A Method for Inverting Leaf Chlorophyll Content Based on New Spectral Features

Abstract

The invention discloses a method for retrieving the chlorophyll content of leaves based on new spectral features. The training data set of new spectral features (△h) is used as an independent variable for retrieving the chlorophyll content of leaves. Based on the parameters obtained in step 3, the set Solo algorithm performs dimensionality reduction processing on this part of the data; use the obtained dimensionality reduction data as the independent variable of the final regression model to perform multiple nonlinear regression, and establish a multiple nonlinear regression model with lasso algorithm constraints, that is, LASSO‑MNR inversion model, using the LASSO-MNR inversion model to invert the chlorophyll content of leaves. The present invention establishes a multiple nonlinear regression model (LASSO‑MNR) based on the lasso algorithm. The model is simple in principle, easy to operate, effectively reduces the redundancy of spectral data, improves inversion accuracy, and has greater flexibility It has strong universality and practicality at the same time.

Description

technical field [0001] The invention belongs to the technical field of inversion of trace element content in leaves of agricultural crops, and in particular relates to a method for inversion of chlorophyll content in leaves based on a combination of a lasso algorithm and a multiple nonlinear regression algorithm. Background technique [0002] The growth of crops such as winter wheat can be expressed by their physiological characteristics, and their physiological characteristics determine their changes in light absorption, transmission and reflection. Therefore, the difference in spectral reflectance data can be used to monitor the chlorophyll content of crops such as winter wheat. Hyperspectral remote sensing can quickly and accurately reflect the physiological and biochemical parameters of crops such as winter wheat, so spectral difference analysis based on hyperspectral has a wider application prospect in the field of vegetation remote sensing research. Hyperspectral data...

AI Technical Summary

Problems solved by technology

Among them, a single vegetation index discards a large amount of data and does not represent the entire hyperspectral information well, which may affect the inversion accuracy; using characteristic parameters such as red edge index and red edge area to invert chlorophyll content is also widely used. However, there are still some deficiencies in the generality of the model; in recent years, some related scholars have used methods such as neural networks to monitor the chlorophyll content of crops, but most of these methods have the influence of general applicability and poor interpretability.

At present, in the relevant research on hyperspectral data, there are few studies on using lasso algorithm to reduce dimensionality and then using multiple nonlinear regression to fit and invert leaf chlorophyll content.

Images