Fares Hedayati and Jike Chong and Kurt Keutzer

EECS Department, University of California, Berkeley

Technical Report No. UCB/EECS-2010-138

November 23, 2010

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-138.pdf

Recognition of Tibetan wood block print is a difficult problem that has many challenging steps. We propose a two stage framework involving image preprocessing, which consists of noise removal and baseline detection, and simultaneous character segmentation and recognition by the aid of a generalized hidden Markov model (also known as gHMM). For the latter stage, we train a gHMM and run the generalized Viterbi algorithm on our image to decode observations. There are two major motivations for using gHMM. First, it incorporates a language model into our recognition system which in turn enforces grammar and disambiguates classification errors caused by printing errors and image noise. Second, gHMM solves the segmentation challenge. Simply put gHMM is an HMM where the emission model allows multiple consecutive observations to be mapped to the same state. For features of our emission model we apply line and circle Hough transform to stroke detection, and use classs-pecific scaling for feature weighing. With gHMM, we find KMQDF to be the most effective distance metric for discriminating character classes. The accuracy of our system is 90.03%.


BibTeX citation:

@techreport{Hedayati:EECS-2010-138,
    Author= {Hedayati, Fares and Chong,  Jike and Keutzer, Kurt},
    Title= {Recognition of Tibetan Wood Block Prints with Generalized Hidden Markov and Kernelized Modified Quadratic Distance Function},
    Year= {2010},
    Month= {Nov},
    Url= {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-138.html},
    Number= {UCB/EECS-2010-138},
    Abstract= {  Recognition of Tibetan wood block print is a difficult problem that has many challenging steps. We propose a two stage framework involving image preprocessing, which consists of noise removal and baseline detection, and simultaneous character segmentation and recognition by the aid of a generalized hidden Markov model (also known as gHMM). For the latter stage, we train a gHMM and run the generalized Viterbi algorithm on our image to decode observations. There are two major motivations for using gHMM. First, it incorporates a language model into our recognition system which in turn enforces grammar
and disambiguates classification errors caused by printing errors and image noise. Second, gHMM solves the segmentation
challenge. Simply put gHMM is an HMM where the emission model allows multiple consecutive observations to be mapped to the same state. For features of our emission model we apply line and circle Hough transform to stroke detection, and use classs-pecific scaling for feature weighing. With gHMM, we find KMQDF to be the most effective distance metric for discriminating character classes. The accuracy of our system is 90.03%.},
}

EndNote citation:

%0 Report
%A Hedayati, Fares 
%A Chong,  Jike 
%A Keutzer, Kurt 
%T Recognition of Tibetan Wood Block Prints with Generalized Hidden Markov and Kernelized Modified Quadratic Distance Function
%I EECS Department, University of California, Berkeley
%D 2010
%8 November 23
%@ UCB/EECS-2010-138
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-138.html
%F Hedayati:EECS-2010-138