Optical logarithmic filtering using inherent film nonlinearity
| Autor: | Thomas Cheng, Francis T. S. Yu, Anthony M. Tai |
|---|---|
| Rok vydání: | 2010 |
| Předmět: |
Signal processing
Halftone Logarithm Spatial filter Computer science business.industry Materials Science (miscellaneous) Filter (signal processing) Industrial and Manufacturing Engineering Multiplicative noise Convolution Transformation (function) Optics Detection theory Business and International Management Optical filter business Image resolution Linear filter |
| Zdroj: | Applied optics. 16(9) |
| ISSN: | 1559-128X |
| Popis: | Linear optical spatial filtering cannot be effectively applied to multiplied and convolved signals. One approach is to first perform a logarithmic transformation to produce a signal in additive form suitable for linear filtering processes [ A. V. Oppenheim , Proc. IEEE56, 1264 ( 1968)]. It was first suggested that a half-tone screen be used to perform such a transformation [ H. Kato J. W. Goodman , Appl. Opt.14, 1813 ( 1975)]. However, the maximum spatial resolution of this halftone screen technique is limited by the resolution of the screen. In this paper, we shall propose a different technique using the inherent film nonlinearity for the logarithmic transformation. Such a technique would enable the transformation of signals of very high spatial resolution, limited only by the resolution limit of the photographic film. This technique is applied to the spatial filtering and detection of signals in multiplicative noise. Experimental comparisons between linear and logarithmic filtering are presented. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|