GA-Neural Network Based Estimation of Materials Concentration in Polluted Water Lake Using Spectral Radiance Reflectance

Abstract

This paper presents a new model based on Neural Network(NN) for estimating distribution of materials concentration in polluted water lake. A spectral radiance reflectance method has been employed to estimate the materials concentration below the water surface. Lake Kasumigaura, Japan, has been selected as a target area in our research observation, and the lake was polluted by rich chlorophyll-a (chl.a) concentration and suspended solid (SS). The essence of our research is to scrutinize a set of wavelength effective to the estimation using NN model and employing Genetic Algorithm (GA). The obtained results show that proposed method is applicable to use in a remote-sensing image method. The results demonstrate that our NN model is capable to provide a pragmatic approach to accommodate existing and future demands for defining the quality of water and environmental monitoring.