Showing papers on "Median filter published in 1985"

Adaptive Noise Smoothing Filter for Images with Signal-Dependent Noise

Abstract: In this paper, we consider the restoration of images with signal-dependent noise. The filter is noise smoothing and adapts to local changes in image statistics based on a nonstationary mean, nonstationary variance (NMNV) image model. For images degraded by a class of uncorrelated, signal-dependent noise without blur, the adaptive noise smoothing filter becomes a point processor and is similar to Lee's local statistics algorithm [16]. The filter is able to adapt itself to the nonstationary local image statistics in the presence of different types of signal-dependent noise. For multiplicative noise, the adaptive noise smoothing filter is a systematic derivation of Lee's algorithm with some extensions that allow different estimators for the local image variance. The advantage of the derivation is its easy extension to deal with various types of signal-dependent noise. Film-grain and Poisson signal-dependent restoration problems are also considered as examples. All the nonstationary image statistical parameters needed for the filter can be estimated from the noisy image and no a priori information about the original image is required.

Generalized median filtering and related nonlinear filtering techniques

Abstract: We consider some generalizations of median filters which combine properties of both the linear and median filters. In particular, L filters and M filters are considered, motivated by robust estimators which are generalizations of the median as a location estimator. A related filter, which we call the modified trimmed mean (MTM) filter, is also described. The filters are evaluated for their performance on noisy signals containing sharp discontinuities or edges. It is shown that M filters can offer a more favorable combination of the running mean and median filters than can L filters, while MTM filters generally have better characteristics than M filters. We also show that an MTM filter is a data-dependent modification of L filters. The concept of double-window filtering is introduced as a refinement of MTM filtering. One representative set of filtered sequences of a test input using these filters are presented to illustrate the performance characterisics of these filters.

Geometric filter for speckle reduction.

Abstract: An algorithm is described which reduces speckle noise in images. It is a nonlinear algorithm based on geometric concepts. Tests were performed on synthetic aperture radar images which show that it compares favorably with a 3 × 3 median filter.

Adaptive filters for digital image noise smoothing: An evaluation*

Abstract: Six adaptive noise filtering algorithms were implemented and evaluated. There are (1) median filtering, (2) K-nearest neighbor averaging, (3) gradient inverse weighted smoothing, (4) sigma filtering, (5) Lee additive and multiplicative filtering, and (6) modified Wallis filtering. For the sake of comparison, the mean filter was also included. All algorithms were tested on noise corrupted copies of a composite image consisting of a uniform field, a bar pattern of periods increasing from 2 to 20 pixels, printed text, and a military tank sitting on desert terrain. In one test, uniformly distributed noise between gray levels of −32 and 32 was added to the composite image and filtered. In a second test, multiplicative Gaussian noise with mean 1.0 and standard deviation 0.25 was introduced, then filtered. A 7×7 pixel processing window was used in all six adaptive algorithms and the mean filter for both tests. An adaptive filter was used iteratively with varying window sizes to demonstrate the success of iterative adaptive smoothing. Filtering results were evaluated from statistics, examination of transects plotted from each filtered bar pattern, and from visual ranking by a group of observers.

Adaptive noise cancellation for a class of nonlinear, dynamic reference channels

Abstract: The computational complexity of nonlinear adaptive noise cancellation can be reduced by restricting the nonlinearity expected in the reference path to the noise canceler. The class of zero memory nonlinearities preceded by linear processors in the reference path is considered. Noniterative and iterative methods of system identification are applied to the determination of processor parameters in the noise canceler. The computational requirements of each of the algorithms are compared, and the iterative method is modified for improved convergence. Experimental results are presented for the modified iterative algorithm.

Root properties and convergence rates of median filters

Abstract: Median filters are a special class of ranked order filters used for smoothing signals. Repeated application of the filter on a quantized signal of finite length ultimately results in a sequence, termed a root signal, which is invariant to additional passes of the median filter. In this paper, the theory is developed both for determining the cardinality of the root signal space of arbitrary window width filters applied to signals with any number of quantization levels and for counting or estimating the number of passes required to produce a root for binary signals.

Equivalent-noise model for contrast detection and discrimination.

Abstract: Models for the detection and the discrimination of low-contrast signals by human observers typically assume that the observer is limited by the filtering action of the visual system and by the noisy character of its processing. For some models both the filtering and the noise can be represented by a noise in the stimulus domain, the input equivalent noise of the model. We derive some formulas for computing this noise, describe the calculation of a sample, and discuss some implications of this approach.

Image preprocessing by modified median filter

Abstract: Preprocessing of a noisy image, such as is obtained by a raster scan, by tracting the median of the pixel values of one or more lines from the medians of predetermined pixel neighborhoods

VLSI Implementation of a fast rank order filtering algorithm

Abstract: A useful class of digital filters is the rank order filter, of which the median filter is the most popular. This paper details the NMOS IC design of a 6-bit rank order filter based upon the approach of Fitch, Coyle, and Gallagher. The chip provides for programmable window lengths of 3, 5, 7, and 9; the selection of any element in the rank order; operation in the nonrecursive or recursive modes; and the ability to function in a "ganged" mode so that two or more identical chips can accommodate samples of length seven or eight bits.

Smoothed median filters with FIR substructures

Abstract: In this paper we introduce a new generalized class of Median Filters (MF) which we call Smoothed Median Filters (SMF). The input signal x(n) is filtered with M linear phase FIR filters and the output of the SMF is the median of the outputs of the FIR filters. The statistical properties of the SMF's are analyzed for input signals with Gaussian, double exponential, and uniform density functions. It is shown that SMF's have essentially the same statistical properties and same type of root signals as the conventional MF's. SMF's preserve corner points and ramps and attenuate impulsive type noise components effectively. An interesting subclass of the SMF's requiring only two scaling multipliers and three data sorting operations irrespective of the filter length is introduced.

VLSI Architecture for a one chip video median filter

Abstract: Real-time image processing in an application environment needs a set of low-cost implementations of various algorithms. This paper presents a one chip VLSI median filter based on a systolic processor and working at video rate. It includes its own memory and can be used without any image memory for on-line processing. The architectural choices have made it possible to design a small size chip with a high performance level.

Output distributions of two-dimensional median filters

Abstract: The output distribution for the separable two-dimensional median filter and the square window two-dimensional median filter are derived for several cases, including first-order and multivariate output distributions with white noise, signals plus white noise, and general multivariate input images. These are then used to compute some specific statistics in several illustrative examples.

Noise capture properties of the constant modulus algorithm

Abstract: In communication applications, the Constant Modulus Algorithm has proven an effective means for adjusting coefficients of an adaptive filter. However, under certain conditions, it may generate a response that discriminates against the signal component in a preference for broadband noise for sustained periods. Work has been done toward characterizing and modeling this phenomenon.

Method of reducing the visibility of the noise in a sequence of video images and device for implementing this method

Abstract: The invention relates to a method and a device for reducing the visibility of the noise in a sequence of video images by filtering the luminance value of the fixed points and of the points whose motion is due to the noise by an inter-image recursive filtering and by filtering the points with true motion by a median filtering consisting in replacing a luminance value by the median value of the luminance values situated within a window centred on the relevant point, these latter values being ordered. The device includes: a motion detector 2; a median filter 4; an inter-image recursive filter 5; an image memory 3; a multiplexer with two inputs and one output 6; and a bidirectional recursive filter 7. The multiplexer 6 transmits: either a filtered value provided by the output of the median filter 4, when the point has a true motion; or a filtered value provided by the inter-image recursive output 5 when the point is fixed or has a motion due to noise. The bidirectional recursive filter 7 filters the values provided by the output of the multiplexer 6 whatever the state of the point processed, so as to smooth out the demarcations between zones in which the filtering differs. Application to the reduction of noise in television images.

A fast sort-selection filter chip with effectively linear hardware complexity

Abstract: A hardware architecture for obtaining the t(th) largest number from a list of N numbers, each B bits long is presented. When t is such that N=2t-1, the system described is a hardware median filter. It is also shown that a pipelined architecture can perform a real-time filtering operation at a rate of one median computation per cycle, where the cycle time of this system is proportional to \log N . This hardware selection architecture is currently being implemented on a VLSI chip (7K gates plus testability structures) to build an N=25 median filter with a throughput rate for image processing of better than 10M pixels/second with 8 bits/pixel. This architecture can also be used as an efficient basis of a hardware sorter.

An alternative editing technique for oceanographic data

Abstract: A specific problem of data processing lies in data errors which inhibit the routine handling of a large amount of data. To edit these data, a special median filter is introduced. Its efficiency for CTD data processing as well as examples of its practical application are described and discussed by means of comparison of two median filter techniques (MEDFIL and MEDDIF) with a conventional method (MAXDIF).

Image restoration by an iterative damped least-squares method with non-negativity constraint.

Abstract: In this paper we describe a new digital method for restoring linearly degraded images in the presence of noise. The restoration procedure is an iterative damped least-squares (DLS) algorithm that is based on the principle of damped least squares. This method acquires the advantage of tolerance to noise by incorporating additional constraints of non-negativity of the object and the adaptive technique. We also use a median filter in the proposed algorithm to suppress spiky noise. After discussing the convergence of the iterative DLS algorithm, we present some results of computer simulations that demonstrate the effectiveness of the proposed method.

Noise Evaluation and Filter Design in CT Images

Abstract: Filtering CT images to remove noise, and thereby enhance the signal-to-noise ratio in the images, is a difficult process because CT noise is of a broad-band spatial-frequency character, overlapping frequencies of interest in the signal. We present a measurement of the noise power spectrum of a CT scanner and show that some form of spatially variant filtering of CT images can be beneficial if the filtering process is based upon the differences between the frequency characteristics of the noise and the signal. For evaluating the performance, we used a percentage standard deviation, an index representing contrast, a frequency spectral pattern, and several CT images processed with the filter.

Identification of image and blur parameters for the restoration of noncausal blurs

Abstract: An optimal statistical parameter estimation technique is presented for the identification of unknown image and blur-model parameters. Maximum likelihood estimates of the unknown parameters are derived both in the absence and in the presence of observation noise. The proposed algorithms are able to locate the zeros of the observed image spectrum on the entire Z 1 -Z 2 plane and therefore, unlike previous algorithms, are not restricted to the Fourier domain. Images restored with the proposed algorithms are shown as examples.

Application of combined median-averaging filters to scintigraphic image processing.

Abstract: Statistical and deterministic properties of median filters are briefly discussed and their inherent advantages as a prospective tool in scintigraphic data processing are pointed out. The ability of median filters of suppressing impulse noise while the edge-like features of an image are preserved, is demonstrated on phantom data. The residual high-frequency noise remaining after median filtering can be subsequently reduced by standard smoothing procedures. A simple algorithm, made up of the superposition of a median and an averaging filter, is presented and shown to be a promising candidate in the quest for fast and easy-to-implement processing routine.

Generalized Median Filtering and Related Nonlinear

Abstract: Abstruct-We consider some generalizations of median filters which combine properties of both the linear and median filters. In particular, L filters and M filters are considered, motivated by robust estimators which are generalizations of the median as a location estimator. A related filter, which we call the modified trimmed mean (MTM) filter, is also described. The filters are evaluated for their performance on noisy signals containing sharp discontinuities or edges. It is shown that M filters can offer a more favorable combination of the running mean and median filters than can L filters, while MTM filters generally have better characteristics than M filters. We also show that an MTM filter is a data-dependent modification of L filters. The concept of double-window filtering is introduced as a refinement of MTM filtering. One representative set of filtered sequences of a test input using these filters are presented to illustrate the performance characterisics of these filters.

Time-Varying Polynomial System Approach to Multichannel Optimal Linear Filtering

Abstract: A new approach to linear estimation in time-varying discrete multivariable systems is described. The signal model is taken to be a time-varying vector difference equation which can be expressed in ARMA polynomial system form. An optimal linear filter and predictor is derived in terms of time-dependent polynomial operators and this can also be implemented as a recursive algorithm using difference equations. The system model and filter are particuarly relevant in self-tuning filtering applications.

The contour plot method for noise reduction in digital video.

Abstract: In this paper a digital image processing method for removing noise from still and moving grey-value pictures is presented and its performance is investigated. The method consists of replacing a particular noisy pixel by a weighted average of pixels in some standard neighborhood of the pixel. In case of a single still picture the neighborhood has two spatial dimensions, while for sequences of pictures, neighborhoods with an additional third (time) dimension are used. With the aid of edge detectors in various directions a king of contour plot is constructed which indicates to what extent a pixel of the standard neighborhood belongs to the same region as the pixel being restored. This contour plot is converted into a set of weight coefficients of a transversal filter. The performance of the method is assessed in terms of gain in signal-to-noise ratio (S/N-ratio) and by visual inspection.

Image restoration using a parallel indentification and filtering procedure

Abstract: In this paper a parallel identification scheme is presented to extend the fast parallel Kalman filter structure of [1] with an on-line identification procedure. It is shown that the identification problem can be specified as a parallel set of one-dimensional autoregressive moving average (ARMA) identification problems. For the case of linear motion blur in the presence of noise a simple and fast parallel identification algorithm is described. Several identification and restoration results are given as examples.

A median filter approach for correcting errors in a vector field

Abstract: Techniques are presented for detecting and correcting errors in a vector field. These methods employ median filters which are frequently used in image processing to enhance edges and remove noise. A detailed example is given for wind field maps produced by a spaceborne scatterometer. The error detection and replacement algorithm was tested with simulation data from the NASA Scatterometer (NSCAT) project.

"Noise cancellation studies using a least-squares lattice filter"

Abstract: The operation of a least-squares lattice filter, configured for noise cancellation, is described. The algorithm employed uses both apriori and a-posteriori unnormalised residuals and requires no auxiliary scalars. A 100-stage version of this filter has been applied to data of 5 and 10 seconds in length, sampled at 8kHz, comprising a passage of speech corrupted by additive noise. The algorithm successfully effected a noise reduction close to the optimum attainable level. By exploiting redundancy in the number of linear prediction lattice stages required, a reduction from 24 operations per stage to 9 operations was possible for the majority of stages, with no appreciable loss of performance.

Segmentation of multilevel images using Gibbs distribution

Abstract: An algorithm for multilevel image segmentation is presented. The main feature is that it is based on repetition of a binary segmentation algorithm, in a fashion similar to the binary expansion of an integer number. The binary segmentation considered assumes a Markov Random Field model for the original scene, and an additive i.i.d, noise signal. Simulation rasults are presented, and compared with other filtering algorithms, such as median filtering.

Picture processing unit

Abstract: PURPOSE:To reduce the amount of hardware by providing a means reading a picture element value from a picture data storage means at a fixed interval and a means writing an output of a median filter circuit at a fixed interval to eliminate a high energy noise with a filter of a small size. CONSTITUTION:Addresses of the 1st and 2nd picture memories 3, 5 are generated by an address counter 1 of a picture processing unit, added to an address conversion circuit 2 in synchronizing with the clock and converted into an address scanned repetitively at each other picture element by the circuit 2. The converted address is applied to the 1st memory 3 inputting a picture data from a picture signal generator of a device such as a video camera. Furthermore, the picture element value including the noise from the memory 3 is read to the median filter circuit 4 at a fixed interval, an output of the filter circuit 4 is written in the memory 5 and the noise with high energy is eliminated by the filter of a small size so as to simplify the constitution of the device.

Two-dimensional orthogonal median filters and applications

Abstract: Directionally selective median filtering may be accomplished either by passing a data-selecting window over the two-dimensional (2D) data or by pre-aligning the data columns and passing a horizontally-directed median filter over the data. If the data has features which appear as constants along some directional path, these constants may be selectively passed or stopped (filter-subtracted) from the data. Directional selectivity and angular resolution may be designed through the concept of orthogonal 2D median filters. The filters in an orthogonal set have fixed-size sample patterns which are mutually median-exclusive (or have minority common count). Synthetic examples of seismic applications are given and show very positive results.