CSDL Home I ICDAR 2005 Proceedings. Eighth International Conference on Document Analysis and Recognition
Subscribe
Seoul, South Korea
Aug. 31, 2005 to Sept. 1, 2005
ISBN: 0-7695-2420-6
pp: 374-378
Jangkyun Park , Division of Computer Science, KAIST, Korea
Younghee Kwon , Division of Computer Science, KAIST, Korea
Jin Hyung Kim , Division of Computer Science, KAIST, Korea
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ICDAR.2005.49
ABSTRACT
This paper presents a prior model for text image superresolution in the Bayesian framework. In contrast to generic image super-resolution task, super-resolution of text images can be benefited from strong prior knowledge of the image class: Firstly, low-resolution images are assumed to be generated from a highresolution image by a sort of degradation which can be grasped through example pairs of the original and the corresponding degradation; Secondly, text images are composed of two homogeneous regions, text and background regions. These properties were represented in a Markov Random Field (MRF) framework. Experiments showed that our model is more appropriate to text image super-resolution than the other prior models.
INDEX TERMS
null
CITATION
Jangkyun Park,
Younghee Kwon,
Jin Hyung Kim,
"An Example-based Prior Model for Text Image Super-resolution", ICDAR, 2005, Proceedings. Eighth International Conference on Document Analysis and Recognition,
Proceedings. Eighth International Conference on Document Analysis and Recognition 2005, pp. 374-378, doi:10.1109/ICDAR.2005.49