%0 Conference Proceedings
%T Determining the appropriate feature set for fish classification tasks
%D 2005
%A Nery, Marcelo Souza,
%A Campos, Mario Fernando Montenegro,
%A Pádua, Flávio Luis Cardeal,
%A Queiroz Neto, José Pinheiro de,
%A Machado, Alexei Manso Correa,
%A Carceroni, Rodrigo Lima,
%@affiliation Departamento de Ciência da Computação - Universidade Federal de Minas Gerais
%@affiliation Pontifícia Universidade Católica de Minas Gerais
%@affiliation Centro Federal de Educação Tecnológica do Amazonas
%E Rodrigues, Maria Andréia Formico,
%E Frery, Alejandro César,
%B Brazilian Symposium on Computer Graphics and Image Processing, 18 (SIBGRAPI)
%C Natal
%8 9-12 Oct. 2005
%I IEEE Computer Society
%J Los Alamitos
%S Proceedings
%K object classification, feature extraction, feature selection, fish classification.
%X We present a novel fish classification methodology based on a robust feature selection technique. Unlike existing works for fish classification, which propose descriptors and do not analyze their individual impacts in the whole classification task, we propose a general set of features and their correspondent weights that should be used as a priori information by the classifier. In this sense, instead of studying techniques for improving the classifiers structure itself, we consider it as a "black box" and focus our research in the determination of which input information must bring a robust fish discrimination. All the experiments were performed with fish species of Rio Grande river in Minas Gerais, Brazil. This work has been developed as part of a wider research, which has as main goal the development of effective fish ladders for the Brazilian dams.
%@language en
%3 paduaf_fishclassification.pdf