A Forest Monitoring Method of Time Series of Remote Sensing Vegetation Index

Abstract

A remote sensing vegetation index time series forest monitoring method, comprising: acquiring landsat8 remote sensing vegetation index (NDVI) time series data, and generating a time series data set spanning several years with daily time steps in a study area; based on the quality of the remote sensing data in landsat8 file, identify and eliminate noise data by using cloud and fog quality and time series difference method; establish the initial remote sensing vegetation index time series fitting diagram based on Gaussian equation for each valid point; The difference between the curve fitting values ​​is weighted, and the Gaussian model is iteratively calculated until the residual reaches the target threshold or the number of iterations reaches the upper limit, and the final vegetation index time series fitting graph is obtained; the forest annual monitoring index is extracted from the vegetation index time series fitting curve.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image information processing, and relates to a remote sensing vegetation index time series forest monitoring method. Background technique [0002] The surface information at a certain moment captured by remote sensing satellites is recorded as remote sensing images, and multiple images of the same area at different times are remote sensing time-series images. Remote sensing time series data has very important research value, for example, it can be used to study the relationship between vegetation and phenology, changes in vegetation, and monitor forest growth. However, affected by multiple factors such as light intensity, atmospheric humidity, cloud thickness, and weather, multi-source noise interference leads to a decline in the quality of remote sensing data, and it is difficult to accurately extract surface information, which has caused great difficulties in remote sensing research. [...

AI Technical Summary

Problems solved by technology

The Fourier filtering method expresses the time curve as a series of linear superposition of cosine waves, and the fitting results are reliable by screening important band information, but this method is not adaptable enough to the situation where the effective time series is sparse

Gaussian filtering method The nonlinear Gaussian function is fitted by the least square method, which overcomes subjectivity and has obvious advantages in information extraction in the long-term range, but this method is sensitive to the noise of the original data points

Images