Matches in SemOpenAlex for { <https://semopenalex.org/work/W1990852007> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W1990852007 abstract "Restoration of Noisy Images With Adaptive Windowing and Nonlinear Filtering*Woo -Jin Song and William A. PearlmanRensselaer Polytechnic InstituteElectrical, Computer, and Systems Engineering DepartmentTroy, New York 12180AbstractA general problem with statistically -based estimators for images degraded by additivenoise is their dependence on average quatities when image intensities vary rapidly andwidely. The Wiener estimator, for example, given the stationary power spectrum on the ob-ject image and the noise, is known to produce a noisy effect in the flat intensity regionsand a blurring or fuzzy effect in the edge regions on the restored image. The power spectraare usually estimated over regions containing both edges and flat regions and therefore arenot truly representative of either regional type. In this work, we accept a nonstationaryimage model and utilize a novel adaptive windowing technique in conjunction with a nonlinearestimator to overcome the cited defects of other estimators. This technique is appliedsuccessively to simulated noisy one -dimensional feature waveforms, an arbitrarily selectednoisy image scan line,noisy images with one -dimensional windowing, and noisy images withtwo -dimensional windowing. In each case, features of the edge and flat regions both arefaithfully reconstructed. In fact, the restored images are remarkably sharp and clean.They appear far superior to the comparable Wiener restorations despite the fact that theirmean -squared error is about the same or slightly larger.IntroductionParameter or spectral estimates required for image estimators are usually derived fromfixed regions of the image. The object image, however, is nonstationary or, at best, quasi -or locally stationary. Hence an image estimate based on such regional averages can not fol-low rapid statistical changes within a region. For example, the Wiener (minimum mean -squarelinear) estimator requires the image power spectrum, which averages quantities belonging tobusy and flat contrast areas of the image. The effect of averaging between these two re-gional types is manifested in restorations which are noisy in the flat areas due to insuf-ficient smoothing and fuzzy in the edge regions due to oversmoothing. If the analysis win -dow for parameter or spectral estimates contains elements truly representative of the sta-tistical class of the current picture element, then the desired estimate of the true objectintensity at that point would be obtained. Here we shall present a scheme for adaptivelychanging the analysis window so that it embraces elementsof the current pixel's statistical" @default.
- W1990852007 created "2016-06-24" @default.
- W1990852007 creator A5005719536 @default.
- W1990852007 creator A5028597605 @default.
- W1990852007 date "1986-11-20" @default.
- W1990852007 modified "2023-09-23" @default.
- W1990852007 title "Restoration Of Noisy Images With Adaptive Windowing And Nonlinear Filtering" @default.
- W1990852007 doi "https://doi.org/10.1117/12.937267" @default.
- W1990852007 hasPublicationYear "1986" @default.
- W1990852007 type Work @default.
- W1990852007 sameAs 1990852007 @default.
- W1990852007 citedByCount "6" @default.
- W1990852007 crossrefType "proceedings-article" @default.
- W1990852007 hasAuthorship W1990852007A5005719536 @default.
- W1990852007 hasAuthorship W1990852007A5028597605 @default.
- W1990852007 hasConcept C105795698 @default.
- W1990852007 hasConcept C106430172 @default.
- W1990852007 hasConcept C11413529 @default.
- W1990852007 hasConcept C115961682 @default.
- W1990852007 hasConcept C138885662 @default.
- W1990852007 hasConcept C153180895 @default.
- W1990852007 hasConcept C154945302 @default.
- W1990852007 hasConcept C174576160 @default.
- W1990852007 hasConcept C18537770 @default.
- W1990852007 hasConcept C185429906 @default.
- W1990852007 hasConcept C2776401178 @default.
- W1990852007 hasConcept C31972630 @default.
- W1990852007 hasConcept C33923547 @default.
- W1990852007 hasConcept C41008148 @default.
- W1990852007 hasConcept C41895202 @default.
- W1990852007 hasConcept C90652560 @default.
- W1990852007 hasConcept C9417928 @default.
- W1990852007 hasConcept C99498987 @default.
- W1990852007 hasConceptScore W1990852007C105795698 @default.
- W1990852007 hasConceptScore W1990852007C106430172 @default.
- W1990852007 hasConceptScore W1990852007C11413529 @default.
- W1990852007 hasConceptScore W1990852007C115961682 @default.
- W1990852007 hasConceptScore W1990852007C138885662 @default.
- W1990852007 hasConceptScore W1990852007C153180895 @default.
- W1990852007 hasConceptScore W1990852007C154945302 @default.
- W1990852007 hasConceptScore W1990852007C174576160 @default.
- W1990852007 hasConceptScore W1990852007C18537770 @default.
- W1990852007 hasConceptScore W1990852007C185429906 @default.
- W1990852007 hasConceptScore W1990852007C2776401178 @default.
- W1990852007 hasConceptScore W1990852007C31972630 @default.
- W1990852007 hasConceptScore W1990852007C33923547 @default.
- W1990852007 hasConceptScore W1990852007C41008148 @default.
- W1990852007 hasConceptScore W1990852007C41895202 @default.
- W1990852007 hasConceptScore W1990852007C90652560 @default.
- W1990852007 hasConceptScore W1990852007C9417928 @default.
- W1990852007 hasConceptScore W1990852007C99498987 @default.
- W1990852007 hasLocation W19908520071 @default.
- W1990852007 hasOpenAccess W1990852007 @default.
- W1990852007 hasPrimaryLocation W19908520071 @default.
- W1990852007 hasRelatedWork W1493054196 @default.
- W1990852007 hasRelatedWork W1536088005 @default.
- W1990852007 hasRelatedWork W1979141523 @default.
- W1990852007 hasRelatedWork W200823940 @default.
- W1990852007 hasRelatedWork W2023108657 @default.
- W1990852007 hasRelatedWork W2023631307 @default.
- W1990852007 hasRelatedWork W2045956739 @default.
- W1990852007 hasRelatedWork W2049909233 @default.
- W1990852007 hasRelatedWork W2070107952 @default.
- W1990852007 hasRelatedWork W2074769753 @default.
- W1990852007 hasRelatedWork W2076749578 @default.
- W1990852007 hasRelatedWork W2104763670 @default.
- W1990852007 hasRelatedWork W2122322733 @default.
- W1990852007 hasRelatedWork W2146981757 @default.
- W1990852007 hasRelatedWork W2165572139 @default.
- W1990852007 hasRelatedWork W2171222618 @default.
- W1990852007 hasRelatedWork W2891679023 @default.
- W1990852007 hasRelatedWork W2949423816 @default.
- W1990852007 hasRelatedWork W3106505466 @default.
- W1990852007 hasRelatedWork W2088740075 @default.
- W1990852007 isParatext "false" @default.
- W1990852007 isRetracted "false" @default.
- W1990852007 magId "1990852007" @default.
- W1990852007 workType "article" @default.