Defect Exclusive Custom Vocabulary for Classification

Terence Sweeney, Dermot Kerr, Sonya Coleman

Research output: Contribution to conferencePaperpeer-review

81 Downloads (Pure)


Automated inspection has become a vital part of quality control in many industries, including during
semiconductor wafer production. Current processes often focus on finding defects by comparing images
with a ‘golden’ image pixel to pixel or, more recently, using shallow or deep learning based approaches.
We present an alternative approach which uses the Bag of Visual Words technique to determine local
features that correspond to specific defects within a wafer image, known as a custom vocabulary. Using
this custom vocabulary combined with machine learning, we can characterise and accurately classify
defects found on wafer images.
Original languageEnglish
Number of pages4
Publication statusPublished (in print/issue) - 30 Aug 2020
EventIrish Machine Vision and Image Processing Conference (IMVIP) 2020 - ITSligo / Virtual , Ireland
Duration: 31 Aug 20202 Sept 2020


ConferenceIrish Machine Vision and Image Processing Conference (IMVIP) 2020
Abbreviated titleIMVIP 2020
Internet address


  • Defect Detection
  • Local Features
  • Classification
  • Bag of Visual Words
  • Machine Vision


Dive into the research topics of 'Defect Exclusive Custom Vocabulary for Classification'. Together they form a unique fingerprint.

Cite this