Summary
International Technical Conference on Circuits/Systems, Computers and Communications
2016
Session Number:T1-5
Session:
Number:4508
Defect Classification of Electronic Board Using Bag of Features and Color Information
Hidenobu Inoue, Yuji Iwahori, Boonserm Kijsirikul, Manas K. Bhuyan ,
pp.463-466
Publication Date:2016/7/10
Online ISSN:2188-5079
DOI:10.34385/proc.61.4508
PDF download (1.1MB)
Summary:
This paper proposes a new approach of defect classification using Bag of Features and color information in order to correspond to the various defect images without using any reference images. The purpose of the paper is to classify the true defect and pseudo defect such as dust on the electronic board. First, features are extracted from each image of data set and histogram features are generated and represented by Bag of Features. Next, noise removal is applied and combined features which consist of Bag of Features and color information are used. After extracting features, SVM (Support Vector Machine) is used for the learning and classification. The usefulness of the proposed approach is confirmed by evaluating the accuracy of defect classification in comparison with the previous approaches with the target images of electronic board images which includes the actual defects.