Modeling and visualizing the quantitative and allometric relationship of individual parts of rice by an integrated neural network

Abstract

To propose an artificial neural network (ANN) to establish quantitative relation and allometric relation of individual parts of rice, namely, stems, roots, leaf, and panicles. A novel quantitative model for calculating the amount of light incident to the leaves is also introduced. The amount of light incident to the leaves can be calculated by applying radiosity rendering technique to a three dimensional architectural model of rice which is constructed by L-system. Computations of light incidence will be used as one of the environment parameters in the artificial neural network process for predicting rice biomass.