Extração de Medidas Biométricas em Nervuras de Órgãos Foliares Vegetais

Rodrigo de Oliveira PlotzeOdemir Martinez Bruno

Computer vision techniques utilized for extracting biometric features in plant leaf structures are described in this paper. The collected biometric data are concentrated in leaf venation system. The methods are applied in real images of plants in two species sets: (i) passion fruits species of genus Passiflora and (ii) vegetable species of Mata Atlântica and Cerrado. The biometric features extracted were introduced in a neural network to perform the classification of the species. The results exhibited a correctly species discrimination more than 90%. The developed methods can be used to assists the taxonomists to perform biometry in plant leaf structures.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: