BDBComp
Parceria:
SBC
Classificação de Cores por Redes Neurais Artificiais: Um Estudo do Uso de Diferentes Sistemas de Representação de Cores no Futebol de Robôs Móveis Autônomos

Alexandre da Silva SimõesAnna Helena Reali Costa

In previous works we have investigated the influence of the use of different color representations in the performance of a multilayer perceptron Artificial Neural Network used for color image segmentation. The RGB, HSV and YUV color representations were tested in tasks of color-based segmentation, aiming at robustness to variations in the illumination of the environment. In the present work we verify the usefulness of this approach applied to the robotic soccer domain. Our goal consists in defining the most suitable color representation to this domain, since it is a real-time domain that presents serious restrictions about the calibration time. Clique no link abaixo para buscar o texto completo deste trabalho na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD