SAVR法──影像表面分维估计的新方法

陶闯; 林宗坚

doi:10.11834/jrs.1994027

浏览量 : 0 下载量: 220 CSCD: 4

PDF
导出
分享
收藏
专辑

SAVR法──影像表面分维估计的新方法
SAVR Method: A New Technique to Estimate the Fractal Dimension of Intensity Surfaces
1994年第3期
纸质出版：1994
DOI： 10.11834/jrs.1994027
稿件说明：

移动端阅览

[1]陶闯,林宗坚.SAVR法──影像表面分维估计的新方法[J].环境遥感,1994(03):203-208. DOI： 10.11834/jrs.1994027.

SAVR Method: A New Technique to Estimate the Fractal Dimension of Intensity Surfaces[J]. Journal of Remote Sensing, 1994, (3). DOI： 10.11834/jrs.1994027.

摘要

本文在分形表面积与体积关系的基础上，导出了一种估计影像表面分维数的新方法──ＳＡＶＲ法。通过对合成的已知维数分形影像数据的测试，表明新方法在分维估值准确性和无标度区特性方面表现较优，进一步与ＢＬＡＮＫＥＴ法分维估计实验比较得出，ＳＡＶＲ法具有较强的保持多分辨影像分维估值的稳定性。最后提出了基于ＳＡＶＲ法方向分维的纹理特征度量并给出了在方向纹理分析中的应用实例。

Abstract

Based on the fractal relation between the superficial area and volume

a new method called the SAVR Method is derived for 3-D surfaces.The compara-tive studies and experiments show the new method behaves well both in the esti-mating accuracy of fractal dimension and the characteristics of limit scales by testingthe synthesized the fractal images with the known dimensions.Especially

the SAVRmethod could obtain the stable estimates of intensity surfaces with varing the diffe-rent resolutions.At the last

we proposed a kind of texture feature measures basedon the SAVR directional fractal dimension for texture analysis

and the experimentsin boundary detection of texture image are also present.