Abstract An automatic character detection system for IC test handler is designed to recognize the characters on the surface of IC chip based on active learning SVM. Firstly, industrial camera is employed to collect a large number of chips’ surface image. Secondly, image preprocessing is carried out, including image grayscale, binarization and filter processing. Thirdly, the features of the preprocessed image are extracted. To reduce the annotation cost for training data and improve recognition rate, active learning algorithm is used to label the training data, while support vector machine algorithm is used to classify those data. Comparison with SVM algorithm, template matching and BP neural network shows the effectiveness of the proposed algorithm.
Automatic Character Detection System for IC Test Handler Based on Active Learning SVM
2017-01-01
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch