Evaluation of Flower by Neural Network

  • Ikeda, Y. (Department of Agricultural Engineering Faculty of Agriculture Kyoto University) ;
  • Sawada, T. (Department of Agricultural Engineering Faculty of Agriculture Kyoto University)
  • Published : 1993.10.01

Abstract

The color image of the rose was segmented by the cluster analysis on the color space into the characteristic sub-regions, the degree of bloom of the flower was represented numerically base on the segmented image and judged by the artificial neural network system whose input variable were the characteristic regions. Judgement by neural system were compared with that of the farmers and it was found that degree of agreement were fairly good.

Keywords