various electronic components mounted on printed circuit boards (PCB). Experimental results show that it can achieve both robustness and accuracy. In addition, the accuracy of component positioning increases with increasing number of sample images used for building the model, and peaks at the threshold number of images, which varies with quality of the EASM in representing the detected component. Our findings validate the feasibility of applying the developed approach to production environment.
Keywords :
automatic optical inspection; image sampling; object detection; printed circuits; shape recognition; AOI; EASM; automatic optical inspection; enhanced active shape model; image sampling; in-situ; printed circuit boards; target detection; Accuracy; Active shape model; Biomedical imaging; Facial features; Robustness; Shape; Training;
Conference_Titel :
Genetic and Evolutionary Computing (ICGEC), 2010 Fourth International Conference on
Conference_Location :
Shenzhen
Print_ISBN :
978-1-4244-8891-9
Electronic_ISBN :
978-0-7695-4281-2
DOI :
10.1109/ICGEC.2010.108