Bare soil change identification method combining multiple remote sensing indexes

Abstract

The invention relates to a bare soil change identification method combining multiple remote sensing indexes. The method comprises the following steps: S1, building a vegetation near-infrared reflection index time series data set; S2, establishing a bare soil index time series data set; S3, establishing a characteristic index time series data set; S4, evaluating and selecting characteristic indexes; S5, calculating the change trend of the characteristic indexes; S6, establishing a bare soil change identification scheme; and S7, identifying the bare soil change of the research area to obtain a bare soil change spatial distribution map of the research area: judging whether the bare soil is changed pixel by pixel according to the bare soil change identification scheme established in step S6, and finally generating a bare soil change identification result map of the research area. The method is suitable for large-range bare soil change monitoring, the space-time continuity is high, bare soil does not need to be extracted in advance, and a long-time sequence bare soil change spatial distribution map can be rapidly obtained.

Description

technical field [0001] The invention relates to the technical field of time-series remote sensing change monitoring, in particular to a bare soil change identification method combined with multiple remote sensing indexes. Background technique [0002] Soil erosion is a severe ecological environment problem facing the world today, and various disasters induced by it have also brought threats and damage to people's production and life, and bare soil is the main surface landscape in soil erosion areas. In this regard, it is especially important to quickly and accurately locate the bare soil area and analyze the temporal and spatial changes of the bare soil on the surface for the control of soil erosion. In recent years, remote sensing technology has become an indispensable high-tech for soil erosion control and evaluation. In this regard, many scholars use remote sensing technology to study the change of bare soil, and the main methods include three categories: (1) identify th...

AI Technical Summary

Problems solved by technology

[0003] At present, many scholars have studied the identification method of bare soil change, but there are still the following deficiencies: 1. The method based on manual visual interpretation has a large workload, and it is difficult to obtain large-scale bare soil information.

2. Based on the method of supervised classification, it is difficult to select training sample areas, while the accuracy of classification results obtained by unsupervised classification methods is low

3. Due to the heterogeneity within the class, this method is not suitable for large-scale bare soil identification; 4. When analyzing the spatio-temporal evolution of bare soil, the comparison method after classification will bring large errors

Images