DocumentCode :
3491649
Title :
Analysis of optimal estimation algorithms efficiency of anisotropic textures orientation
Author :
Gruzman, I.S.
fYear :
2012
fDate :
2-4 Oct. 2012
Firstpage :
19
Lastpage :
21
Abstract :
Based on the Cramer-Rao inequality analysis of the potential accuracy of the estimate of anisotropic texture orientation observed on the background of additive white noise. The brightness function of the anisotropic texture with one dominant direction was approximated by a deterministic two-dimensional harmonic function. A method for analyzing the efficiency of optimal algorithms for estimating the orientation of anisotropic texture, allowing to determine the ultimate capabilities of methods based on the gradient structure tensor. The method is based on the Gaussian approximation for the components of the gradient structure tensor calculated in the frequency domain.
Keywords :
AWGN; approximation theory; brightness; frequency-domain analysis; gradient methods; image texture; tensors; Cramer-Rao inequality analysis; Gaussian approximation; additive white noise background; anisotropic textures orientation estimation; brightness function; deterministic 2D harmonic function; frequency domain; gradient structure tensor; optimal estimation algorithms efficiency; Accuracy; Additive white noise; Algorithm design and analysis; Approximation algorithms; Cramer-Rao bounds; Tensile stress;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Actual Problems of Electronics Instrument Engineering (APEIE), 2012 11th International Conference on
Conference_Location :
Novosibirsk
Print_ISBN :
978-1-4673-2842-5
Type :
conf
DOI :
10.1109/APEIE.2012.6628985
Filename :
6628985
Link To Document :
بازگشت