Bank of Wiener filters for adaptive image restoration

Sung Jea Ko, Yong Hoon Lee, Adly T. Fam

Research output: Contribution to conferencePaper

Abstract

Summary form only given, as follows. The authors introduce a new type of bank of Wiener filters, based on a nonstationary multiple image model, for the restoration of images degraded by signal-independent additive noise or signal-dependent noise. In the restoration algorithm, the local-minimum mean-square-error (LMMSE) decision rule is used to choose the best output in the MSE sense at each pixel among the outputs of a bank of Wiener filters. This filtering algorithm is shown to be very effective in suppressing noise while preserving signal characteristics such as edges, due to its locally adaptive structure. Finally, the generalized homomorphic transformation to make signal-dependent noise independent of the signal is combined with the bank of Wiener filters technique to process images degraded by signal-dependent noise.

Original languageEnglish
Pages238
Number of pages1
Publication statusPublished - 1988
Externally publishedYes

ASJC Scopus subject areas

  • Engineering(all)

Fingerprint Dive into the research topics of 'Bank of Wiener filters for adaptive image restoration'. Together they form a unique fingerprint.

  • Cite this