Process Biotechnology

Plant Biotechnology

Electronic Journal of Biotechnology ISSN: 0717-3458 Vol. 8 No. 3, Issue of December 15, 2005
© 2005 by Pontificia Universidad Católica de Valparaíso -- Chile Received February 15, 2005 / Accepted June 22 , 2005
DOI: 10.2225/vol8-issue3-fulltext-1  
RESEARCH ARTICLE

Comparison of shape analysis methods for Guinardia citricarpa ascospore characterization

Mário Augusto Pazoti
Institute of Mathematics and Computer Science
University of São Paulo
Av. Trabalhador São-Carlense,400, Centro CP 668
CEP 13560-970
São Carlos, São Paulo, Brazil
Tel: 55 16 33739671
E-mail: pazoti@icmc.usp.br

Rogério Eduardo Garcia
Institute of Mathematics and Computer Science
University of São Paulo
Av. Trabalhador São-Carlense,400, Centro CP 668
CEP 13560-970
São Carlos, São Paulo, Brazil
Tel: 55 16 33739671
E-mail: regarcia@icmc.usp.br

José Dalton Cruz Pessoa
Embrapa Instrumentação Agropecuária
Rua XV de Novembro, 1452
CEP 13560-970
São Carlos, SP, Brazil
Tel: 55 16 33739671
E-mail: dalton@cnpdia.embrapa.br

Odemir Martinez Bruno*
Institute of Mathematics and Computer Science
University of São Paulo
Av. Trabalhador São-Carlense,400, Centro CP 668
CEP 13560-970
São Carlos, São Paulo, Brazil
Tel: 55 16 33739671
E-mail: bruno@icmc.usp.br

*Corresponding author


Financial support: FAPESP - the State of São Paulo Research Funding Agency - Grant 03/04826-9, Brazil.

Keywords: computer vision, curvature, pattern recognition, shape analysis, shape signature and projection.

Abstract
Full Text

Among the diseases affecting the commercial citrus production, the citrus black spot (CBS) is considered to cause substantial losses. The analyses of particles in suspension in the orchards and collected into a disc have been applied as a preventive action trying to identify the presence of fungus spores before symptom appearance. In this paper, we show the results of several shape analysis methods applied to the fungus, the first step to the aimed computer aided vision system, capable to assist the identification process. Experiments and comparative results among the methods are presented in this paper, showing that better results were obtained applying the curvature method.

 
Supported by UNESCO / MIRCEN network