基于分形理论的遥感影像分类研究

分形理论是非线性科学领域的一大支柱,它的提出为人们解决非线性世界的问题提供了新的思想和方法,由于分形维数与人类对图像表面纹理粗糙度的感知是一致的,利用分形理论进行遥感影像分类具有潜在的理论和应用价值。本文根据遥感影像特点,引入分形方法来描述其纹理结构特征,利用双毯覆盖模型来提取遥感影像的纹理特征,在此基础上,采用K-means算法将纹理结构特征与光谱特征相结合,进行遥感影像分类,其分类效果优于单纯采用光谱特征的分类。

Study on Classification of Remote Sensing Image Based on Fractal Theory

Fractal theory is the pillar of nonlinear science,which has been proposed for people to solve the nonlinear problems of the world,provides the new ideas and methods.Because the fractal dimension and human perception to the image texture roughness is consistent,so fractal theory has the potential for remote sensing image classification theory and application.This article based on the characteristics of remote sensing image introduced how to describe and capture the texture features of ground objects by Doubl...