PCB Inspection using Competitive Learning and Fuzzy Associative Memories

Abstract

Visual printed circuit board inspection in digital images is viewed as a pattern classification problem. The process involves a two level classification of the printed circuit board image sub-patterns into either standard non-defective class or a defective class. The patterns that are identified as being defective in the first level are thoroughly checked for defects in the second level. And the patterns that are non-defective are checked for dimensional verification for the classes it has been identified and assigned to the correct class. This classification uses fuzzy information of the sub-patterns by extracting features from the scan-line grid. A prototype system is development using fuzzy associative memories for dimensional verification. Rules are generated using the features extracted from the scan-line grid with an adaptive vector quantization algorithm that uses differential competitive learning (DCL) in updating winning synaptic vectors. The objective in using fuzzy feature vectors is to drastically reduce the computational time required for inspection, which is a major problem in the conventional inspection systems. The paper concludes with experimental results and directions for future work.

Department(s)

Engineering Management and Systems Engineering

Second Department

Computer Science

Document Type

Article - Journal

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2024 The Authors, All rights reserved.

Publication Date

01 Dec 1994

This document is currently not available here.

Share

 
COinS