Title :
Font clustering and classification in document images
Author :
Ozturk, Serdar ; Sankur, Billent ; Abak, A.Toygar
Author_Institution :
Boğaziçi University, Department of Electrical-Electronic Engineering, Bebek, Istanbul, Turkey
Abstract :
Clustering and identification of fonts in document images impacts on the performance of optical character recognition (OCR). Therefore font features and their clustering tendency are investigated. Font clustering is implemented both from shape similarity and from OCR performance points of view. A font recognition algorithm is developed to identify the font group with which a given text was created.
Keywords :
Character recognition; Clustering algorithms; Discrete cosine transforms; Feature extraction; Optical character recognition software; Text recognition; Vectors;
Conference_Titel :
Signal Processing Conference, 2000 10th European
Conference_Location :
Tampere, Finland
Print_ISBN :
978-952-1504-43-3
