Forest type recognition method based on high-resolution remote sensing images

Abstract

The invention relates to a forest type recognition method based on high-resolution remote sensing images, including: preprocessing the high-resolution remote sensing images to obtain panchromatic spectral images and multi-spectral images; The vegetation index is extracted from the spectral image; combined with the second-class survey data of forest resources, the dominant tree species are analyzed, and the gray value and vegetation index value of the dominant tree species are obtained; value and vegetation index value, respectively establish the relevant forest type identification function model; test and evaluate the correlation function model; apply the correlation function model to forest type identification. The forest type can be identified by using the texture features of the Gaofen-2 image to obtain the gray level and vegetation index, which reasonably reduces the field investigation of forest resources and saves costs and resources.

Description

technical field [0001] The invention relates to a method for identifying forest types based on high-resolution remote sensing images. Background technique [0002] Forest type identification is an important part of forest resources management and monitoring. Traditional forest type identification methods are relatively backward, relying on a large number of manual field identification, which not only has a lot of work intensity, poor timeliness, inaccurate monitoring results, errors and timeliness, but also consumes a lot of manpower, material resources and Financial resources, and did not meet the expected requirements. As time changes, the state of forest resources will show dynamic changes, and traditional methods cannot accurately grasp the changes of forest types. [0003] With the successful launch of my country's domestic high-resolution series of remote sensing satellites, high-resolution remote sensing images have been widely used in earth observation. The textur...

AI Technical Summary

Problems solved by technology

Traditional forest type identification methods are relatively backward, relying on a large number of manual field identification, which not only has a lot of work intensity, poor timeliness, inaccurate monitoring results, errors and timeliness, but also consumes a lot of manpower, material resources and financial resources, and did not meet the expected requirements

As time changes, the state of forest resources will show dynamic changes, and traditional methods cannot accurately grasp the changes in forest types

Images