摘要
本文基于分形几何中的分维值对影像纹理进行分类。系统地介绍了以影像分维值为依据,采用多分辨率分维值,以及本文提出的利用已有空隙(Lacunarity)概念的拟合回归法,对影像纹理进行分类的全过程。试验表明该方法对标准纹理和航空影像纹理的分类都能取得较好的效果。
This paper based on the fractal dimension to classify image textures; We systematically introduce the whole process of image texture classification,the method relys on the fractal di-mension of image. and with the help of multiresolution fractal and the regression of lacunarity’s value to treat images which have the same fractal dimension but belonging to dif-ferent kinds of textures. Whether standard texture or aerial image texture can obtain very good results with this method.
出处
《测绘学报》
EI
CSCD
北大核心
1995年第4期283-291,共9页
Acta Geodaetica et Cartographica Sinica
基金
国家测绘基金
关键词
分维
多分辨率
影像纹理
分类
分形几何
Fractal dimension,Multiresolution fractal dimension,Lacunarity,Image texture classification