Pattern analysis and evaluation of printed circuit boards

An evaluation methodology for printed circuit boards with fine pitches, including defect analysis, is discussed. The evaluation is a big issue for a mass production system. Evaluation accuracy and defect appearance rate decide yield and cost for the manufacturing lot of PC boards. The authors discuss how the inspected PC boards are compared with the original pattern information, which is derived directly from artwork, or CAD data. Topological data of the manufactured PC boards are obtained from an optical inspection system by image processing. Pattern defects are easily detected and their defect types can be also discriminated by the proposed topological comparison method. The authors describe more details about features, and a fast comparison and discrimination method. Results and examples are presented.<<ETX>>

[1]  T. Pavlidis A thinning algorithm for discrete binary images , 1980 .

[2]  M. Ito,et al.  Recognition of pattern defects of printed circuit board using topological information , 1991, [1991 Proceedings] Eleventh IEEE/CHMT International Electronics Manufacturing Technology Symposium.

[3]  Gabriella Sanniti di Baja,et al.  A Width-Independent Fast Thinning Algorithm , 1985, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Carlo Arcelli,et al.  Pattern thinning by contour tracing , 1981 .