Spectral Modeling Method of Yellow River Main Stream Based on Autoencoder and Multilayer Perceptron Network

Abstract

The invention discloses a spectral modeling method for main ice of The Yellow River based on automatic coder and a multilayer perceptor network, wherein the method is used for settling a technical problem of low self-adaptability in an existing modeling method. The method is characterized in that many factors such as the absorption coefficient and the scattering coefficient of a water molecule, the absorption coefficient and the scattering coefficient of a slit particle, incident angle of light and upward scattering angle of water layer are comprehensively considered; and utilizing a neural network structure which is formed through connecting the automatic coder with a multilayer perceptor, studying by means of training data, and finally obtaining the spectral model of the main ice of The Yellow River. The automatic coder has characteristics of effective redundancy information elimination and high robustness to the noise, and therefore obtaining of robustness parameter estimation result is facilitated. According to the spectral model for the main ice of The Yellow River, reasons of the main ice are considered; the factors such as water body, split and illumination are comprehensively considered; and the automatic coder and the multilayer perceptor network are used in the model parameter estimating method; particular model parameters can be automatically acquired according to the data; and high expandability and high self-adaptability are realized.

Description

technical field [0001] The invention relates to a modeling method, in particular to a method for modeling the spectrum of the main stream of the Yellow River based on an automatic encoder and a multi-layer perceptron network. Background technique [0002] Multi-spectral remote sensing images can capture the reflection information in several spectral bands at the same time on the basis of capturing the spatial information of ground objects. In recent years, with the continuous development of remote sensing technology, multispectral remote sensing images have been widely used in the identification and extraction of water body regions, water quality detection, flood monitoring and other fields closely related to the national economy and people's livelihood. The main stream of the Yellow River is the stream with the largest flow velocity, the largest water momentum, the largest sediment concentration, and the deepest part of the riverbed in the Yellow River channel. According t...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the shortcomings of poor adaptability of existing modeling methods, the present invention provides a method for modeling the spectrum of the Yellow River Main Stream based on an autoencoder and a multi-layer perceptron network

Images