Leaf reflectance satellite remote sensing extraction method for eliminating influence of vegetation canopy structure and earth surface background

Abstract

The invention provides a leaf reflectance satellite remote sensing extraction method for eliminating influence of vegetation canopy structures and earth surface backgrounds and belongs to the field ofresearch on vegetation remote sensing retrieval parameter methods. The method comprises the following steps: by using a 4-Scale model, confirming visual probability (PT) of sun leaves corresponding to a remote sensing image pixel spectrum and visual probability (PG) of a lighting background; calculating an angle index (AI) of the spectrum, performing related analysis on PT and PG, and establishing an estimation model based on Al, PT and PG; by taking a leaf area index (LAI), PT and PT as three retrieval items, establishing a checking table (related to wavelengths) for multi-time scattering factors M in a simulation manner by using the 4-Scale model; finally, by using the established estimation model and the checking table, calculating average leaf reflectance of remote sensing image vegetation pixels. By adopting the method provided by the invention, reflection spectrums of leaves can be extracted through remote sensing images; compared with an optimal iterative computation method, the method is high in calculation efficiency, and compared with a table checking method with multiple steps of checking, the method is relatively simple in calculation process and relatively high in efficiency.

Description

1. Technical field [0001] The present invention uses the satellite remote sensing image data to quantitatively invert the visibility probability of the sun leaf and the illumination background of the pixel, describes the influence of the shadow component and the multiple scattering reflection with the multiple scattering reflection factor, and utilizes the theoretical development based on the 4-scale geometric optics model The method of extracting the average leaf reflectance of the pixel by the spectral downscaling model belongs to the research field of vegetation remote sensing inversion parameter method. 2. Background technology [0002] The pixel-scale reflection signal of satellite observation is affected by factors such as leaf reflectivity, canopy structure, and background reflectivity. Using satellite remote sensing data to quantitatively invert the biochemical parameters of vegetation leaves and canopies, it is first necessary to eliminate the vegetation canopy struc...

AI Technical Summary

Problems solved by technology

[0004] These two methods are based on the physical model of remote sensing physical mechanism, in which the calculation efficiency of the optimal iterative calculation method is low, and the operability is not strong in the application of satellite remote sensing imagery in large areas; the look-up table method performs a large number of simulation calculations in advance, A lookup table is established to simplify the calculation process and improve the efficiency of the inversion calculation, but it is required to input the canopy structure parameters of each pixel and the geometric angle of observation as the retrieval items, and multi-step search is required, the process is cumbersome, and the lookup table is There are limitations, even when other search items are fixed, PT, PG and LAI are not in a one-to-one mapping relationship, and the lookup table gives the value under the maximum possible probability

Images