No-Reference Quality Assessment for Screen Content Images Using Visual Edge Model and AdaBoosting Neural Network

In this paper, a competitive no-reference metric is proposed to assess the perceptive quality of screen content images (SCIs), which uses the human visual edge model and AdaBoosting neural network. Inspired by the existing theory that the edge information which reflects the visual quality of SCI is effectively captured by the human visual difference of the Gaussian (DOG) model, we compute two types of multi-scale edge maps via the DOG operator firstly. Specifically, two types of edge maps contain contour and edge information respectively. Then after locally normalizing edge maps, L -moments distribution estimation is utilized to fit their DOG coefficients, and the fitted L -moments parameters can be regarded as edge features. Finally, to obtain the final perceptive quality score, we use an AdaBoosting back-propagation neural network (ABPNN) to map the quality-aware features to the perceptual quality score of SCIs. The reason why the ABPNN is regarded as the appropriate approach for the visual quality assessment of SCIs is that we abandon the regression network with a shallow structure, try a regression network with a deep architecture, and achieve a good generalization ability. The proposed method delivers highly competitive performance and shows high consistency with the human visual system (HVS) on the public SCI-oriented databases.

Characterizing Architectural Distortion in Mammograms by Linear Saliency

Architectural distortion (AD) is a common cause of false-negatives in mammograms. This lesion usually consists of a central retraction of the connective tissue and a spiculated pattern radiating from it. This pattern is difficult to detect due the complex superposition of breast tissue. This paper presents a novel AD characterization by representing the linear saliency in mammography Regions of Interest (ROI) as a graph composed of nodes corresponding to locations along the ROI boundary and edges with a weight proportional to the line intensity integrals along the path connecting any pair of nodes. A set of eigenvectors from the adjacency matrix is then used to extract discriminant coefficients that represent those nodes with higher salient lines. A dimensionality reduction is further accomplished by selecting the pair of nodes with major contribution for each of the computed eigenvectors. The set of main salient lines is then assembled as a feature vector that inputs a conventional Support Vector Machine (SVM). Experimental results with two benchmark databases, the mini-MIAS and DDSM databases, demonstrate that the proposed linear saliency domain method (LSD) performs well in terms of accuracy. The approach was evaluated with a set of 246 RoI extracted from the DDSM (123 normal tissues and 123 AD) and a set of 38 ROI from the mini-MIAS collections (19 normal tissues and 19 AD) respectively. The classification results showed respectively for both databases an accuracy rate of 89 % and 87 %, a sensitivity rate of 85 % and 95 %, and a specificity rate of 93 % and 84 %. Likewise, the area under curve (A _z ) of the Receiver Operating Characteristic (ROC) curve was 0.93 for both databases.

Risk prediction for oral potentially malignant disorders using fuzzy analysis of cytomorphological and autofluorescence alterations in habitual smokers

AIM

This study aims to develop a novel noninvasive method for early cancer trend diagnosis in habitual smokers by corroborating cytomorphological and autofluorescence alterations.

MATERIALS & METHODS

A total of 120 subjects were included and categorized into nonsmoker, smoker and clinically diagnosed oral potentially malignant disorder (OPMD) patients. Oral exfoliative epithelial cells were studied through differential interference contrast and fluorescence microscopy. Fuzzy trend analysis was performed using measured parameters for determining the risk factors among smokers.

RESULTS

The risk assessment in this study showed a positive correlation of smoking duration with early cancer risk factors with a correlation co-efficient of 0.86.

CONCLUSION

Alterations in cellular morphology and autofluorescence intensities showed positive correlation with OPMD. The present study will benefit to investigate early prediction of OPMD among susceptible individuals.

Spatiotemporal Rank Filtering Improves Image Quality Compared to Frame Averaging in 2-Photon Laser Scanning Microscopy

Live imaging of biological specimens using optical microscopy is limited by tradeoffs between spatial and temporal resolution, depth into intact samples, and phototoxicity. Two-photon laser scanning microscopy (2P-LSM), the gold standard for imaging turbid samples in vivo, has conventionally constructed images with sufficient signal-to-noise ratio (SNR) generated by sequential raster scans of the focal plane and temporal integration of the collected signals. Here, we describe spatiotemporal rank filtering, a nonlinear alternative to temporal integration, which makes more efficient use of collected photons by selectively reducing noise in 2P-LSM images during acquisition. This results in much higher SNR while preserving image edges and fine details. Practically, this allows for at least a four fold decrease in collection times, a substantial improvement for time-course imaging in biological systems.

Coupled Learning for Facial Deblur

Blur in facial images significantly impedes the efficiency of recognition approaches. However, most existing blind deconvolution methods cannot generate satisfactory results due to their dependence on strong edges, which are sufficient in natural images but not in facial images. In this paper, we represent point spread functions (PSFs) by the linear combination of a set of pre-defined orthogonal PSFs, and similarly, an estimated intrinsic (EI) sharp face image is represented by the linear combination of a set of pre-defined orthogonal face images. In doing so, PSF and EI estimation is simplified to discovering two sets of linear combination coefficients, which are simultaneously found by our proposed coupled learning algorithm. To make our method robust to different types of blurry face images, we generate several candidate PSFs and EIs for a test image, and then, a non-blind deconvolution method is adopted to generate more EIs by those candidate PSFs. Finally, we deploy a blind image quality assessment metric to automatically select the optimal EI. Thorough experiments on the facial recognition technology database, extended Yale face database B, CMU pose, illumination, and expression (PIE) database, and face recognition grand challenge database version 2.0 demonstrate that the proposed approach effectively restores intrinsic sharp face images and, consequently, improves the performance of face recognition.

Edge-Aware BMA Filters

There has been continuous research in edge-aware filters which have found many applications in computer vision and image processing. In this paper, we propose a principled-approach for the development of edge-aware filters. The proposed approach is based on two well-established principles: 1) optimal parameter estimation and 2) Bayesian model averaging (BMA). Using this approach, we formulate the problem of filtering a pixel in a local pixel patch as an optimal estimation problem. Since a pixel belongs to multiple local patches, there are multiple estimates of the same pixel. We combine these estimates into a final estimate using BMA. We demonstrate the versatility of this approach by developing a family of BMA filters based on different settings of cost functions and log-likelihood and log-prior functions. We also present a new interpretation of the guided filter and develop a BMA guided filter which includes the guided filter as a special case. We show that BMA filters can produce similar smoothing results as those of the state-of-the-art edge-aware filters. Two BMA filters are computationally as efficient as the guided filter which is one of the fastest edge-aware filters. We also demonstrate that the BMA guided filter is better than the guided filter in preserving sharp edges. A new feature of the BMA guided filter is that the filtered image is similar to that produced by a clustering process.

Blind prediction of natural video quality

We propose a blind (no reference or NR) video quality evaluation model that is nondistortion specific. The approach relies on a spatio-temporal model of video scenes in the discrete cosine transform domain, and on a model that characterizes the type of motion occurring in the scenes, to predict video quality. We use the models to define video statistics and perceptual features that are the basis of a video quality assessment (VQA) algorithm that does not require the presence of a pristine video to compare against in order to predict a perceptual quality score. The contributions of this paper are threefold. 1) We propose a spatio-temporal natural scene statistics (NSS) model for videos. 2) We propose a motion model that quantifies motion coherency in video scenes. 3) We show that the proposed NSS and motion coherency models are appropriate for quality assessment of videos, and we utilize them to design a blind VQA algorithm that correlates highly with human judgments of quality. The proposed algorithm, called video BLIINDS, is tested on the LIVE VQA database and on the EPFL-PoliMi video database and shown to perform close to the level of top performing reduced and full reference VQA algorithms.

Image enhancement using the hypothesis selection filter: theory and application to JPEG decoding

We introduce the hypothesis selection filter (HSF) as a new approach for image quality enhancement. We assume that a set of filters has been selected a priori to improve the quality of a distorted image containing regions with different characteristics. At each pixel, HSF uses a locally computed feature vector to predict the relative performance of the filters in estimating the corresponding pixel intensity in the original undistorted image. The prediction result then determines the proportion of each filter used to obtain the final processed output. In this way, the HSF serves as a framework for combining the outputs of a number of different user selected filters, each best suited for a different region of an image. We formulate our scheme in a probabilistic framework where the HSF output is obtained as the Bayesian minimum mean square error estimate of the original image. Maximum likelihood estimates of the model parameters are determined from an offline fully unsupervised training procedure that is derived from the expectation-maximization algorithm. To illustrate how to apply the HSF and to demonstrate its potential, we apply our scheme as a post-processing step to improve the decoding quality of JPEG-encoded document images. The scheme consistently improves the quality of the decoded image over a variety of image content with different characteristics. We show that our scheme results in quantitative improvements over several other state-of-the-art JPEG decoding methods.

Blind image quality assessment: a natural scene statistics approach in the DCT domain

We develop an efficient, general-purpose, blind/noreference image quality assessment (NR-IQA) algorithm using a natural scene statistics (NSS) model of discrete cosine transform (DCT) coefficients. The algorithm is computationally appealing, given the availability of platforms optimized for DCT computation. The approach relies on a simple Bayesian inference model to predict image quality scores given certain extracted features. The features are based on an NSS model of the image DCT coefficients. The estimated parameters of the model are utilized to form features that are indicative of perceptual quality. These features are used in a simple Bayesian inference approach to predict quality scores. The resulting algorithm, which we name BLIINDS-II, requires minimal training and adopts a simple probabilistic model for score prediction. Given the extracted features from a test image, the quality score that maximizes the probability of the empirically determined inference model is chosen as the predicted quality score of that image. When tested on the LIVE IQA database, BLIINDS-II is shown to correlate highly with human judgments of quality, at a level that is competitive with the popular SSIM index.

Efficient stereoscopic ranging via stochastic sampling of match quality

We present an efficient method that computes dense stereo correspondences by stochastically sampling match quality values. Nonexhaustive sampling facilitates the use of quality metrics that take unique values at noninteger disparities. Depth estimates are iteratively refined with a stochastic cooperative search by perturbing the estimates, sampling match quality, and reweighting and aggregating the perturbations. The approach gains significant efficiencies when applied to video, where initial estimates are seeded using information from the previous pair in a novel application of the Z-buffering algorithm. This significantly reduces the number of search iterations required. We present a quantitative accuracy evaluation wherein the proposed method outperforms a microcanonical annealing approach by Barnard and a cooperative approach by Zitnick and Kanade , while using fewer match quality evaluations than either. The approach is shown to have more attractive memory usage and scaling than alternatives based on exhaustive sampling.

A novel image smoothing filter using membership function

This paper presents a new class of image noise smoothing algorithms utilizing the membership information of the neighboring pixels. The basic idea of this method is to compute the smoothed output using neighboring pixels from the same cluster to avoid image blurring. A fuzzy c-means algorithm is first applied to the image to separate the image pixels into a certain number of clusters. A membership function is defined as the probability that a pixel belongs to a cluster. The proposed method uses this membership function as a weight to calculate the weighted sum of the pixel values from its neighboring pixels. The results of the application of this algorithm to various images show that it can smooth images with edge enhancement. The smoothness of the resultant images can be controlled by the cluster number and window size.

Moving window-based double Haar wavelet transform for image processing

Image denoising is a lively research field. The classical nonlinear filters used for image denoising, such as median filter, are based on a local analysis of the pixels within a moving window. Recently, the research of image denoising has been focused on the wavelet domain. Compared to the classical nonlinear filters, it is based on a global multiscale analysis of images. Apparently, the wavelet transform can be embedded in a moving window. Thus, a moving window-based local multiscale analysis is obtained. In this paper, based on the Haar wavelet, a class of nonorthogonal multi-channel filter bank with its corresponding wavelet shrinkage called Lee shrinkage is derived. As a special case of this filter bank, the double Haar wavelet transform is introduced. Examples show that it is suitable for a moving window-based local multiscale analysis used for image denoising, edge detection, and edge enhancement.

A universal noise removal algorithm with an impulse detector

We introduce a local image statistic for identifying noise pixels in images corrupted with impulse noise of random values. The statistical values quantify how different in intensity the particular pixels are from their most similar neighbors. We continue to demonstrate how this statistic may be incorporated into a filter designed to remove additive Gaussian noise. The result is a new filter capable of reducing both Gaussian and impulse noises from noisy images effectively, which performs remarkably well, both in terms of quantitative measures of signal restoration and qualitative judgements of image quality. Our approach is extended to automatically remove any mix of Gaussian and impulse noise.

An optimal nonlinear extension of linear filters based on distributed arithmetic

Distributed arithmetic (DA)-based implementation of linear filters relies on the linear nature of this operation and has been suggested as a multiplication free solution. In this work, we introduce a nonlinear extension of linear filters optimizing under mean-square error criterion the memory function [(MF) multivariate Boolean function with not only binary output] which is in the core of DA based implementation. Such an extension will improve the filtering of noise which may contain non-Gaussian components without increasing the complexity of implementation. Experiments on real images have shown the superiority of the proposed filters over the optimal linear filters. Different versions of these filters are also considered for an impulsive noise removal, faster processing, and filtering using large input data windows.

An analog neural network implementation in fixed time of adjustable-order statistic filters and applications

In this paper, we show a neural network implementation in fixed time of adjustable order statistic filters, including sorting, and adaptive-order statistic filters. All these networks accept an array of N numbers Xi = S(Xi)M(Xi)2E(Xi) as input (where S(Xi) is the sign of Xi, M(Xi) is the mantissa normalized to m digits, and Ex is the exponent) and employ two kinds of neurons, the linear and the threshold-logic neurons, with only integer weights (most of the weights being just +1 or -1) and integer threshold. Therefore, this will greatly facilitate the actual hardware implementation of the proposed neural networks using currently available very large scale integration technology. An application of using minimum filter in implementing a special neural network model neural network classifier (NNC) is given. With a classification problem of l classes C1, C2,.. ., C1, NNC classifies in fixed time an unknown vector to one class using a minimum-distance classification technique.

Comparison of hemodynamic response nonlinearity across primary cortical areas

Hemodynamic responses to auditory and visual stimuli and motor tasks were assessed for the nonlinearity of response in each of the respective primary cortices. Five stimulus or task durations were used (1, 2, 4, 8, and 16 s), and five male subjects (aged 19 +/- 1.9 years) were imaged. Two tests of linearity were conducted. The first test consisted of using BOLD responses to short stimuli to predict responses to longer stimuli. The second test consisted of fitting ideal impulse response functions to the observed responses for each event duration. Both methods show that the extent of the nonlinearity varies across cortices. Results for the second method indicate that the hemodynamic response is nonlinear for stimuli less than 10 s in the primary auditory cortex, nonlinear for tasks less than 7 s in the primary motor cortex, and nonlinear for stimuli less than 3 s in the primary visual cortex. In addition, neural adaptation functions were characterized that could model the observed nonlinearities.

Generalization of median root prior reconstruction

Penalized iterative algorithms for image reconstruction in emission tomography contain conditions on which kind of images are accepted as solutions. The penalty term has commonly been a function of pairwise pixel differences in the activity in a local neighborhood, such that smooth images are favored. Attempts to ensure better edge and detail preservation involve difficult tailoring of parameter values or the penalty function itself. The previously introduced median root prior (MRP) favors locally monotonic images. MRP preserves sharp edges while reducing locally nonmonotonic noise at the same time. Quantitative properties of MRP are good, because differences in the neighboring pixel values are not penalized as such. The median is used as an estimate for a penalty reference, against which the pixel value is compared when setting the penalty. In order to generalize the class of MRP-type of priors, the standard median was replaced by other order statistic operations, the L and finite-impluse-response median hybrid (FMH) filters. They allow for smoother appearance as they apply linear weighting together with robust nonlinear operations. The images reconstructed using the new MRP-L and MRP-FMH priors are visually more conventional. Good quantitative properties of MRP are not significantly altered by the new priors.

Development of the local maximum variety of ranked-order filters

A form of ranked-order filters is introduced as the local maximum filter. The construction of the local maximum filter is described, followed by a discussion of its function and some of its more important properties, and an example application of a two-dimensional local maximum filter is provided to illustrate the detection of single-pixel targets against a cloud clutter background. The closing discussion provides a mathematical development of the filter.

Topological median filters

This paper describes the definition and testing of a new type of median filter for images. The topological median filter implements some existing ideas and some new ideas on fuzzy connectedness to improve, over a conventional median filter, the extraction of edges in noise. The concept of alpha-connectivity is defined and used to create an algorithm for computing the degree of connectedness of a pixel to all the other pixels in an arbitrary neighborhood. The resulting connectivity map of the neighborhood effectively disconnects peaks in the neighborhood that are separated from the center pixel by a valley in the brightness topology. The median of the connectivity map is an estimate of the median of the peak or plateau to which the center pixel belongs. Unlike the conventional median filter, the topological median is relatively unaffected by disconnected features in the neighborhood of the center pixel. Four topological median filters are defined. Qualitative and statistical analyses of the four filters are presented. It is demonstrated that edge detection can be more accurate on topologically median filtered images than on conventionally median filtered images.

Hybrid order statistic filter and its application to image restoration

We introduce a new nonlinear filter for signal and image restoration, the hybrid order statistic (HOS) filter. Because it exploits both rank- and spatial-order information, the HOS realizes the advantages of nonlinear filters in edge preservation and reduction of impulsive noise components while retaining the ability of the linear filter to suppress Gaussian noise. We show that the HOS filter exhibits improved performance over both the linear Wiener and the nonlinear L filters in reducing mean-squared error in the presence of contaminated Gaussian noise. In many cases it also performs favorably compared with the Ll and rank-conditioned rank selection filters.

Developments in generalized ranked-order filters

A general formulation of ranked-order filters is developed in two parts: part 1, signal-to-noise-ratio analysis and part 2, construction and analysis of a ranked-order-filter function based on a mathematical logic approach. The filter function is analyzed to define the structure of filter roots for one-dimensional (1-D) and two-dimensional (2-D) window filters as data patterns that are invariant of the filter. The 1-D and 2-D coded window filters defined for roots of repeated patterns of binary data are defined and analyzed. The analysis concludes with an application of the coded window filter to a computer-generated 2-D noisy image containing a binary pattern and an application for feature extraction by a 2-D filter constrained by a predicate function to select only fixed-point root data structures.

Nonlinear image estimation using piecewise and local image models

We introduce a new approach to image estimation based on a flexible constraint framework that encapsulates meaningful structural image assumptions. Piecewise image models (PIMs) and local image models (LIMs) are defined and utilized to estimate noise-corrupted images, PIMs and LIMs are defined by image sets obeying certain piecewise or local image properties, such as piecewise linearity, or local monotonicity. By optimizing local image characteristics imposed by the models, image estimates are produced with respect to the characteristic sets defined by the models. Thus, we propose a new general formulation for nonlinear set-theoretic image estimation. Detailed image estimation algorithms and examples are given using two PIMs: piecewise constant (PICO) and piecewise linear (PILI) models, and two LIMs: locally monotonic (LOMO) and locally convex/concave (LOCO) models. These models define properties that hold over local image neighborhoods, and the corresponding image estimates may be inexpensively computed by iterative optimization algorithms. Forcing the model constraints to hold at every image coordinate of the solution defines a nonlinear regression problem that is generally nonconvex and combinatorial. However, approximate solutions may be computed in reasonable time using the novel generalized deterministic annealing (GDA) optimization technique, which is particularly well suited for locally constrained problems of this type. Results are given for corrupted imagery with signal-to-noise ratio (SNR) as low as 2 dB, demonstrating high quality image estimation as measured by local feature integrity, and improvement in SNR.

MRL-filters: a general class of nonlinear systems and their optimal design for image processing

A class of morphological/rank/linear (MRL)-filters is presented as a general nonlinear tool for image processing. They consist of a linear combination between a morphological/rank filter and a linear filter. A gradient steepest descent method is proposed to optimally design these filters, using the averaged least mean squares (LMS) algorithm. The filter design is viewed as a learning process, and convergence issues are theoretically and experimentally investigated. A systematic approach is proposed to overcome the problem of nondifferentiability of the nonlinear filter component and to improve the numerical robustness of the training algorithm, which results in simple training equations. Image processing applications in system identification and image restoration are also presented, illustrating the simplicity of training MRL-filters and their effectiveness for image/signal processing.

An adaptive order-statistic noise filter for gamma-corrected image sequences

Original video signals are often corrupted by a certain amount of noise originating from the camera electronics. As a result of the gamma correction in cameras, the observed noise is signal dependent. We present a spatio-temporal order-statistic (OS) noise filter that takes into account the gamma correction in the camera. The calculation of the filter coefficients requires higher-order order-statistics (HOOS) of the noise process. We make use of a range test (RT) to determine locally from which neighboring signal values an estimate should be formed. The noise filter that we arrive at is adaptive and computationally efficient.

A robust approach to image enhancement based on fuzzy logic

In this paper, we propose a robust approach to image enhancement based on fuzzy logic that addresses the seemingly conflicting goals of image enhancement: (i) removing impulse noise, (ii) smoothing out nonimpulse noise, and (iii) enhancing (or preserving) edges and certain other salient structures. We derive three different filters for each of the above three tasks using the weighted (or fuzzy) least squares (LS) method, and define the criteria for selecting each of the three filters. The criteria are based on the local context, and they constitute the antecedent clauses of the fuzzy rules. The overall result of the fuzzy rule-based system is the combination of the results of the individual filters, where each result contributes to the degree that the corresponding antecedent clause is satisfied. This approach gives us a powerful and flexible image enhancement paradigm. Results of the proposed method on several types of images are compared with those of other standard techniques.

Adaptive LMS L-filters for noise suppression in images

Several adaptive least mean squares (LMS) L-filters, both constrained and unconstrained ones, are developed for noise suppression in images and compared in this paper. First, the location-invariant LMS L-filter for a nonconstant signal corrupted by zero-mean additive white noise is derived. It is demonstrated that the location-invariant LMS L-filter can be described in terms of the generalized linearly constrained adaptive processing structure proposed by Griffiths and Jim (1982). Subsequently, the normalized and the signed error LMS L-filters are studied. A modified LMS L-filter with nonhomogeneous step-sizes is also proposed in order to accelerate the rate of convergence of the adaptive L-filter. Finally, a signal-dependent adaptive filter structure is developed to allow a separate treatment of the pixels that are close to the edges from the pixels that belong to homogeneous image regions.

A new efficient approach for the removal of impulse noise from highly corrupted images

A new framework for removing impulse noise from images is presented in which the nature of the filtering operation is conditioned on a state variable defined as the output of a classifier that operates on the differences between the input pixel and the remaining rank-ordered pixels in a sliding window. As part of this framework, several algorithms are examined, each of which is applicable to fixed and random-valued impulse noise models. First, a simple two-state approach is described in which the algorithm switches between the output of an identity filter and a rank-ordered mean (ROM) filter. The technique achieves an excellent tradeoff between noise suppression and detail preservation with little increase in computational complexity over the simple median filter. For a small additional cost in memory, this simple strategy is easily generalized into a multistate approach using weighted combinations of the identity and ROM filter in which the weighting coefficients can be optimized using image training data. Extensive simulations indicate that these methods perform significantly better in terms of noise suppression and detail preservation than a number of existing nonlinear techniques with as much as 40% impulse noise corruption. Moreover, the method can effectively restore images corrupted with Gaussian noise and mixed Gaussian and impulse noise. Finally, the method is shown to be extremely robust with respect to the training data and the percentage of impulse noise.

Noise reduction of image sequences using motion compensation and signal decomposition

In this paper, a new spatio-temporal filtering method for removing noise from image sequences is proposed. This method combines the use of motion compensation and signal decomposition to account for the effects of object motion. Because of object motion, image sequences are temporally nonstationary, which requires the use of adaptive filters. By motion compensating the sequence prior to filtering, nonstationarities, i.e., parts of the signal that are momentarily not stationary, can be reduced significantly. However, since not all nonstationarities can be accounted for by motion, a motion-compensated signal still contains nonstationarities. An adaptive algorithm based on order statistics is described that decomposes the motion-compensated signal into a noise-free nonstationary part and a noisy stationary part. An RLS filter is then used to filter the noise from the stationary signal. Our new method is experimentally compared with various noise filtering approaches from literature.

A multidimensional nonlinear edge-preserving filter for magnetic resonance image restoration

The paper presents a multidimensional nonlinear edge-preserving filter for restoration and enhancement of magnetic resonance images (MRI). The filter uses both interframe (parametric or temporal) and intraframe (spatial) information to filter the additive noise from an MRI scene sequence. It combines the approximate maximum likelihood (equivalently, least squares) estimate of the interframe pixels, using MRI signal models, with a trimmed spatial smoothing algorithm, using a Euclidean distance discriminator to preserve partial volume and edge information. (Partial volume information is generated from voxels containing a mixture of different tissues.) Since the filter's structure is parallel, its implementation on a parallel processing computer is straightforward. Details of the filter implementation for a sequence of four multiple spin-echo images is explained, and the effects of filter parameters (neighborhood size and threshold value) on the computation time and performance of the filter is discussed. The filter is applied to MRI simulation and brain studies, serving as a preprocessing procedure for the eigenimage filter. (The eigenimage filter generates a composite image in which a feature of interest is segmented from the surrounding interfering features.) It outperforms conventional pre and post-processing filters, including spatial smoothing, low-pass filtering with a Gaussian kernel, median filtering, and combined vector median with average filtering.

Nonlinear scale-space filtering and multiresolution system

We derive and demonstrate a nonlinear scale-space filter and its application in generating a nonlinear multiresolution system. For each datum in a signal, a neighborhood of weighted data is used for clustering. The cluster center becomes the filter output. The filter is governed by a single scale parameter that dictates the spatial extent of nearby data used for clustering. This, together with the local characteristic of the signal, determines the scale parameter in the output space, which dictates the influences of these data on the output. This filter is thus adaptive and data driven. It provides a mechanism for (a) removing impulsive noise, (b) improved smoothing of nonimpulsive noise, and (c) preserving edges. Comparisons with Gaussian scale-space filtering and median filters are made using real images. Using the architecture of the Laplacian pyramid and this nonlinear filter for interpolation, we construct a nonlinear multiresolution system that has two features: (1) edges are well preserved at low resolutions, and (2) difference signals are small and spatially localized. This filter implicitly presents a new mechanism for detecting discontinuities differing from techniques based on local gradients and line processes. This work shows that scale-space filtering, nonlinear filtering, and scale-space clustering are closely related and provides a framework within which further image processing, image coding, and computer vision problems can be investigated.

A robust nonlinear filter for image restoration

A class of nonlinear regression filters based on robust estimation theory is introduced. The goal of the filtering is to recover a high-quality image from degraded observations. Models for desired image structures and contaminating processes are employed, but deviations from strict assumptions are allowed since the assumptions on signal and noise are typically only approximately true. The robustness of filters is usually addressed only in a distributional sense, i.e., the actual error distribution deviates from the nominal one. In this paper, the robustness is considered in a broad sense since the outliers may also be due to inappropriate signal model, or there may be more than one statistical population present in the processing window, causing biased estimates. Two filtering algorithms minimizing a least trimmed squares criterion are provided. The design of the filters is simple since no scale parameters or context-dependent threshold values are required. Experimental results using both real and simulated data are presented. The filters effectively attenuate both impulsive and nonimpulsive noise while recovering the signal structure and preserving interesting details.

A structure for adaptive order statistics filtering

In applications such as smoothing and enhancement of images, adaptive filtering techniques offer the flexibility needed for good performance with non-stationary observations. Many adaptive schemes can be based on the idea of determining the local statistics of the signal through appropriate tests on the data, to aid in the selection of a filtering procedure that is suited to the data. In the paper, the authors consider decision-directed or data-dependent adaptive filtering schemes that are based on order statistics. A general formulation for such a class of adaptive order statistics filters is presented. Approximate statistical performance analysis, especially in the presence of edges, may be carried out for this entire class of filters. The authors give examples of some existing filters that fit into this framework. The formulation also accommodates filters that employ multiple windows in their operation. To illustrate the potential of this class of multiple window (MW) filters, they construct and analyze simple filters, like the triple window median (TW-MED) and the triple window median of means (TW-MOM) filters, that are shown to yield useful performance. The class of mean-median hybrid (MMH) filters is also presented as a simple example which may be extended to give interesting performance.

Tree-structured nonlinear filters in digital mammography

A new class of nonlinear filters with more robust characteristics for noise suppression and detail preservation is proposed for processing digital mammographic images. The new algorithm consists of two major filtering blocks: (a) a multistage tree-structured filter for image enhancement that uses central weighted median filters as basic sub-filtering blocks and (b) a dispersion edge detector. The design of the algorithm also included the use of linear and curved windows to determine whether variable shape windowing could improve detail preservation. First, the noise-suppressing properties of the tree-structured filter were compared to single filters, namely the median and the central weighted median with conventional square and variable shape adaptive windows; simulated images were used for this purpose. Second, the edge detection properties of the tree-structured filter cascaded with the dispersion edge detector were compared to the performance of the dispersion edge detector alone, the Sobel operator, and the single median filter cascaded with the dispersion edge detector. Selected mammographic images with representative biopsy-proven malignancies were processed with all methods and the results were visually evaluated by an expert mammographer. In all applications, the proposed filter suggested better detail preservation, noise suppression, and edge detection than all other approaches and it may prove to be a useful tool for computer-assisted diagnosis in digital mammography.

Multistage enhancement of surface recordings of canine gastric electrical signals

This article describes a multistage signal processing scheme to enhance the quality of canine gastric signals recorded from the abdominal surface. The scheme involves a cascade application of linear prediction followed by a nonlinear processing known as alpha-TM filtering. The linear prediction is used to separate, in the minimum mean square error sense, the slow wave from other uncorrelated interference signals. We make novel use of the order versus frequency response characteristics of linear predictors to achieve this separation. The nonlinear filtering is used to suppress the residual wide band impulsive noise. Our studies have indicated that such an optimized signal enhancement scheme produces a clean time domain signal, which is easy to interpret visually. It not only preserves the periodicity of the slow wave, but also seems to track any irregularities in the periods. We believe that this last feature, namely the potential to track nonstationarities in the signal, is the main contribution of our approach.

OSNet: a neural network implementation of order statistic filters

A dedicated neural network model called OSNet which finds the kth largest element in an array of real numbers is proposed. Its overall processing time is constant irrespective of the number of elements in the array and is four times the processing time of a single neuron. Networks of this kind may be used as building blocks for hardware implementation of order statistic filters. Examples of using OSNet for implementing various order statistic filters and for sorting are shown.

Combination median filter

A detail- and structure-preserving smoothing filter is introduced. It is called the combination median filter as it uses directional median, multilevel median, and median filters to smooth different regions of the image. The decision about the region is made by robust Dixon's r-test which is well known in statistics for outlier detection. The threshold value of Dixon's test can be kept constant. As a result, the filtering algorithm operates like a quasi-nonadaptive filter, and no computation of local statistics is involved. Some properties of the filter as well as detailed experimental results that demonstrate its superior performance are presented.

Robust adaptive parameter estimators in arrhythmia detection

The authors consider the statistical analysis of threshold crossing intervals, as applied to estimation of tachycardia rates from intracavitary electrograms. The authors developed a class of robust algorithms designed to produce minimum variance estimates for tachycardia rates. The authors formulated the algorithms using order statistic filters, and obtained the minimum variance unbiased order statistic estimator. The potential gain in efficiency achieved by this approach is demonstrated via a representative example. The results indicated that the order statistics operator can produce dramatic reductions for typical errors in error variance as compared to linear estimators.

Analysis of order statistic filters applied to ultrasonic flaw detection using split-spectrum processing

Split-spectrum processing of broadband ultrasonic signals coupled with order statistic filtering has proven to be effective in improving the flaw-to-clutter ratio of backscattered signals. It is shown that an optimal rank can be obtained with a prior knowledge of flaw-to-clutter ratio and the underlying distributions. The order statistic filter performs well where the flaw and clutter echoes have good statistical separation in a given quantile region representing a particular rank (e.g. minimum, median, maximum). Order statistic filters are analyzed for the situation in which the observations do not contain equivalent statistical information. Experimental and simulated results are presented to show how effectively the order statistic filter can utilize information contained in different frequency bands to improve flaw detection.