Vegetation identification method and application

Abstract

The invention belongs to the technical field of environmental protection, and particularly relates to a vegetation identification method and application. Due to the fact that mature and young palm trees and other vegetations are difficult to distinguish accurately by using optical and SAR data, and an existing classification algorithm has strong dependence on values of input parameters when monitoring and recognizing the vegetations. The vegetation identification method is applied to identification of a palm plantation, and comprises the following steps: fusing a first satellite optical remote sensing image and a second satellite synthetic aperture radar image to obtain fused image data; extracting an information feature combination from the fused data image; calculating weight values of the features, and selecting an optimal subset as a model training set; optimizing a random forest algorithm to obtain an improved random forest algorithm, and training a vegetation identification and classification model by adopting the improved random forest algorithm and the model training set; and verifying the vegetation identification classification model to obtain an identification result. The identification efficiency can be remarkably improved.

Description

technical field [0001] The application belongs to the technical field of environmental protection, and in particular relates to a vegetation identification method and its application. Background technique [0002] At present, many studies have used satellite remote sensing data to identify vegetation. Optical-based methods rely on information extracted from phenological or image features of vegetation, including phenological methods and image recognition methods. Phenology-based methods mainly use temporal changes in vegetation spectra to monitor the expansion of vegetation such as palm trees. In order to solve the identification and monitoring problems of vegetation such as palm plantations in tropical areas, the all-weather global observation Synthetic Aperture Radar (SAR) has become the object of attention. These studies used radar satellite data (L-band in ALOS PALSAR and C-band in Sentinel-1) as the main basis for identifying vegetation. Using radar satellite data, d...

AI Technical Summary

Problems solved by technology

[0005] Based on the fact that it is difficult to accurately distinguish between mature and young palm trees and other vegetation using optical remote sensing and SAR data, and the existing classification algorithms have a strong dependence on the value of input parameters when monitoring and identifying vegetation. , the application provides a vegetation recognition method and its application

Images