Red edge spectral feature-based method and system for remote sensing quantitative inversion of chlorophyll a

Abstract

The application discloses a red edge spectral feature-based remote sensing quantitative inversion method and system for chlorophyll a. A water quality proxy variable reflecting water body optical types is constructed. On the basis of performing strict non-negative bottom line constraint on red light band reflectivity, adaptive fusion of red light absorption valley and red edge reflection peak features is realized through the variable and continuous weight function with specific first derivative constraint, and the fusion process does not need manual intervention weight parameters. On this basis, based on the fusion obtained feature fusion index and measured chlorophyll a concentration data, a mathematical regression model with first derivative greater than zero is used to establish a quantitative mapping relationship, and then the spatial distribution of chlorophyll a concentration of the target water body pixel is obtained. The application effectively solves the physical contradiction of weak low-concentration signal and high-concentration saturation, and completely eliminates the spatial step fault from the mathematical mechanism, and improves the precision and spatial consistency of chlorophyll a remote sensing inversion in complex inland water areas.

Images