%0 Conference Proceedings
%4 sid.inpe.br/banon/2004/08.19.00.32
%2 sid.inpe.br/banon/2004/08.19.00.32.37
%@doi 10.1109/SIBGRA.2004.1352947
%T Handwritten Recognition with Multiple Classifiers for Restricted Lexicon
%D 2004
%A Oliveira Júnior, José Josemar,
%A Kapp, Marcelo Nepomoceno,
%A Freitas, Cinthia Obladen de Almendra,
%A Carvalho, João Marques de,
%A Sabourin, Robert,
%@affiliation Universidade Federal de Campina Grande, Coordenação de Pós-Graduação em Engenharia Elétrica, Caixa Postal 10105, 58109-970, Campina Grande, PB - Brazil,
%@affiliation Pontíficia Universidade Católica do Paraná, R. Imaculada Conceição 1155, 80215-901, Curitiba, PR - Brazil,
%@affiliation Ècole de Technologie Superieure, 1100 Rue Notre Dame Ouest, H3C 1K3, Montreal, QC - Canada,
%E Araújo, Arnaldo de Albuquerque,
%E Comba, João Luiz Dihl,
%E Navazo, Isabel,
%E Sousa, Antônio Augusto de,
%B Brazilian Symposium on Computer Graphics and Image Processing, 17 (SIBGRAPI) - Ibero-American Symposium on Computer Graphics, 2 (SIACG)
%C Curitiba, PR, Brazil
%8 17-20 Oct. 2004
%I IEEE Computer Society
%J Los Alamitos
%S Proceedings
%K pattern recognition, multiple classifiers, handwritten recognition.
%X This paper prsents a multiple classifier system applied to the handwritten word recognition (HWR) problem. The goal is to analyse the influence of different global classifiers taken isolatedly as well as combined in a particular HWR task. The application proposed is the recognition of the Portuguese handwritten names of the months. The strategy takes advantage of the complementarity mechanisms of three different classifiers: Conventional Neural Network, Class-Modular Neural Network and Hidden Markov Models, yielding a multiple classifier that is more efficient than either individual technique. The recognition rates obtained vary from 75.9% using the stand alone HMM classifier to 96.0% considering the classifiers combination.
%@language en
%3 4382_oliveira_jose.pdf