An Atmospheric Correction Method for Class II Water Body Based on Neural Network Quadratic Optimization

Abstract

The invention discloses a class-II water atmospheric correction method based on neural network quadratic optimization and relates to the technical field of remote sensing image data processing. The method comprises the steps of extracting geometric information and wavelength information of an acquired hyperspectral image and an initial optical thickness value T(550)<0> of aerosol at 550nm; inputting the extracted parameters into a neutral network model, and outputting atmospheric transmittance t and total contribution Rho(path) in a simulation manner; simulating apparent reflectance Rho(toa)<sim> according to off-water reflectance Rho(w)(NIR) of a near infrared band together with the t and the Rho(path); performing spectral optimization on the Rho(toa)<sim> and real apparent reflectance Rho(toa)<mes> extracted from the image to finally obtain optimal solutions T(550)<pt>, R<opt> and n<opt>; inputting the T(550)<pt> into the neutral network model to obtain atmospheric diffuse transmittance t<opt> of all hyperspectral bands and the total contribution Rho(path)<opt> of atmospheric molecules and the aersol; estimating the off-water reflectance of the hyperspectral image according to the real apparent reflectance Rho(toa)<mes> of the image. According to the class-II water atmospheric correction method based on the neural network quadratic optimization, the practicability of the model is improved, the input parameters are reduced, and the estimation precision is improved.

Description

technical field [0001] The invention relates to the technical field of remote sensing image data processing, in particular to a second-class water body atmospheric correction method based on neural network secondary optimization. Background technique [0002] Atmospheric correction has always been an important issue in data preprocessing in water color remote sensing. Morel & Prieur (1977) divided seawater into first-class water bodies and second-class water bodies. Most of the first-class water bodies are open ocean water bodies that are far away from land and less affected by human activities. The chlorophyll in the water plays a decisive role in its optical properties; The characteristics are determined by colored soluble organic matter, suspended inorganic matter and chlorophyll, which are complex and changeable. Gordon, et al. (1994) developed a standard based on dark pixels, assuming that all the signal values ​​in the near-infrared band come from Rayleigh scattering...

AI Technical Summary

Problems solved by technology

However, these methods all rely on the selection of clean pixels, and for turbid second-class water bodies, overcorrection often occurs

Images