Maximum entropy image restoration revisited

Author

Willis, Matthew ; Jeffs, B.D. ; Long, David G.

Author_Institution

Dept. of Electr. & Comput. Eng., Brigham Young Univ., Provo, UT, USA

Volume

1

fYear

2000

fDate

2000

Firstpage

89

Abstract

This paper presents a new non-iterative, closed-form approximation to the maximum entropy (ME) image restoration method. A fast frequency domain implementation of this closed form approach is developed for the case of circular convolutional blur. This result dramatically reduces the computational demands compared to conventional iterative ME algorithms such as MART. Some limitations and advantages of ME restoration are investigated, including its dismal performance for high resolution restoration of decimated or randomly sampled blurred observations

Keywords

approximation theory; convolution; frequency-domain analysis; image resolution; image restoration; image sampling; iterative methods; maximum entropy methods; MART; circular convolutional blur; computational demands reduction; decimated blurred observations; fast frequency domain implementation; high resolution restoration; maximum entropy image restoration; noniterative closed-form approximation; performance; randomly sampled blurred observations; Constraint optimization; Entropy; Equations; Frequency domain analysis; Image restoration; Iterative algorithms; Least squares approximation; Least squares methods; Null space; Radio astronomy;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing, 2000. Proceedings. 2000 International Conference on

Conference_Location

Vancouver, BC

ISSN

1522-4880

Print_ISBN

0-7803-6297-7

Type

conf

DOI

10.1109/ICIP.2000.900899

Filename

900899