LoboVault Home

Cramer-Rao bound analysis of multi-frame blind deconvolution


Please use this identifier to cite or link to this item: http://hdl.handle.net/1928/3273

Cramer-Rao bound analysis of multi-frame blind deconvolution

Show full item record

Title: Cramer-Rao bound analysis of multi-frame blind deconvolution
Author: Haji, Alim
Advisor(s): Osinski, Marek
Committee Member(s): Hayat, Majeed
Matson, Charles
Department: University of New Mexico. Dept. of Electrical and Computer Engineering
Subject: Imaging
Multi-frame blind deconvolution
cramer-rao bounds
LC Subject(s): Image reconstruction--Digital techniques.
Image processing--Digital techniques.
Degree Level: Masters
Abstract: This thesis explores how support constraints and multiple frames affect multi-frame blind deconvolution. Previous research in non-blind deconvolution, which seeks to estimate an object from a blurred and noisy image, characterized how the use of support constraints exploited spatial noise correlations to reduce noise in the estimate of the object. In multi-frame blind deconvolution, the blurring function is unknown and must be estimated along with the object. Applying a support constraint to both the object and the blurring functions, when using blind deconvolution, is one way to ensure a unique solution. The effects on the estimate of the object as a function of the size of the supports are analyzed. Also, the benefit in noise reduction in the estimate of the object from including multiple blurred and noisy images is considered. Cramer-Rao Bound theory is employed to provide an algorithm-independent metric to analyze the effects from these parameters. The Cramer-Rao bound is a lower limit to the variance of any estimate of an unknown parameter. In this research, the unknown parameters are the intensities of the object which is estimated.
Graduation Date: May 2007
URI: http://hdl.handle.net/1928/3273

Files in this item

Files Size Format View
Thesis printed.doc 1.844Mb Microsoft Word View/Open

This item appears in the following Collection(s)

Show full item record

UNM Libraries

Search LoboVault


My Account