Showing papers on "Median filter published in 1991"

Center weighted median filters and their applications to image enhancement

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. >

A class of fast Gaussian binomial filters for speech and image processing

Abstract: The authors present an efficient, in-place algorithm for the batch processing of linear data arrays. These algorithms are efficient, easily scaled, and have no multiply operations. They are suitable as front-end filters for a bank of quadrature mirror filters and for pyramid coding of images. In the latter application, the binomial filter was used as the low-pass filter in pyramid coding of images and compared with the Gaussian filter devised by P.J. Burt (Comput. Graph. Image Processing, vol.16, p.20-51, 1981). The binomial filter yielded a slightly larger signal-to-noise ratio in every case tested. More significantly, for an (L+1)*(L+1) image array processed in (N+1)*(N+1) subblocks, the fast Burt algorithm requires a total of 2(L+1)/sup 2/N adds and 2(L+1)/sup 2/ (N/2+1) multiplies. The binomial algorithm requires 2L/sup 2/N adds and zero multiplies. >

Multivariate ordering in color image filtering

Abstract: Multivariate data ordering and its use in color image filtering are presented. Several of the filters presented are extensions of the single-channel filters based on order statistics. The statistical analysis of the marginal order statistics is presented for the p-dimensional case. A family of multichannel filters based on multivariate data ordering, such as the marginal median, the vector median, the marginal alpha -trimmed mean, and the multichannel modified trimmed mean filter, is described in detail. The performance of the marginal median and the vector median filters in impulsive noise filtering is investigated. Simulation examples of the filters under study are described. >

Multistage order statistic filters for image sequence processing

Abstract: The application of multistage order statistic filters (MOS) to the task of noise suppression in time-varying imagery is studied. It is shown that MOS filters efficiently preserve image structures under motion without motion compensation preprocessing. In particular, the families of multistage median and weighted median filters are considered. Motion preservation and statistical smoothing measures are derived. It is shown that spatiotemporal filtering allows for a significant improvement over both spatial and temporal filtering in terms of output image resolution and noise suppression. >

Gain calibrating nonuniform image-array data using only the image data

Abstract: An algorithm is developed for calibrating the spatial nonuniformity of image-array (CCD-type) detectors. Like other techniques this approach uses multiple, spatially displaced images. In circumstances where high-precision flat fields are not available by other means (i.e., sky flats) this technique is advantageous as it uses the data frames for gain calibration even when the array images extended, nonuniform, sources. Numerical experiments and direct observations with intrinsically uniform and quite nonuniform detectors show that this algorithm is useful when data frames are crowded with sources - circumstance where 'median filtering' flatfielding techniques often fail. The algorithm described is robust and efficiently uses information from multiple data frames to determine pixel gain variations, using visible and IR array observations of extended sources.

Image signal processing with digital filtering to minimize aliasing caused by image manipulation

Abstract: An image signal processing apparatus has an address generator that generates addresses for successive words of a digital input signal representing an input image to be manipulated. A computation circuit monitors the addresses for successive word sets which each comprise first and second words relatively horizontally spaced in the input image and a third word aligned vertically with the second word, and computes from each set both horizontal and vertical local scaling factors representing the extent of compression in the horizontal and vertical directions of a corresponding local area of the manipulated image. A bandwidth controller is responsive to the successive scaling factors to vary the horizontal and vertical bandwidths of a digital filter which two-dimensionally filters the input image, thereby to minimize aliasing that would be caused by the manipulation of the local areas. The computation circuit computes the scaling factors for each local area such that the horizontal factor corresponds to the length of a line that extends perpendicularly from a line passing through the positions of the second and third words in the manipulated image to the position of the first word and the vertical factor corresponds to the length of a line that extends perpendicularly from a line passing through the positions of the first and second words to the position of the third word.

Segmentation method for use against moving objects

Abstract: Three image frames containing an object of interest and background clutter are taken at successive time intervals and stored in memory. The background of images A, B and C are registered preferably using an area correlator 12. A median filter 16 is used to select a median value from the registered image frames. Then, subtractor 18 serves to subtract the median pixel values from one of the image frames. This difference output is then thresholded to provide a binary signal whose pixel values exceeding the threshold levels are generally associated with the position of the moving object.

On estimation of noise variance in two-dimensional signal processing

Abstract: Estimation of noise variance is an important component of digital signal processing, in particular of image processing. In this paper we develop methods for estimating the variance of white noise in a two-dimensional degraded signal. We discuss optimal configurations of pixels for difference-based estimation, and describe asymptotically optimal selection of weights for the component pixels. After extensive analysis of possible configurations we recommend averaging linear configurations over a variety of different orientations (usually two or four). This approach produces estimators with properties of both statistical and numerical efficiency.

Comparison of algorithms for standard median filtering

Abstract: Standard median filtering that is searched repeatedly for a median from a sample set which changes only slightly between the subsequent searches is discussed. Several well-known methods for solving this running median problem are reviewed, the (asymptotical) time complexities of the methods are analyzed, and simple variants are proposed which are especially suited for small sample sets, a frequent situation. Although the discussion is restricted to the one-dimensional case, the ideas are easily extended to higher dimensions. >

Two-dimensional adaptive block Kalman filtering of SAR imagery

Abstract: A method for removing speckle from synthetic aperture radar (SAR) imagery by using 2-D adaptive block Kalman filtering is introduced. The image process is represented by an autoregressive model with a nonsymmetric half-plane (NSHP) region of support. New 2-D Kalman filtering equations are derived which taken into account not only the effect of speckles as multiplicative noise but also the effects of the additive receiver thermal noise and the blur. This method assumes local stationarity within a processing window, whereas the image can be assumed to be globally nonstationary. A recursive identification process using the stochastic Newton approach is also proposed which can be used on-line to estimate the filter parameters based upon the information within each new block of the image. Simulation results on several images are provided to indicate the effectiveness of the proposed method when used to remove the effects of speckle noise as well as those of the additive noise. >

Bio-potential signal processor for magnetic resonance imaging

Abstract: An adaptive filtering system is used to reduce the electrical noise on low power ECG signals and the like as generated by rapidly switched magnetic gradient fields in magnetic resonance imaging A correlated noise reference signal is derived from the inputs to the gradient coils by a combination differentiator and low pass filter The noise reference signal is received by a filter having adjustable coefficients and the result subtracted from the low power signal to produce an error signal used to adjust the coefficients of the filter Three separate adaptive filters may be placed in series each with a separate correlated noise reference signal to reduce the noise from three gradient coils

Experimental studies in transmission ultrasound computed tomography.

Abstract: The reconstruction of the speed-of-sound distribution within a target can be achieved by CT techniques from measurements on transmitted ultrasonic pulses. The mathematical relationship between speed-of-sound imaging and the conventional CT situation is explained. An experimental system, which has been developed to investigate speed-of-imaging and other forms of in-vivo ultrasound CT, is described, along with the technique used for data acquisition and image reconstruction. These include measurement of pulse time-of-flight by the threshold or cross-correlation methods. Techniques for reducing artifacts in speed-of-sound images are also described, such as median filtering and modified Shepp-Logan filtering. These techniques have been used to obtain high quality speed-of-sound images of various phantoms.

Spatial prediction filtering in the t-x and f-x domains

Abstract: The predictability of seismic signals from nearby traces can be a powerful tool for reducing random or locally coherent noise. The choice of algorithm to reduce noise for a given application is a function of the data signal and noise characteristics. When the signal and noise are relatively consistent over a given design window, an f-x domain Wiener‐filter approach can be used. For cases in which the data are time‐ or space‐varying, a new approach using 2-D adaptive filtering in the t-x domain can be very effective. In either of these approaches, a prediction trace‐gap can often be successfully used to remove locally coherent noise when lateral signal changes are not too rapid.

Efficient serial and parallel algorithms for median filtering

Abstract: A serial algorithm for separable median filtering is developed that requires only two comparisons per element when the window size is three. In addition, fast parallel concurrent-read-exclusive-write parallel random-access machine (CREW PRAM) algorithms with good processor-time product are developed for separable median filtering and two-dimensional median filtering. >

Complete classification of roots to one-dimensional median and rank-order filters

Abstract: The set of roots to the one-dimensional median filter is completely determined. Let 2N+1 be the filter window width. It has been shown that if a root contains a monotone segment of length N+1, then it must be locally monotone N+2. For roots with no monotone segment of length N+1, it is proved that the set of such roots is finite, and that each such root is periodic. The methods used are constructive, so given N, one can list all possible roots of this type. The results developed for the median filter also apply to rank-order filters. >

Mechanism for improving television display of still images using image motion-dependent filter

Abstract: An image motion-dependent mechanism selectively inserts an image modification operator, such as a median filter, into the image signal path of a video display unit, so that loss of definition and introduction of flicker resulting from the use of a median filter in the course of the display of a still or stationary image on an IDTV are effectively eliminated. The selectively inserted image modification operator, which responds to an index code or bit value representative of whether or not the image frame is repeated, may also serve to prevent the generation of cross-luminance and cross-color decoding artifacts, as well as reducing the amount of random noise in the image.

Noise reduction using adaptive spatial filtering in photorefractive two-beam coupling

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.

Optimization of the weighted median filter by learning.

Abstract: The weighted median filter (WMF) is a generalization of the median filter. The WMF is more effective for image processing than the conventional median filter. However, the design of the parameters of the WMF is a difficult problem. A novel method of optimizing the WMF is proposed that utilizes the close relation between the nonrecursive WMF and the feed-forward neural network with shift-invariant weight coefficients. The optimization problem of the WMF results in the learning of the interconnection weights of the network.

Median based idempotent filters

Abstract: Idempotent filters produce a root signal in a single filter pass, i.e. the filter output is invariant to further filterings with the same filter. In this paper median based idempotent filter structures are introduced. Two approaches to generate these filters are studied: weighted median filters and median filter cascades. Two subclasses of n-dimensional idempotent weighted median filters, called Class 1 and Class 2 filters in the paper, are introduced. It is shown that both Class 1 and Class 2 filters suppress impulsive noise from n-dimensional input signals and yet have almost no effect on the non-corrupted parts of the signal. These filters are therefore well-suited for example for preprocessing purposes. An application to speech processing is described. Other likely applications of these filters are in image processing and, also, in image sequence processing, where the filter mask is typically 3-dimensional. Sufficient conditions for a filter cascade to be idempotent are given. Two idempotent median filter cascades and their advantages are discussed.

3-dimensional median filters for image sequence processing

Abstract: Two 3-D median-based filtering algorithms have been developed that preserve the motion in the image sequence while attenuating noise effectively. Some observations are made on the root signals in the binary domain based on the positive Boolean functions corresponding to the filters. From the Boolean expressions the output distribution functions are derived. The performance of both filters under various noise types is examined theoretically and experimentally. The structures are simulated on a video sequencer (DVSR 100) on real image sequences. Comparisons are made with other 2- and 3-D algorithms from the literature based on mean square error, mean absolute error, and subjective criteria. >

Hierarchical motion-compensated deinterlacing

Abstract: This paper introduces a new method of converting interlaced video to a progressively scanned video.The missing pixel values of the interlaced sequence are interpolated from past fields according to motion vectors found between the present and past. Hierarchical block-matching motion estimation is used in finding these motion vectors. This approach gives improved results over conventional de-interlacing meth-ods such as median filtering and linear spatial iflterj)ola.tion. 1. INTRODUCTION Interlaced video is widely used in transmitting a sequence ofimages. Most television systems incorporateinterlace for its many advantages, including the reduction of baudwith. However, some image processingtasks require that interlaced video l)e converted to progressively scanned video. Conversion of frame ratesbetween different interlaced television standards is a. good example. For television systems implementingprogressively scanned display, de-interlaciiig is certainly a. necessity.Motion estimation has found use in video coding and image sequence restoration or deblurring, and awide variety of algorithms have been developed over the past several years[1-6J. In particular, [6] appliesa two-level hierarchy to the problem of interpolating missing fields in a reduced-resolution test sequence.Motion estimation has also been applied to the problem of de-interlacing. In this paper we apply a QMFpyramid based motion estimation technique[fl to the dc-interlacing problem in a novel manner.

Non-linear filtering techniques for narrow-band interference rejection in direct sequence spread-spectrum systems

Abstract: The authors concentrate on the problem of interference rejection for the cases of single and multitone jammers. They propose the use of a conditional nonlinear median filter operating in the transform domain, for the detection and suppression of narrowband signals of sufficient power, without regard to their center frequency, bandwidth, or peak power. It is shown that this approach offers several significant advantages over similar techniques that have been used before. >

Edge preserving filtering by combining nonlinear mean and median filters

Abstract: A very general class of nonlinear filters, called mapping order-statistics filters (MOSFs), is introduced. A subclass of the MOSFs called L/sub p/-mean median filters (L/sub p/-MMFs) is discussed. The L/sub p/-MMF shows better properties in preserving edges and attenuating positive and negative impulse noise simultaneously than do the nonlinear mean filters. In the analysis of the deterministic properties of the L/sub p/-MMF, the concept of a monotonic trend region is introduced. A necessary condition for a signal to be a root signal is given. This necessary condition is also valid for the known finite-impulse-response median hybrid filter. >

Nonlinear spatio-temporal noise suppression techniques with applications in image sequence processing

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. >

Robust median filter with adaptive window length

Abstract: The authors introduce a novel adaptive median filter algorithm which efficiently exploits the benefits of the variable size data window. The method exploits the noise rejecting properties of the long window sizes and the edge preservation of the short windows. Adaptive window lengths reduce the variance of the filtered signal considerably without degrading the accurate representation of the discontinuities. Finally, a computationally efficient implementation of the proposed adaptive median filter is presented. Examples show that the proposed method outperforms the conventional fixed window length median operators. >

New Concept Of Signal Averaging In Time Domain

Abstract: Noise reduction is one of the main problems to overcome in order to obtain a correct interpretation of biological signal Signal averaging is a very common method for improvement the signal-tonoise ratio (SNR) Usually noise is assumed to be stationary but generally speaking that is not true The main idea of the method is that cycles have not equal weights in averaging process

Synthesis of optimal detail-restoring stack filters for image processing

Abstract: A two-step method is presented to synthesize optimal stack filters under the mean absolute error (MAE) criterion. First, the probabilities needed in the optimal filter design are estimated based on images. Second, the linear program (LP) required for finding the best filter is avoided by a 'reasonably good' suboptimal routine which only involves data comparisons. A sufficient condition under which the suboptimal routine results in optimal solutions is given and shown to hold in most practical cases. The proposed method is then applied to synthesize a family of optimal stack filters for the task of restoring an image in impulsive noise. Testing results show that the synthesized filters lead to a greatly improved image-detail restoration compared to standard median filters, although median filters remove noise better. >