Showing papers on "Salt-and-pepper noise published in 1991"
TL;DR: The center weighted median (CWM) filter as discussed by the authors is a weighted median filter that gives more weight only to the central value of each window, which can preserve image details while suppressing additive white and/or impulsive-type noise.
Abstract: The center weighted median (CWM) filter, which is a weighted median filter giving more weight only to the central value of each window, is studied. This filter can preserve image details while suppressing additive white and/or impulsive-type noise. The statistical properties of the CWM filter are analyzed. It is shown that the CWM filter can outperform the median filter. Some relationships between CWM and other median-type filters, such as the Winsorizing smoother and the multistage median filter, are derived. In an attempt to improve the performance of CWM filters, an adaptive CWM (ACWM) filter having a space varying central weight is proposed. It is shown that the ACWM filter is an excellent detail preserving smoother that can suppress signal-dependent noise as well as signal-independent noise. >
1,071 citations
TL;DR: The results indicate that the colored noise Kalman filters provide a significant gain in signal-to- noise ratio (SNR), a visible improvement in the sound spectrogram, and an audible improvement in output speech quality, none of which are available with white-noise-assumption Kalman and Wiener filters.
Abstract: Scalar and vector Kalman filters are implemented for filtering speech contaminated by additive white noise or colored noise, and an iterative signal and parameter estimator which can be used for both noise types is presented. Particular emphasis is placed on the removal of colored noise, such as helicopter noise, by using state-of-the-art colored-noise-assumption Kalman filters. The results indicate that the colored noise Kalman filters provide a significant gain in signal-to-noise ratio (SNR), a visible improvement in the sound spectrogram, and an audible improvement in output speech quality, none of which are available with white-noise-assumption Kalman and Wiener filters. When the filter is used as a prefilter for linear predictive coding, the coded output speech quality and intelligibility are enhanced in comparison to direct coding of the noisy speech. >
302 citations
TL;DR: A new idea, enhancing speech based on auditory evidence, is explored for the problem of enhancing speech degraded by stationary and nonstationary additive white noise; a significant reduction of such noise and an improvement in speech quality are achieved.
Abstract: A new idea, enhancing speech based on auditory evidence, is explored for the problem of enhancing speech degraded by stationary and nonstationary additive white noise. Distinguishing different objectives for heavy and light noise interference, two related algorithms are developed. For speech degraded by heavy noise, the improvement in signal-to-noise ratio (SNR) is as high as 12 dB; for lightly noisy speech, the improvement is modest and decreases as the SNR of the noisy speech increases. Quantizing noise is used to assess the capacity for reducing nonstationary noise using these algorithms; a significant reduction of such noise and an improvement in speech quality are achieved. The advantages of the proposed algorithms for speech enhancement include no need for prior knowledge of the noise and only a modest computational requirement. >
101 citations
TL;DR: Results are presented on restoring photographic blurred images using the proposedfilter in the exposure domain, whereas the use of the classical Wiener filter in the density domain, with additive noise assumption, does not yield any visible improvement.
Abstract: A linear minimum mean-square-error deconvolution filter in the presence of multiplicative noise is derived. The importance of incorporating the nonlinear sensor characteristics into the restoration of noisy and blurred scanned photographic images is discussed. It is proposed to restore images in the 'exposure domain' where a linear convolutional relationship between the original and the observed images can be established. Results are presented on restoring photographic blurred images using the proposed filter in the exposure domain, whereas the use of the classical Wiener filter in the density domain, with additive noise assumption, does not yield any visible improvement. >
51 citations
TL;DR: In this paper, continuous-time integrated filters with a maximum dynamic range at the audio-frequency range were obtained using MOSFET-C filters in combination with a well-designed BiCMOS amplifier, where the detrimental influence of 1/f noise was eliminated and the filter noise was completely determined by the unavoidable thermal noise of the filter resistors.
Abstract: Continuous-time integrated filters having a maximum dynamic range at the audio-frequency range are obtained using MOSFET-C filters. Using MOSFET-C filters in combination with a well-designed BiCMOS amplifier, the detrimental influence of 1/f noise can practically be eliminated and the filter noise is completely determined by the unavoidable thermal noise of the filter resistors. A 98-dB dynamic range (total harmonic distortion >
31 citations
TL;DR: An optical signal-processing technique for additive noise reduction that uses the noisy signal and a Gaussian reference beam to produce an adaptive Wiener filter is presented.
Abstract: We present an optical signal-processing technique for additive noise reduction that uses the noisy signal and a Gaussian reference beam to produce an adaptive Wiener filter. We experimentally demonstrate an improvement from 1 to 8 in the signal-to-noise ratio by using nonlinear gain in two-beam coupling in barium titanate to transmit 50% of the signal and 6% of the noise.
30 citations
TL;DR: In this article, the optimal trade-off inplane rotation invariant filters for noise robustness and sharpness of the correlation peaks are determined, and it is demonstrated that the classical harmonic filters are overspecialized.
29 citations
TL;DR: The algorithm presented here for nonstationary constrained least-squares filtering deals with the transmission photon counting noise problem in the presence of limited dosages and highly nonhomogeneous fields by exchanging artifacts for less colored noise, which is more easily overcome by the viewer.
Abstract: The algorithm presented here for nonstationary constrained least-squares filtering deals with the transmission photon counting noise problem in the presence of limited dosages and highly nonhomogeneous fields. This technique consists of a design for a set of nonstationary filters, tuned to local noise autocorrelation functions in the reconstructed image. Estimates of noise properties can be taken from object detection information. The approach yields improvement in the tradeoff between noise levels and resolution of the image by exchanging artifacts for less colored noise, which is more easily overcome by the viewer. >
23 citations
TL;DR: In this article, the authors derived an exact theory of the mean first-passage time for an arbitrary one-dimensional dynamical system driven by a multiplicative external noise with finite correlation time (colored noise).
Abstract: We derive an exact theory of the mean first-passage time for an arbitrary one-dimensional dynamical system driven by a multiplicative external noise with finite correlation time (colored noise). Three different cases of colored noise are discussed: the random telegraph signal, the pre-Gaussian noise, and the Ornstein-Uhlenbeck diffusion process. Using random telegraph signals as a tool, we fully solve the difficult problem of non-Markovian boundary conditions associated with such a problem. The analytic solution with these boundary conditions give a complete solution of the escape time in the presence of colored noise.
21 citations
11 Jun 1991
TL;DR: In this paper, the spatiotemporal center weighted median (CWM) filter for image sequences is investigated and it is statistically shown that the CWM filter preserves image structures under motion at the expense of noise suppression.
Abstract: Statistical properties of the spatiotemporal center weighted median (CWM) filter for image sequences are investigated. It is statistically shown that the CWM filter preserves image structures under motion at the expense of noise suppression. To improve the CWM filter, an adaptive CWM (ACWM) filter having a variable central weight is presented. It is shown that the ACWM filter can preserve image structures under motion while suppressing noise, and thus can be effectively used in image sequence filtering. >
16 citations
TL;DR: This approach is compared with five other well known filter algorithms taken from the literature and it is shown that the new method is simple and fast and smoothes the objects within an image without losing edge information and without creating undesired structures.
23 Sep 1991
TL;DR: In this article, a source consistency filtering (SCF) is proposed to distinguish noise from signal by using the constraints provided by the redundancy of leads of multichannel recordings to distinguish signal from noise.
Abstract: A novel filter, called SCF (source consistency filtering), is introduced. In its spatial form, SCF uses the constraints provided by the redundancy of leads of multichannel recordings to distinguish noise from signal. Quantitatively, each lead is predicted using least-squares-method-derived coefficients, and a multichannel coherence of leads and their respective predictions results in an effective noise reduction in electrocardiogram recording. >
01 Aug 1991
TL;DR: A morphological filtering analog to the classic Wiener filter is described, designed to exploit differences in the spatial nature of the objects in the ideal noiseless images as compared to the noise/clutter objects.
Abstract: Filtering by morphological operations is particularly suited for removal of clutter and noise objects which have been introduced into noiseless binary images. The morphological filtering is designed to exploit differences in the spatial nature (shape, size, orientation) of the objects (connected components) in the ideal noiseless images as compared to the noise/clutter objects. Since the typical noise models (union, intersection set difference, etc.) for binary images are not additive, and the morphological processing is strongly nonlinear, optimal filtering results conventionally available for linear processing in the presence of additive noise are not directly applicable to morphological filtering of binary images. In this paper, a morphological filtering analog to the classic Wiener filter is described.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.
01 Jan 1991
TL;DR: In this paper, three-dimensional weighted median filters for noise reduction in gray-level image sequences are presented, which are designed to retain completely non-moving regions in image sequences.
Abstract: Three-dimensional weighted median filters for noise reduction in gray-level image sequences are presented. The filters are designed to retain completely nonmoving regions in image sequences. In moving regions, 3-D filter structures adapt automatically to changes in image sequence content and perform a filtering operation in moving regions. Noise attenuation capability of the filters is studied. The performance of the filters in terms of detail preservation and noise reduction is examined using real image sequences. >
TL;DR: The iterative method for removing signal-independent, additive noise from digital images is described, which successfully processes degraded images by filtering noise from regions of uniform intensity while preserving texture pixels and edges.
Abstract: An iterative method for removing signal-independent, additive noise from digital images is described. The underlying concept of this fully automated method is noise filtering by use of local statistics. Assuming that the noise is statistically stationary, noise variance is estimated from an input image by utilizing its smallest local variances. Noise filtering is performed iteratively and terminates when the estimated noise variance converges to zero. The method successfully processes degraded images by filtering noise from regions of uniform intensity while preserving texture pixels and edges. >
11 Jun 1991
TL;DR: The authors develop a novel adaptive algorithm for noise filtering that combines the mean filter with the median filter and achieves all kinds of noise removal as well as edge preservation.
Abstract: The authors develop a novel adaptive algorithm for noise filtering. This filter combines the mean filter with the median filter and achieves all kinds of noise removal as well as edge preservation. This filter reduces both impulsive and nonimpulsive noise. This filter is constructed from simple filters and there is no need for a priori information on the input signals. >
11 Jun 1991
TL;DR: In this article, the authors outline the development of this theory from its beginnings in the study of the noise removal properties and structural behavior of the median filter to the recently developed theory of optimal stack filtering.
Abstract: Within the last two decades a useful, nontrivial theory of nonlinear signal processing has been built around the median filter. The authors outline the development of this theory from its beginnings in the study of the noise removal properties and structural behavior of the median filter to the recently developed theory of optimal stack filtering. A recent application of stack filters to the problem of edge detection in noisy images is provided to demonstrate the effectiveness of this new theory. >
TL;DR: In this paper, the rotation invariant minimum noise and correlation energy (RI-MINACE) filter was proposed to provide sharp, easily detected correlation peaks and excellent discrimination against other false objects.
Abstract: Advanced filter algorithms using in-plane rotation invariance as an additional constraint during filter synthesis are presented. The new rotation-invariant Minimum Average Correlation Energy (RI-MACE) filter provides sharp, easily detected correlation peaks and excellent discrimination against other false objects. In the presence of background noise and clutter, the rotation-invariant Minimum Noise and Correlation Energy (RI-MINACE) filter (modified RI- MACE filter) uses the noise information during filter synthesis to obtain improved noise performance and also maintain easily detected correlation peaks. New test results are presented to show the improved discrimination capability of the RI-MACE filter and the improved noise performance of the RI-MINACE filter.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.
09 Apr 1991
TL;DR: Pertinent criteria to remove noise in color images with median filters and reconstruct the perceptually optimum output of the median filters are derived and justified and experiments demonstrate noise suppression and minimization of color distortion.
Abstract: The failure of well-developed techniques for grey-level scalar images when directly applied to remove impulse noise with median filters in color vector images is illustrated. Pertinent criteria to remove noise in color images with median filters and reconstruct the perceptually optimum output of the median filters are derived and justified. The experiments demonstrate noise suppression and minimization of color distortion. >
30 Sep 1991
TL;DR: The authors start from the Bayes approach and the assumption that surfaces are piecewise smooth and corrupted by a combination of white Gaussian and salt and pepper noise and show that such surfaces can be modelled by introducing an outlier process that is capable of 'throwing away' data.
Abstract: The authors discuss the problem of detecting outliers from a set of surface data. They start from the Bayes approach and the assumption that surfaces are piecewise smooth and corrupted by a combination of white Gaussian and salt and pepper noise. They show that such surfaces can be modelled by introducing an outlier process that is capable of 'throwing away' data. They make use of mean field techniques to finally obtain a deterministic network. The experimental results with real images support the model. >
28 Aug 1991
TL;DR: The modified gain function and center clipping are proposed to reduce 'residual noise' and it is shown that these algorithms can improve the speech SNR of 5- -10dB by 3.4- 12dB, and provide enhanced speech with colorless noise.
Abstract: Three types of noisy speech enhancement algorithms based on Minimum Mean-Square Error Short-Time Spectral Amplitude estimation (MMSE-STSA) are experimentally compared, especially under low SNR. Among them, MMSE-STSA and MMSE-LOG-STSA have been reported [1],[2] and MMSE Short-Time Relative Spectral Amplitude (MMSE-REL-STSA) estimation is derived in this paper. The modified gain function and center clipping are proposed to reduce 'residual noise'. It is shown that these algorithms can improve the speech SNR of 5- -10dB by 3.4- 12dB, and provide enhanced speech with colorless noise.
TL;DR: In this article, a new median filter for the restoration of images corrupted with additive Gaussian noise is described, and experimental results show that the performance of this improved median filter is better than that of the standard median filter.
Abstract: A new median filter for the restoration of images corrupted with additive Gaussian noise is described. Statistical analysis and experimental results show that the performance of this improved median filter is better than that of the standard median filter for erasing Gaussian noise, especially at high noise levels.
TL;DR: A realizable inverse filter of the received signal is derived and by using this filter a high resolution of superposed signals is obtained and the noise power estimation at the inverse filter output posses the interesting NAR property, even in presence of overlapped signals.
TL;DR: An approach using an entropy measure on edges to differentiate between variations in the image due to edge information as compared against noise, which uses entropy calculated against the spatial contour variations of edges in the window.
Abstract: A recurring problem in adaptive filtering is selection of control measures for parameter modification. A number of methods reported thus far have used localized order statistics to adaptively adjust filter parameters. The most effective techniques are based on edge detection as a decision mechanism to allow the preservation of edge information while noise is filtered. In general, decision-directed adaptive filters operate on a localized area within an image by using statistics of the area as a discrimination parameter. Typically, adaptive filters are based on pixel to pixel variations within a localized area that are due to either edges or additive noise. In homogeneous areas within the image where variances are due to additive noise, the filter should operate to reduce the noise. Using an edge detection technique, a decision directed adaptive filter can vary the filtering proportional to the amount of edge information detected. We show an approach using an entropy measure on edges to differentiate between variations in the image due to edge information as compared against noise. The method uses entropy calculated against the spatial contour variations of edges in the window.
01 Jan 1991
TL;DR: In this paper, a simple iterative scheme for estimating colored sequences is presented, in which the colored plant noise is modeled as the output of a shaping filter excited by white noise.
Abstract: In many deconvolution problems, the signal to be estimated is modeled as the input to a known plant and assumed white. There are, however, situations in which this signal is not white. A simple iterative scheme for estimating colored sequences is presented. In this scheme, the colored plant noise is modeled as the output of a shaping filter excited by white noise. The shaping filter is considered as part of the plant while applying Mendel's minimum variance deconvolution (MVD) algorithm based on the Kalman filter to estimate the plant noise. To begin with, the shaping filter is just an identity filter. The estimated plant noise is then used to update its coefficients iteratively until the change in the coefficient values is small. The iterative scheme has been tested using simulated data under different conditions, and is found to perform quite well under certain situations. >
23 Sep 1991
TL;DR: In this article, a stack filter based method was proposed for edge detection in which intensity edges were obtained by thresholding the difference between the image and a dilated version of the image.
Abstract: Recently, a morphological method was proposed for edge detection in which intensity edges were obtained by thresholding the difference between the image and a dilated version of the image. While this technique is promising, it is quite sensitive to noise. To improve noise immunity and robustness, we propose using stack filters to estimate the dilated and eroded versions of the image, and then threshold the difference between these two images. Comparisons between this stack filter based technique and some standard edge detectors are provided. For instance, we find that this approach yields results comparable to those obtained with the Canny operator for images with additive Gaussian noise, but works much better when the noise is impulsive. Extensive simulations with many different images and different types of noise were performed. Pratt's figure of merit was used as an objective measure of performance on synthetic images. Many natural scenes were also used to test the performance of this technique. The results indicate that this approach is robust with respect to changes in both the image and the noise. In other words, filters obtained by training on one image and one type of noise work well even when both the image and noise statistics vary.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.
16 Jun 1991
TL;DR: In this paper, a new structure for adaptive noise cancellation is suggested to improve the convergence characteristics of extracting a non-stationary correlated signal corrupted by random noise, which composes of two linear phase filters which are connected in parallel and both filters are being adapted independently according to Widrow's least-mean-square algorithm.
Abstract: A new structure for adaptive noise cancellation is suggested to improve the convergence characteristics of extracting a non-stationary correlated signal corrupted by random noise. The canceller composes of two linear phase filters which are connected in parallel and both filters are being adapted independently according to Widrow's least-mean-square (LMS) algorithm. The merits of the new configuration are discussed and a comparison with the conventional structure that consists of a single transversal filter is also included. Experimental results show that the proposed canceller has a superior performance in the context of removing contaminated noise since gradient fluctuation during adaptation could be suppressed considerably which in turn speeds up the rate of convergence. >
16 Jun 1991
TL;DR: In this article, the authors proposed a class of nonlinear filters based on the M-estimates of location parameters in statistical theory, which are called moving-window M-type filters.
Abstract: Proposes a class of nonlinear filters-M-type filters, which are based on the M-estimates of location parameters in statistical theory. Moving-window M-filters can be defined starting from M-estimates of location. Then, several modified structures are given. It is shown that recursive M-type filter can perform better than nonrecursive M-type filter, running mean and median filter in white noise suppression, while it can be designed to be comparable to the median filter in edge preservation in the presence of noise. (To suppress impulse noise, the proposed filter generates output with a modified reference output: to reduce calculation complexity, M-type filter with linear FIR substructure is introduced.) Finally, experimental results for one- and two-dimensional signals are presented. >
01 Jun 1991
TL;DR: A novel noise smoothing method based on a nonparametric statistic runs test that is able to smooth only noisy areas without reducing the spatial resolution in the image.
Abstract: In this paper we describe a novel noise smoothing method based on a nonparametric statistic runs test. We assume that the data bits of a pixel can be divided into signal bits and noise bits. The signalcomprises the most significant bits and the noise bits are the least significant ones. The idea in this smoothing method is to preserve the signal bits and only modify the noise bits. The number of noise bitsof each pixel is determined based on the runs in the neighborhood. If the number of noise bits is zero then no smoothing is necessary. The degree of smoothing is a function of the number of noise bits. Using this technique we are able to smooth only noisy areas without reducing the spatial resolution in the image. The algorithm is easy to implement. The application of the smoothing algorithm on a chest image was given. 1. INTRODUCTION Image noise may come from the x-ray source, optical to electrical conversion, quantization error, or from thermal noise in the electronic amplifier. Noise degrades image quality and reduces spatial