CAPACITY DIMENSION ESTIMATING IN 3D

被引：0

作者：

Adamcik, Josef ^{[1
]}

Kukal, Jaromir ^{[1
]}

Nerad, Jakub ^{[1
]}

机构：

[1] Czech Tech Univ, Fac Nucl Sci & Phys Engn, Dept Software Engn Econ, Prague 12000 2, Czech Republic

来源：

16TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING MENDEL 2010 | 2010年

关键词：

capacity dimension; fractal analysis; point estimate; bias compensation; Alzheimer disease; 3D;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Minkowski-Bouligand (capacity) dimension can be estimated from the shape of 3D binary digital image. But the point estimate is biased typically to lower value. This effect was studied to obtain a relationship between the estimated and theoretical values of capacity dimension. Asymmetric Minkowski sausage and its volume were studied in the case of fBm 3D shapes generated from regular dodecahedron using random midpoint displacement technique. The resulting corrected estimate of capacity dimension was used for the study of Alzheimer diseased patients against controls from 3D SPECT images of human brain.

引用

页码：129 / 133

页数：5