Title :
Evaluation of binarization methods for document images
Author :
Trier, Øivind Due ; Taxt, Torfinn
Author_Institution :
Dept. of Comput. Sci., Michigan State Univ., East Lansing, MI, USA
fDate :
3/1/1995 12:00:00 AM
Abstract :
This paper presents an evaluation of eleven locally adaptive binarization methods for gray scale images with low contrast, variable background intensity and noise. Niblack´s method (1986) with the addition of the postprocessing step of Yanowitz and Bruckstein´s method (1989) added performed the best and was also one of the fastest binarization methods
Keywords :
document image processing; image processing; document images; gray scale images; locally adaptive binarization methods; low contrast; noise; postprocessing; variable background intensity; Availability; Background noise; Data mining; Engineering drawings; Image analysis; Labeling; Pixel; Postal services; Stochastic resonance; Text analysis;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
