Method for extracting winter wheat planting area based on spectrum reconstruction technology

Abstract

The invention discloses a method for extracting a winter wheat planting area based on spectrum reconstruction technology. The method comprises the following steps: step 1, acquiring a multi-scene high-quality remote sensing image of the whole winter wheat growth period in a research area and preprocessing a multi-scene high-quality remote sensing image; step 2, calculating a normalized vegetationindex of each image, and obtaining NDVI time sequence curve data of the remote sensing images in the research area through band synthesis, SG filtering and other processing; step 3, selecting a part of winter wheat sample points and obtaining NDVI time sequence curve data of the winter wheat sample points as a winter wheat training data set; step 4, performing singular vector decomposition on thewinter wheat training data set to obtain a plurality of front singular components of the winter wheat NDVI time sequence curve; step 5, reconstructing the NDVI time sequence curve of the whole image pixel by pixel by utilizing the obtained singular components; and step 6, comparing the similarity between the reconstructed NDVI time series data of the whole image and the original NDVI time series data. The phenological law of winter wheat growth is used for extracting the planting area, the principle is simple, operation is convenient, and high universality is achieved.

Description

technical field [0001] The invention belongs to the technical field of remote sensing extraction of crop planting area, in particular to a method for extracting winter wheat planting area based on spectral reconstruction technology. Background technique [0002] Winter wheat is one of the main grain crops in my country, and its sown area and output both occupy a pivotal position in the national grain production. In recent years, with the in-depth adjustment of agricultural structure and the development of agricultural market economy, great changes have taken place in the rural planting structure, and the development of winter wheat production is facing new challenges. Timely and accurate access to winter wheat planting information is of great significance for winter wheat production management, growth monitoring, yield estimation, and ensuring national food security. [0003] At present, remote sensing methods for crop planting area extraction can be divided into three cate...

AI Technical Summary

Problems solved by technology

In recent years, relevant scholars have also conducted some research on remote sensing image classification based on spectral similarity (such as spectral correlation (SCM), spectral angle (SAM), spectral information divergence (SID), etc.), but these methods are mostly used for Land cover type change monitoring or hyperspectral remote sensing image classification, although a small number of scholars also apply it to the extraction of crop planting area, most of them describe the difference between the original spectral curve of the image target pixel and the existing reference spectral curve of the spectral library. There are few studies on calculating the similarity between pixels and pixels to extract crop planting area.

Images