In this paper, we propose an adaptive sub-layer compensation (ASLC) based facial edge detection for human emotion recognition. We modify the Marr-Hildreth edge detector with Wiener filtering, sub-layer compensation and hysteresis analysis to compensate the negative effects of the Laplacian of Gaussian (LoG) operator such as image degradation, high response to unwanted details, and disconnected edges. We investigate a Discriminative Isomap (D-Isomap) based approach that combines the ASLC feature and deformable elastic body spline (EBS) feature for the final decision. RML emotion database and Cohn-Kanade database are used for the experiment and the results demonstrate the effectiveness of the proposed method.

In this paper, we investigate kernel based methods for multimodal information analysis and fusion. We introduce a novel approach, kernel cross-modal factor analysis, which identifies the optimal transformations that are capable of representing the coupled patterns between two different subsets of features by minimizing the Frobenius norm in the transformed domain. The kernel trick is utilized for modeling the nonlinear relationship between two multidimensional variables. We examine and compare with kernel canonical correlation analysis which finds projection directions that maximize the correlation between two modalities, and kernel matrix fusion which integrates the kernel matrices of respective modalities through algebraic operations. The performance of the introduced method is evaluated on an audiovisual based bimodal emotion recognition problem. We first perform feature extraction from the audio and visual channels respectively. The presented approaches are then utilized to analyze the cross-modal relationship between audio and visual features. A hidden Markov model is subsequently applied for characterizing the statistical dependence across successive time segments, and identifying the inherent temporal structure of the features in the transformed domain. The effectiveness of the proposed solution is demonstrated through extensive experimentation.

This paper presents a novel approach for multimodal information fusion. The proposed method is based on kernel cross-modal factor analysis (KCFA), in which the optimal transformations that represent the coupled patterns between two different subsets of features are identified by minimizing the Frobenius norm in the transformed domain. It generalizes the linear cross-modal factor analysis (CFA) method via the kernel trick to model the nonlinear relationship between two multidimensional variables. The effectiveness of the introduced solution is demonstrated through experimentation on an audiovisual based emotion recognition problem. Experimental results show that the proposed approach outperforms the concatenation based feature level fusion, the linear CFA, as well as the canonical correlation analysis (CCA) and kernel CCA methods.

In this paper, we introduce a new method for audiovisual based multimodal emotion recognition. The proposed method identifies the optimal transformations that are capable of representing the coupled patterns between audio and visual information through cross-modal association. Specifically, kernel machine technique is utilized for capturing the nonlinear relationship between two different subsets of features. A hidden Markov model is subsequently applied for characterizing the statistical dependence across successive time segments, and identifying the inherent temporal structure of the features in the transformed domain. Information fusion at the feature and score levels are examined and compared. The effectiveness of the introduced solution is demonstrated through extensive experimentation.

This paper outlines a new method for automatic detection of microcalcification clusters in mammograms. The presence of microcalcification clusters, which appear as small bright spots in mammographic images, is considered a very important sign in breast cancer diagnosis. However, such clusters can be hard to detect due to their size and low contrast from surrounding normal tissue. This work presents a new fuzzy based method for the detection of microcalcification clusters. The proposed method consists of four major steps. First, the breast area is extracted. Then a powerful fuzzy contrast adaptation is employed to highlight the contrast of the microcalcification spots. Next, a thresholding method based on fuzzy sets type II is used to extract the candidate points. Finally, the features of these points are extracted and a support vector machine classifier distinguishes the location of real microcalcifications. During these steps, the selection of appropriate parameters is performed based on local image characteristics. The results are promising and show that this method can detect microcalcifications effectively, making it useful towards computer-aided breast cancer diagnosis.

This paper presents a novel scheme for mass detection in mammography images. In this method, a mean shift-based algorithm is used to cluster pixels in the image. The extraction of the breast border is the first step. Image pixels are then clustered using a mean shift algorithm that employs intensity information to extract a set of high density points in the feature space. This is followed by further stages involving mode fusion. Due to its non-parametric nature, mean shift algorithm can work effectively with non-convex regions resulting in better candidates for a reliable segmentation. The proposed method has been validated on standard datasets and the results show that this method can detect masses in mammography images, making it useful for breast cancer detection systems.

This work presents an image reconstruction technique for noise suppression in FLAIR MRI with white matter lesions (WML). The technique utilizes a fuzzy edge estimate to initially localize edge information. Edge and intensity information are coupled through the conditional expectation operator, resulting in a robust and global description of the edge content in the image. As this global measure separates noise and useful edge information, a threshold is used to suppress the irrelevant details and integration is used to reconstruct the “noisefree” image. The threshold is automatically determined based on an objective function that minimizes noise (within class scatter) while maximizing contrast between the WML and brain tissue classes (between class variance). The result is an edge preserving smoothing filter, since the image is reconstructed based on the edge map. The proposed method was compared to the bilateral filter and was found to provide on average a 44.39% increase of noise attenuation in flat regions (smoothness) and a 34.14% increase in edge amplification (enhancement).

Indoor positioning is an enabling technology for delivery of location-based services in mobile computing environments. This paper proposes a positioning solution using received signal strength in indoor wireless local area networks. In this application, an explicit measurement equation and the corresponding noise statistics are unknown because of the complexity of the indoor propagation channel. To address these challenges, we introduce a new state-space Bayesian filter: the nonparametric information (NI) filter. This filter effectively tracks motion in situations where the Kalman filter and its variants are inapplicable, while maintaining a computational complexity comparable to that of the Kalman filter. To deal with the noisy nature of the indoor propagation environment, the NI filter is used in the design of an intelligent dynamic WLAN tracking system. The system anticipates future position values and adapts its sensing and estimation parameters accordingly. Our experimental results conducted on measurements from a real office environment indicate that the combination of the intelligent design and the NI filter results in significant improvements over the Kalman and particle filters.

This paper proposes an uncorrelated multilinear discriminant analysis (UMLDA) framework for the recognition of multidimensional objects, known as tensor objects. Uncorrelated features are desirable in recognition tasks since they contain minimum redundancy and ensure independence of features. The UMLDA aims to extract uncorrelated discriminative features directly from tensorial data through solving a tensor-to-vector projection. The solution consists of sequential iterative processes based on the alternating projection method, and an adaptive regularization procedure is incorporated to enhance the performance in the small sample size (SSS) scenario. A simple nearest-neighbor classifier is employed for classification. Furthermore, exploiting the complementary information from differently initialized and regularized UMLDA recognizers, an aggregation scheme is adopted to combine them at the matching score level, resulting in enhanced generalization performance while alleviating the regularization parameter selection problem. The UMLDA-based recognition algorithm is then empirically shown on face and gait recognition tasks to outperform four multilinear subspace solutions (MPCA, DATER, GTDA, TR1DA) and four linear subspace solutions (Bayesian, LDA, ULDA, R-JD-LDA).

This paper proposes an uncorrelated multilinear principal component analysis (UMPCA) algorithm for unsupervised subspace learning of tensorial data. It should be viewed as a multilinear extension of the classical principal component analysis (PCA) framework. Through successive variance maximization, UMPCA seeks a tensor-to-vector projection (TVP) that captures most of the variation in the original tensorial input while producing uncorrelated features. The solution consists of sequential iterative steps based on the alternating projection method. In addition to deriving the UMPCA framework, this work offers a way to systematically determine the maximum number of uncorrelated multilinear features that can be extracted by the method. UMPCA is compared against the baseline PCA solution and its five state-of-the-art multilinear extensions, namely two-dimensional PCA (2DPCA), concurrent subspaces analysis (CSA), tensor rank-one decomposition (TROD), generalized PCA (GPCA), and multilinear PCA (MPCA), on the tasks of unsupervised face and gait recognition. Experimental results included in this paper suggest that UMPCA is particularly effective in determining the low-dimensional projection space needed in such recognition tasks.

Positioning in wireless networks has gained significant ground as an enabling technology for various applications such as event detection and context awareness. Since these positioning systems rely on radio features to locate a mobile, they are susceptible to non-cryptographic attacks resulting from malicious alteration of the propagation environment. This paper proposes a sensor selection scheme for increasing the resilience of fingerprinting-based positioning systems to RSS-based attacks in the context of wireless local area networks (WLAN). A distributed positioning scheme is proposed whereby an estimate is obtained from each WLAN access point (AP). Sensor selection is performed based on a nonparametric estimate of the Fisher information. Experimental results indicate superior performance compared to existing methods and graceful performance degradation in presence of RSS attacks.

Tensorial data are frequently encountered in various machine learning tasks today and dimensionality reduction is one of their most important applications. This paper extends the classical principal component analysis (PCA) to its multilinear version by proposing a novel unsupervised dimensionality reduction algorithm for tensorial data, named as uncorrelated multilinear PCA (UMPCA). UMPCA seeks a tensor-to-vector projection that captures most of the variation in the original tensorial input while producing uncorrelated features through successive variance maximization. We evaluate the UMPCA on a second-order tensorial problem, face recognition, and the experimental results show its superiority, especially in low-dimensional spaces, through the comparison with three other PCA-based algorithms.

This paper introduces a multilinear principal component analysis (MPCA) framework for tensor object feature extraction. Objects of interest in many computer vision and pattern recognition applications, such as 2-D/3-D images and video sequences are naturally described as tensors or multilinear arrays. The proposed framework performs feature extraction by determining a multilinear projection that captures most of the original tensorial input variation. The solution is iterative in nature and it proceeds by decomposing the original problem to a series of multiple projection subproblems. As part of this work, methods for subspace dimensionality determination are proposed and analyzed. It is shown that the MPCA framework discussed in this work supplants existing heterogeneous solutions such as the classical principal component analysis (PCA) and its 2-D variant (2-D PCA). Finally, a tensor object recognition system is proposed with the introduction of a discriminative tensor feature selection mechanism and a novel classification strategy, and applied to the problem of gait recognition. Results presented here indicate MPCA's utility as a feature extraction tool. It is shown that even without a fully optimized design, an MPCA-based gait recognition module achieves highly competitive performance and compares favorably to the state-of-the-art gait recognizers.

The recent proliferation of Location-Based Services (LBSs) has necessitated the development of effective indoor positioning solutions. In such a context, Wireless Local Area Network (WLAN) positioning is a particularly viable solution in terms of hardware and installation costs due to the ubiquity of WLAN infrastructures. This paper examines three aspects of the problem of indoor WLAN positioning using received signal strength (RSS). First, we show that, due to the variability of RSS features over space, a spatially localized positioning method leads to improved positioning results. Second, we explore the problem of access point (AP) selection for positioning and demonstrate the need for further research in this area. Third, we present a kernelized distance calculation algorithm for comparing RSS observations to RSS training records. Experimental results indicate that the proposed system leads to a 17 percent (0.56 m) improvement over the widely used K-nearest neighbor and histogram-based methods.

This paper proposes a dynamic MMSE estimator for tracking mobile users in indoor wireless local area networks (WLAN) based on received signal strength (RSS). The method uses a training-based static estimate obtained by an adaptive kernel density estimator as the input into a Kalman Filter. Predictions from the filter are used during the next iteration to adaptively select a subset of training data, contained in a radio map, for the static estimator. Experimental results show that the combination of the Kalman filter and the adaptive radio map technique results in nearly 0.5 m (20%) improvement in root mean square location accuracy when compared to static localization

This paper presents a localized coarse-to-fine algorithm for efficient and accurate pedestrian localization and silhouette extraction for the gait challenge data sets. The coarse detection phase is simple and fast. It locates the target quickly based on temporal differences and some knowledge on the human target. Based on this coarse detection, the fine detection phase applies a robust background subtraction algorithm to the coarse target regions and the detection obtained is further processed to produce the final results. This algorithm has been tested on 285 outdoor sequences from the gait challenge data sets, with wide variety of capture conditions. The pedestrian targets are localized very well and silhouettes extracted resemble the manually labeled silhouettes closely

It is well-known that supervised learning techniques such as linear discriminant analysis (LDA) often suffer from the so called small sample size problem when apply to solve face recognition problems. This is due to the fact that in most cases, the number of training samples is much smaller than the dimensionality of the sample space. The problem becomes even more severe if only one training sample is available for each subject. In this paper, followed by the well-known unsupervised technique, kernel principal component analysis (KPCA), a novel feature selection scheme is proposed to establish a discriminant feature subspace in which the class separability is maximized. Extensive experiments performed on the FERET database indicate that the proposed scheme significantly boosts the recognition performance of the traditional KPCA solution

A small world search agent employing peer-to-peer (P2P) concepts borrowed from sociology is employed for performing image retrievals in a small world distributed media index. The Small World Indexing Method (SWIM) allows for a highly networked architecture where index information does not exist as a separate entity on a specific server, but rather is stored within the actual media objects themselves. Since each media object is only responsible for a small portion of the overall index, the loss of portions of the overall network (data objects) accounts for only a small degradation in the overall retrieval performance. Building upon previous work, the graceful degradation which is provided by the SWIM system is addressed here for retrievals which are performed using small world user agents on a large set of MPEG-7 described images.

In this paper, we propose a novel ensemble-based approach to boost performance of traditional Linear Discriminant Analysis (LDA)-based methods used in face recognition. The ensemble-based approach is based on the recently emerged technique known as "boosting". However, it is generally believed that boosting-like learning rules are not suited to a strong and stable learner such as LDA. To break the limitation, a novel weakness analysis theory is developed here. The theory attempts to boost a strong learner by increasing the diversity between the classifiers created by the learner, at the expense of decreasing their margins, so as to achieve a tradeoff suggested by recent boosting studies for a low generalization error. In addition, a novel distribution accounting for the pairwise class discriminant information is introduced for effective interaction between the booster and the LDA-based learner. The integration of all these methodologies proposed here leads to the novel ensemble-based discriminant learning approach, capable of taking advantage of both the boosting and LDA techniques. Promising experimental results obtained on various difficult face recognition scenarios demonstrate the effectiveness of the proposed approach. We believe that this work is especially beneficial in extending the boosting framework to accommodate general (strong/weak) learners.

A small world search agent employing peer-to-peer concepts borrowed from sociology is employed for performing image retrievals in a small world distributed media index. The small world indexing method (SWIM) allows for a highly networked architecture where index information does not exist as a separate entity on a specific server, but rather is stored within the actual media objects themselves. Since each media object is only responsible for a small portion of the overall index, the loss of portions of the overall network (data objects) accounts for only a small degradation in the overall retrieval performance. Building upon previous work, the graceful degradation which is provided by the SWIM system is addressed here for retrievals which are performed using small world user agents on a large set of MPEG-7 described images.

The location of mobile terminals in cellular networks is an important problem with applications in resource allocation, location sensitive browsing, and emergency communications. Finding cost effective location estimation techniques that are robust to non-line of sight (NLOS) propagation, quantization, and measurement noise is a key problem in this area. Quantized time difference of arrival (TDoA) and received signal strength (RSS) measurements can be made simultaneously by CDMA cellular networks at low cost. The different sources of errors for each measurement type cause RSS and TDoA measurements to contain independent information about mobile terminal location. This paper applies data fusion to combine the information of RSS and TDoA measurements to calculate a superior location estimate. Nonparametric estimation methods, that are robust to variations of measurement noise and quantization, are employed to calculate the location estimates. It is shown how the data fusion location estimators are robust, provide lower error than the estimators based on the individual measurements, and have low implementation cost.

In many real applications for face recognition, such as surveillance photo identification, each subject only has one image sample for training which makes many supervised learning techniques fail to apply. Furthermore, since subject appearance has large variabilities due to aging, illumination and camera viewpoints, the face images to be identified are usually different from the stored templates. In this paper, a novel solution to this problem is proposed based on the well known unsupervised methodology, eigenface. We proposed a criterion to select the eigenfaces forming a feature subspace in which the intrapersonal variation is small compared to interpersonal variation and as well as most discriminating power is retained. The selection criterion maximizes the ratio between inter and intra personal variation, and at the same time takes total inter variation into account. Extensive experimentation following the FERET evaluation protocol indicates that the proposed scheme improves significantly the recognition performance.

A practical component to an upper year electrical and computer engineering course in information engineering or multimedia systems is presented. Developed to introduce students to image indexing and retrieval, the experiment uses a Microsoft Windows binary of the freely available MPEG-7 Experimentation Model from ISO for image description. The experiment is broken into three interdependent stages spanning a three week period. Each stage demonstrates various concepts associated with image indexing and retrieval. Students require C/C++ programming skills to parse binary encoded MPEG-7 descriptions, output ASCII description data, and perform retrievals on indexed information. The laboratory is distributed in nature, and each participating student is made to work on a small, mutually exclusive image set. The experiment drives home the urgency of the media indexing problem in light of the rapidly growing market need for multimedia management applications resulting from the huge influx of multimedia data. Through feedback, student views indicate that despite being long and time consuming, the experiment provides an interesting, original and stimulating experience.

The use of low-level visual features such as colour and shape in content-based image retrieval system leads to several ambiguities. Specifically, due to the many-to-many mapping between the low-level feature space and high-level user concepts, a conceptual user query may not be modelled as a single point in the feature space. Furthermore, conceptually similar images may not fall close to each other in the low-level feature space. This work addresses these issues by proposing an interactive technique, query feedback. This method employs user input to represent a high-level conceptual query as various points in the low-level space. Thus, similarity to a given conceptual query is obtained as the fusion of several low-level representations in the feature space.

The rapid growth of image archives increases the need for efficient and fast tools that can retrieve and search through a large amount of visual data. In this paper, we propose an efficient method of extracting the image color content, which serves as an image digital signature, allowing us to efficiently index and retrieve the content of large, heterogeneous multimedia Internet based databases. We apply the proposed method for the retrieval of images from the WEBMUSEUM Internet database, containing a collection of fine art images and show that the new method of image color representation is robust to image distorsions caused by resizing and compression and can be incorporated into existing retrieval systems which exploit the information on color content in digital images.

We propose a novel face detection algorithm which detects faces in color images using a combination of feature and decision fusion mechanisms. In addition to commonly used skin color information, two additional features, namely average face template matching score and horizontal edge template matching score are utilized. A mean shift algorithm operating on the combined feature space is used to determine face candidate areas. The face candidate and its flipped pattern are then inputted to a multiple layer perceptron based classifier. Two outputs along with the correlation value between the candidate and its flipped pattern are then combined to give the final decision. Experimentation on two different test databases indicates that the proposed method performs well under a variety of scale, expression and environmental conditions, outperforming commonly used approaches.

An open peer-to-peer architecture for performing distributed image indexing and retrieval is proposed. The system employs a sociological model of human acquaintance networks (small world theory) and concepts derived from the nature of the World Wide Web. Retrieval is performed using agents, and a node hopping algorithm is employed that exploits node referrals established from descriptor data stored locally by each node. A general framework for this small world image miner (SWIM) is presented along with a realization using MPEG-7 color structure descriptor data for 2400 images. Results related to search agent path length and network node degree are presented.

It is well-known that the applicability of linear discriminant analysis (LDA) to high-dimensional pattern classification tasks such as face recognition (FR) often suffers from the so-called "small sample size" (SSS) problem arising from the small number of available training samples compared to the dimensionality of the sample space. In this paper, we propose a new LDA method that effectively addresses the SSS problem using a regularization technique. In addition, a scheme of expanding the representational capacity of the face database is introduced to overcome the limitation that the LDA based algorithms require at least two samples per class available for learning. Extensive experimentation performed on the FERET database indicates that the proposed methodology outperforms traditional methods such as eigenfaces and direct LDA in a number of SSS setting scenarios.

In this paper, a novel face recognition method is proposed for surveillance photo identification applications. In such a case, only a limited number of images per subject is available for training purposes. Furthermore, surveillance photos are usually different from the stored templates mostly due to aging, illumination and pose variations. It is common practice to apply unsupervised techniques such as principle component analysis (PCA) when the sample size for each subject is small. However, since PCA is performed without sample label considerations, the captured variation between images contains not only interpersonal variation but also intrapersonal variation which has an adverse effect on recognition performance. To overcome the problem, feature selection is performed in the PCA space to obtain a representation in which intrapersonal variation is minimized and interpersonal variation is maximized. Extensive experimentation following the FERET evaluation protocol indicates that the proposed scheme improves significantly the recognition performance.

From a perceptual standpoint, the subjectivity inherent in understanding and interpreting visual content in multimedia indexing and retrieval motivates the need for online interactive learning. Since efficiency and speed are important factors in interactive visual content retrieval, most of the current approaches impose restrictive assumptions on similarity calculation and learning algorithms. Specifically, content-based image retrieval techniques generally assume that perceptually similar images are situated close to each other within a connected region of a given space of visual features. This paper proposes a novel method for interactive image retrieval using query feedback. Query feedback learns the user query as well as the correspondence between high-level user concepts and their low-level machine representation by performing retrievals according to multiple queries supplied by the user during the course of a retrieval session. The results presented in this paper demonstrate that this algorithm provides accurate retrieval results with acceptable interaction speed compared to existing methods.

The large volumes of artistic visual data available to museums, art galleries, and online collections motivate the need for effective means to retrieve relevant information from such repositories. This paper proposes a decision making framework for content-based retrieval of art images based on a combination of low-level features. Traditionally, the similarity among two images has been calculated as a weighted distance between two feature vectors. This approach, however, may not be mathematically and computationally appropriate and does not provide enough flexibility in modeling user queries. This paper proposes a framework that generalizes a wide set of previous approaches to similarity calculation including the weighted distance approach. In this framework, image similarities are obtained through a decision making process based on low-level feature distances using fuzzy theory. The analysis and results of this paper indicate that the aggregation technique presented here provides an effective, general, and flexible tool for similarity calculation based on the combination of individual descriptors and features.

Linear Discriminant Analysis (LDA) is derived from the optimal Bayes classifier when classes are assumed to be Gaussian with identical covariance matrices. However, it is well known that the distribution of face images under a perceivable variation in viewpoint, illumination or facial ex-pression, is highly nonlinear and complex. The Quadratic Discriminant Analysis (QDA) which relaxes the identical covariance assumption and allows for nonlinear discrimi-nant boundaries to be formed, seems to be a better choice. However, the applicability of QDA to problems, such as face recognition, where the number of training samples is much smaller than the dimensionality of the sample space is problematic due to the increased number of parameters to be learned. In this paper, we propose a new regularized discriminant analysis method that effectively solves the so-called “small sample size ” problem in very high-dimensional face image space. Extensive experimentation performed on the FERET database indicates that the proposed method-ology outperforms traditional methods such as Eigenfaces, QDA and Direct LDA in a number of application scenarios. 1.

Adaptive nonlinear filtering methods are preferred in situations when it is necessary to adapt a filter behavior for varying signal and noise statistics. In the case of impulsive noise corruption, the problem is stated often as searching for the switching function that allows the filtering effect only to noisy samples. Thus, the undesired smoothing of non-corrupted image areas, which results in blurring, especially of small image structures and details, is reduced. We provide a new adaptive framework between a basic vector directional filter and an identity operation based on the angular multichannel definition of J.S. Lee's sigma filter (see Comp Vision, Graphics, Image Proc., vol.24, no.2, p.255-69, 1983).

The paper presents a new filtering scheme for the removal of impulsive noise in color images. It is based on estimating the probability density function for color pixels in a filter window by means of the kernel density estimation method. A quantitative comparison of the proposed filter with the vector median filter shows its excellent ability to reduce noise while simultaneously preserving fine image details.

In this paper, we propose a new algorithm to boost performance of traditional linear discriminant analysis (LDA)-based face recognition (FR) methods in complex FR tasks, where highly nonlinear face pattern distributions are often encountered. The algorithm embodies the principle of "divide and conquer", by which a complex problem, is decomposed into a set of simpler ones, each of which can be conquered by a relatively easy solution. The Ad-aBoost technique is utilized within this framework to: 1) generalize a set of simple FR sub-problems and their corresponding LDA solutions; 2) combine results from the multiple, relatively weak, LDA solutions to form a very strong solution. Experimentation performed on the FERET database indicates that the proposed methodology is able to greatly enhance performance of the traditional LDA-based method with an averaged improvement of correct recognition rate (CRR) up to 9% reported.

In this paper a novel approach to the problem of impulsive noise reduction in color images based on the nonparametric density estimation is presented. The basic idea behind the new image filtering technique is the maximization of the similarities between pixels in a predefined filtering window. The new method is faster than the standard vector median filter (VMF) and preserves better edges and fine image details. Simulation results show that the proposed method outperforms standard algorithms of the reduction of impulsive noise in color images.

This paper presents a new filtering scheme for the removal of impulsive noise in multichannel images. It is based on the estimation of the probability density function for pixels in a filtering window, by means of the kernel density estimation method. The filtering algorithm itself is based on the comparison of pixels with their neighborhood in a sliding window. The results obtained with the new filter show its excellent ability to reduce impulsive noise while simultaneously preserving fine image details.

In this paper we provide a new filtering scheme for the detection and the removal of impulsive noise in digital color images. The proposed adaptive nonlinear vector filters take the advantages of the robust order-statistic theory, generalized directional distance filter and standard sigma filter concept. The principles of the design are explained in detail. Simulation studies indicate that the proposed method is computationally attractive and is able to achieve excellent balance between the image-detail preservation and the noise attenuation.

In this paper a novel approach to the problem of impulsive noise removal in color images based on the nonparametric density estimation is presented. The basic idea behind the new image filtering technique is the maximization of the similarities between pixels in a predefined filtering window. The new method is faster than the standard vector median filter and preserves better edges and fine image details. Simulation results show that the proposed method outperforms other standard algorithms of the reduction of impulsive noise in color images.

Techniques that can introduce low-dimensional feature representation with enhanced discriminatory power is of paramount importance in face recognition (FR) systems. It is well known that the distribution of face images, under a perceivable variation in viewpoint, illumination or facial expression, is highly nonlinear and complex. It is, therefore, not surprising that linear techniques, such as those based on principle component analysis (PCA) or linear discriminant analysis (LDA), cannot provide reliable and robust solutions to those FR problems with complex face variations. In this paper, we propose a kernel machine-based discriminant analysis method, which deals with the nonlinearity of the face patterns' distribution. The proposed method also effectively solves the so-called "small sample size" (SSS) problem, which exists in most FR tasks. The new algorithm has been tested, in terms of classification error rate performance, on the multiview UMIST face database. Results indicate that the proposed methodology is able to achieve excellent performance with only a very small set of features being used, and its error rate is approximately 34% and 48% of those of two other commonly used kernel FR approaches, the kernel-PCA (KPCA) and the generalized discriminant analysis (GDA), respectively.

Low-dimensional feature representation with enhanced discriminatory power is of paramount importance to face recognition (FR) systems. Most of traditional linear discriminant analysis (LDA)-based methods suffer from the disadvantage that their optimality criteria are not directly related to the classification ability of the obtained feature representation. Moreover, their classification accuracy is affected by the "small sample size" (SSS) problem which is often encountered in FR tasks. In this paper, we propose a new algorithm that deals with both of the shortcomings in an efficient and cost effective manner. The proposed method is compared, in terms of classification accuracy, to other commonly used FR methods on two face databases. Results indicate that the performance of the proposed method is overall superior to those of traditional FR approaches, such as the eigenfaces, fisherfaces, and D-LDA methods.

Abstract Estimating the position of mobile terminals is an important problem for cellular networks. A low cost method of locating the mobile terminal is to use measurements of the radio path loss. The distribution of radio path loss is, unfortunately, a non‐linear function of the mobile terminal location. The non‐linearity results from large obstacles to radio‐wave propagation such as buildings or hills. This paper demonstrates how the conditional density of the location given measured path loss can be approximated as a sum of kernel density functions based on radio propagation data collected from propagation surveys or estimated from computer models. Using these approximate density functions an accurate location estimate of a mobile terminal can be estimated from measured path loss values contaminated by measurement noise. Copyright © 2002 John Wiley & Sons, Ltd.

The obstacle of generating hybrid queries within the context of content-based image retrieval is still very real. In attempts to overcome this, fuzzy aggregation can be used to combine single, simple index queries into larger, more complex ones. The paper outlines the use of a fuzzy aggregation technique for hybrid querying which has the ability to adjust its behavior according to operator-controlled parameters. The resulting aggregator can be viewed as a feature-adaptive overall similarity measure. We limit the scope of the aggregator to queries involving color content, color coverage, and horizontal/vertical trends, and apply it to a media database comprised of Corel images of fixed size. Preliminary results show promise and illustrate that hybrid queries using the fuzzy aggregator are effective in their ability to retrieve relevant images while suppressing erroneous retrievals when compared to simple, single-feature queries. In addition, the results obtained are at a minimum comparable to multiple-feature queries generated using a weighted mean approach, but exhibiting scalability and greater flexibility in parameter adjustment.

Techniques that can introduce low-dimensional feature representation with enhanced discriminatory power is of paramount importance in face recognition applications. It is well known that the distribution of face images, under a perceivable variation in viewpoint, illumination or facial expression, is highly nonlinear and complex. It is therefore, not surprising that linear techniques, such as those based on principle component analysis (PCA) or linear discriminant analysis (LDA) cannot provide reliable and robust solutions to those complex face recognition problems. We propose a kernel machine based discriminant analysis method, which deals with the nonlinearity of the face patterns' distribution. The proposed method also effectively solves the "small sample size" (SSS) problem which exists in most face recognition tasks. The new algorithm has been tested, in terms of error rate performance, on the multi-view UMIST Face Database. Results indicate that the proposed methodology outperform other commonly used approaches, such as the kernel-PCA (KPCA) and the generalized discriminant analysis (GDA).

We propose a new algorithm of noise reduction in color images. The new technique of multichannel image enhancement is capable of reducing impulse and Gaussian noise and it outperforms the basic methods based on vector median used for noise reduction in color images. A new smoothing operator, based on a random walk model and on a fuzzy similarity measure between pixels connected by a digital geodesic path is introduced. The efficiency of the proposed method was tested on the standard color images using the widely used objective image quality.

We propose a method for shape description of objects in color images. Our method employs angular maps to identify significant changes of color within the image, which are then used to drive snake models. Experimental results show that our angular map-driven snake method not only yields an accurate description of an object shape, but also it is computationally efficient.

We discuss parametric snake-based methods for the localization of object boundaries in color images. We illustrate the discussion using examples from our on-going work.

We propose a method for shape description of objects in color images. Our method employs angular maps in order to identify significant changes of color within the image, which are then used to drive snake models. To obtain an angular map, the angle values of the vectors corresponding to color image pixels are first computed with respect to a reference vector, and organized in a two-dimensional matrix. To identify significant color changes within the original image, the edges of the angular map are next extracted. The resulting edge map is then presented to a snake model. Distance and gradient vector flow snake models have been employed in this work. Experimental results show, not only that the resulting object shape descriptions are accurate and quite similar, but also that our method is computationally efficient and flexible.

A color image segmentation technique is presented for use in coding and/or compression of video-conferencing sequences. The proposed technique utilizes the perceptual HSI (hue, saturation, intensity) color space and incorporates an unsupervised seed determination algorithm that finds the seed pixels that are in the spatial center of the regions in the image. The algorithm employs a hierarchy system where seeds found at the highest level produce the best results. The region growing algorithm uses these seeds to grow regions by appending to each seed pixel those neighboring pixels that satisfy a certain homogeneity criterion. The technique is found to be robust and relatively computationally inexpensive. The effectiveness of the algorithm is found to be much improved over techniques used in the past.

It has been asserted that temporal subband coding (TSB) is inferior to predictive coding for regionally motion compensated (e.g. block-based MC) temporally scalable compressed video. There are two major disadvantages of TSB coding: temporal filtering distortions, and 'open-loop' predictive coding of covered and uncovered regions. The 'open-loop' structure of TSB coding, however, affords two major advantages not enjoyed by MCP coding: simple optimal bit-allocation, non-existence of quantization error feedback. A new adaptive temporal subband (TSB) motion compensated predictive (MCP) coder is proposed. Hierarchical variable-sized block-matched regions with low predictive error are TSB coded, while poorly predicted regions are 'open-loop' MCP coded. Simulation results demonstrate that the adaptive coder substantially improves the temporal scalability of TSB coding, retains an advantageous 'open-loop' structure and provides comparable or superior PSNR to both MCP and TSB coding at MPEG-1 quality bitrates.

In this paper, we present an automatic, multi-frame, region-feature based motion segmentation technique. Region features are extracted from the first two frames and tracked over several frames. A few motion models are obtained by trajectory clustering on the set of trajectories of the region features' centroids. Final segmentation is obtained by region classification based on these motion models. The proposed technique combines the advantages of feature based methods and gradient based methods. Unlike the traditional two-frame formulation which is widely used in the literature, our method is a multi-frame formulation, which provides a more coherent interpretation of the scene over time and generates more reliable segmentation result.

Processing multichannel signals using digital signal processing techniques has received increased attention lately due to its importance in applications such as multimedia technologies and telecommunications. The objective of this paper is twofold: 1) to introduce adaptive filtering techniques to the reader who is just beginning in this area and 2) to provide a review for the reader who may be well versed in signal processing. The perspective of the topic offered here is one that comes primarily from work done in the field of multichannel (color) image processing. Hence, many of the techniques and works cited here relate to image processing with the emphasis placed primarily on filtering algorithms based on fuzzy concepts, multidimensional scaling, and order statistics-based designs. It should be noted, however, that multichannel signal processing is a very broad field and thus contains many other approaches that have been developed from different perspectives, such as transform domain filtering, classical least-square approaches, neural networks, and stochastic methods, just to name a few. We present a general formulation based on fuzzy concepts, which allows the use of adaptive weights in the filtering structure, and we discuss different filter designs. The strong potential of fuzzy adaptive filters for multichannel signal applications, such as color image processing, is illustrated with several examples.

We present a technique for coarsely extracting the regions of natural color images which contain directional detail, e.g., edges, texture, etc., which we then use for image database indexing. As a measure of color activity, we use a perceptually modified distance measure based on the sum-of-angles criterion. We then apply histogram thresholding techniques to separate the image into smooth color regions and busy regions where edge, texture and colour activity exists. Database indices are then created from the busy regions using the directional detail histogram technique and retrieval is performed using these.

The selection of the filter bank in wavelet compression is crucial, affecting the image quality and system design. The biorthogonal coiflet (cooklet) family of wavelet filters has been constructed, and explicit frequency domain formulae have been developed in the Bernstein polynomial basis. We use the Bernstein basis for the frequency domain design and construction of biorthogonal nearly coiflet wavelet bases. In particular, we construct a previously unpublished nearly coiflet 17/11 biorthogonal wavelet filter pair. Key filter quality evaluation metrics due to Villasenor (see IEEE Trans. on Image Proc., vol.4, no.8, p.1053-1060, 1995) demonstrate this filter pair to be well suited for image compression. Comparison is made to the 17/11 biorthogonal coiflet (cooklet), Villasenor 10/18, Odegard 9/7, and classical CDF 9/7 wavelet bases. Simulation results with the SPIHT algorithm due to Said and Pearlman (see IEEE Trans. on Circ. and Systems, vol.6, no.3, p.243-250, 1996), and with our SR/sub SFQ/, confirm that the new 17/11 wavelet basis outperforms the others for still image compression.

We address the issue of image database retrieval based on color using various vector distance metrics. Our system is based on color segmentation where only a few representative color vectors are extracted from each image and used as image indices. These vectors are then used with vector distance measures to determine similarity between a query color and a database image. We test numerous popular vector distance measures in our system and find that directional measures provide the most accurate and perceptually relevant retrievals.

A 2-D content-based triangular mesh structure is proposed for object-based video coding schemes. The mesh design process utilizes a fast color segmentation technique which allows the boundaries of the wireframe patches to effectively match the scene content. The proposed method essentially consists of three components: (i) a face localization stage (for videophone-type applications), (ii) a general color segmentation technique, and() a nodal point placement and triangulation scheme. The proposed approach not only improves the coding efficiency, but it can also support content-based functionalities of future object-based coders.

Wavelet-transform lossy coding achieves state-of-the-art performance for still images. For motion-compensated interframes, however, block-transform coding is preferred due to its ability to intracode poorly predicted blocks. Space-frequency half-pixel block matching is a new approach that enables hybrid block-motion-compensated wavelet-coded interframes to efficiently code blocks with low temporal correlation of high-frequency details. Architectural compatibility is maintained and little additional complexity is required to achieve superior performance. Simulation results with both fixed-size block matching and hierarchical-variable-sized block matching indicate that a block size-constrained adaptive choice of interpolation filter improves the PSNR by 0.75 dB for hybrid wavelet interframe coding (versus +0.25 dB for H.263) at multimedia quality bit rates. For H.263, bit rate reduction can approach 30% at higher rates. The method is perceptually successful where a block-translational motion model fails, efficiently reducing falsely introduced detail.

The problem of narrowband interference suppression in impulsive noise environments is considered. To suppress the interference we introduce a new nonlinear estimator. The proposed estimator can be applied to any discrete time, linear system which is observed in additive non-Gaussian measurement noise. The new filter is recursive, computationally efficient and with significantly improved performance over other linear and nonlinear schemes which are currently used for interference suppression.

Color is the characteristic which is most used for image indexing and retrieval. Due to its simplicity, the color histogram remains the most commonly used method for color indexing and retrieval. However, the lack of good perceptual histogram similarity measures, the global color content of histograms and the erroneous retrieval results due to gamma nonlinearity, calls for improved methods. We implement a vector angular-based distance measure for image retrieval based on color. We build distance vectors in a multidimensional query space in which the retrieval ranking of each image is determined. Our system exhibits high flexibility by allowing all types of queries, including query by color, query by multiple colors and query by example. In addition, colors can be excluded in a query, without requiring an additional level of analysis.

We consider the performance of frequency-hopping spread spectrum (FH SS) radio networks in a Poisson field of interfering terminals using the same modulation and power. The problem is relevant to wireless random-access communication systems where little information about transmitters requires stochastic modeling of their positions. Assuming that the signal strength is attenuated over distance r on average as 1/r/sup m/, we show that the interference in the network could be modeled as a mixture of Gaussian and /spl alpha/-stable noise. Based on this modeling, we derive expressions for the probability of error (P/sub e/) for systems with M-ary frequency shift keying (FSK) which use conventional envelope detectors. Because conventional envelope detectors are optimum only in Gaussian noise and are suboptimum in the noise considered, we also investigate noncoherent detectors which offer improved performance. We examine receivers with limiting nonlinearities and detectors which are optimal in Cauchy noise. Numerical calculations and Monte Carlo simulations are provided to confirm the accuracy of the analysis presented. The results obtained are useful in the performance evaluation of multiple-access radio networks in environments varying from urban settings to office buildings with deterministic and stochastic propagation laws such as lognormal shadowing and Rayleigh fading.

Deconvolution is one of the most important aspects of seismic signal processing. The objective of the deconvolution procedure is to remove the obscuring effect of the wavelet's replica making up the seismic trace and therefore obtain an estimate of the reflection coefficient sequence. This paper introduces a new deconvolution algorithm. Optimal distributed estimators and smoothers are utilized in the proposed solution. The new distributed methodology, perfectly suitable for a multisensor environment, such as the seismic signal processing, is compared to the centralized approach, with respect to computational complexity and architectural efficiency. It is shown that the distributed approach greatly outperforms the currently used centralized methodology offering flexibility in the design of the data fusion network.

A probabilistic approach to edge reinforcement is proposed that is based on Bayesian networks of two-dimensional (2-D) fields of variables. The proposed net is composed of three nodes, each devoted to estimating a field of variables. The first node contains available observations. The second node is associated with a coupled random field representing the estimates of the actual values of observed data and of their discontinuities. At the third node, a field of variables is used to represent parameters describing the membership of a discontinuity into a group. The edge reinforcement problem is stated in terms of minimization of local functionals, each associated with a different node, and made up of terms that can be computed locally. It is shown that a distributed minimization is equivalent to the minimization of a global reinforcement criterion. Results concerning the reinforcement of straight lines in synthetic and real images are reported, and applications to synthetic aperture radar (SAR) images are described.

A new approach to the problem of time series classification is discussed. A new adaptive classification scheme is introduced and compared with existing approaches, such as the Bayesian approach and the incremental credit assignment approach. Simulation results are included to demonstrate the effectiveness of the new methodology.

A simple image compression scheme based on the wavelet transform and properties of the human visual system (HVS) is proposed. The scheme achieves better performance results than those achieved with the well established JPEG compression technique for very high image quality levels. In achieving those results, the wavelet scheme tries to identify the wavelet filter that would result in the smallest number of nonzero coefficients (thus giving a high compression ratio), uses an HVS based processing module to adaptively quantize wavelet coefficients, and efficiently codes the processed wavelet transform. The results show that the scheme outperforms JPEG in terms of the achievable compression ratios, and image quality.

The choice of quantization method in wavelet image compression is a crucial issue that affects performance, quality, and system design. Space-frequency scalar quantization of zero-trees achieves excellent coding efficiency. Stack-run coding is an efficient alternative to zero-trees which maintains independence between subbands. We present a new approach to wavelet quantization which enhances the stack-run coding method. Low addressing complexity, independence between subbands, and fast, parallel decoding are preserved while superior performance is obtained. The most important features are the optimization of dead-zone scalar quantizers, raster scan pattern selection, and the local (spatial) optimization of quantization coefficients. The local optimization is not spatially restricted (as with zero-trees) and new non-recursive optimal algorithms are now possible. Simulation results indicate that the new technique is strongly PSNR competitive with the best of current lossy wavelet image coders. The new framework also allows insight into the nature of the performance gains achieved by space-frequency quantization.

This paper discusses an approach for detecting edges in color images. A color image is represented by a vector field, and the color image edges are detected as differences in the local vector statistics. These statistical differences can include local variations in color or spatial image properties. The proposed approach can easily accommodate concepts, such as multiscale edge detection, as well as the latest developments in vector order statistics for color image processing. A distinction between the proposed approach and previous approaches for color edge detection using vector order statistics is that, besides the edge magnitude, the local edge direction is also provided. Note that edge direction information is a relevant feature to a variety of image analysis tasks (e.g., texture analysis).

New adaptive filters for color image processing are introduced and analyzed. The proposed adaptive methodology constitutes a unifying and powerful framework for multichannel signal processing. Using the proposed methodology, color image filtering problems are treated from a global viewpoint that readily yields and unifies previous, seemingly unrelated, results. The new filters utilize Bayesian techniques and nonparametric methodologies to adapt to local data in the color image. The principles behind the new filters are explained in detail. Simulation studies indicate that the new filters are computationally attractive and have excellent performance.

This paper addresses the implementation of image processing and computer vision techniques to automate tracer flow extraction in images obtained by the photochromic dye technique. This task is important in modeled arterial blood flow studies. Currently, it is performed via manual application of B-spline curve fitting. However, this is a tedious and error-prone procedure and its results are nonreproducible. In the proposed approach, active contours, snakes, are employed in a new curve-fitting method for tracer flow extraction in photochromic images. An algorithm implementing snakes is introduced to automate extraction. Utilizing correlation matching, the algorithm quickly locates and localizes all flow traces in the images. The feasibility of the method for tracer flow extraction is demonstrated. Moreover, results regarding the automation algorithm are presented showing its accuracy and effectiveness. The proposed approach for tracer flow extraction has potential for real-system application.

We present a new, suboptimum, method for the detection of signals in noise for which standard methods of inference are difficult to implement. The method is based on the empirical characteristic function (ECF) and exploits the correlation structure of the ECF evaluated at the finite number of points. We assume discrete time detection so that N i.i.d. data samples are available per symbol. We examine the performance of the proposed method for the mixture of /spl alpha/-stable and Gaussian noise. The detectors based on the ecf show substantial improvements in performance compared to linear detectors and exhibit better performance than the locally optimum (LO) detectors.

In this paper, we present a fast algorithm for perceptually based fractal image compression. The algorithm is based on a refinement of the fractal code from an initial coarse level of a pyramid. Assuming the block matching error is modeled as a first order Laplacian autoregressive process, we derive the threshold sequence for the objective function in each pyramidal level. Computational efficiency depends on the depth of the pyramid and the search step size, and could be improved by up to two orders of magnitude over the computational effort required for a full search of the original image. The algorithm is quasi-optimal, in terms of minimizing the weighted least absolute error. Its main advantage is the greatly decreased computational complexity, when compared to full search algorithms.

This paper addresses the problem of noise attenuation for multichannel data, such as color images. The proposed filter utilizes adaptive data dependent non-parametric techniques. Simulation results indicate that the new filter suppresses impulsive as well as Gaussian noise and preserves edges and details.

A non-linear digital filter is presented in this paper: this filter aims at extending the concept of scalar rank-ordering in the case of multichannel images. The filter is based on two steps: (1) a transformation from a p-dimensional steps to a one-dimensional space by means of a space filling curve; (2) a scalar median filtering step. Results which demonstrate the advantages and the good restoration computational performances of the filter are shown.

This paper addresses the problem of noise attenuation for multichannel data. The proposed filter utilizes adaptively determined data-dependent coefficients based on a novel distance measure which combines vector directional with vector magnitude filtering. The special case of color image processing is studied as an important example of multichannel signal processing.

The processing of color image data using directional information is studied. The class of vector directional filters (VDF), which was introduced by the authors in a previous work, is further considered. The analogy of VDF to the spherical median is shown, and their relation to the spatial median is examined. Moreover, their statistical and deterministic properties are studied, which demonstrate their appropriateness in image processing. VDF result in optimal estimates of the image vectors in the directional sense; this is very important in the case of color images, where the vectors' direction signifies the chromaticity of a given color. Issues regarding the practical implementation of VDF are also considered. In addition, efficient filtering schemes based on VDF are proposed, which include adaptive and/or double-window structures. Experimental and comparative results in image filtering show very good performance measures when the error is measured in the L*a*b* space. L*a*b* is known as a space where equal color differences result in equal distances, and therefore, it is very close to the human perception of colors. Moreover, an indication of the chromaticity error is obtained by measuring the error on the Maxwell triangle; the results demonstrate that VDF are very accurate chromaticity estimators.

Color edge detection is approached in this paper using vector order statistics. Based on the R-ordering method, a class of color edge detectors is defined. These detectors function as vector operators as opposed to component-wise operators. Specific edge detectors can be obtained as special cases of this class. Various such detectors are defined and analyzed. Experimental results show the noise robustness of the vector order statistics operators. A quantitative evaluation and comparison to other color edge detectors favors our approach. Edge detection results obtained from real color images demonstrate the effectiveness of the proposed approach in real applications.

This paper considers the performance of FH SS radio networks in a Poisson field of interfering terminals using the same modulation and power. Assuming 1/r/sup m/ attenuation of signal strength over distance, the interference in the network is modeled as a mixture of Gaussian and circularly symmetric /spl alpha/-stable noise. The problem is relevant to mobile communication systems where the mobility of users requires random modeling of transmitter positions in the network. We derive formulas for the probability distributions of the interference in FH SS radio networks. We generalize the results of Sousa (see IEEE Trans. I.T., vol.38, no.6, p.1743-1754, 1992) where attention was limited to Cauchy random variables, a special subclass of stable distributions, and where the effect of a background (Gaussian) noise was neglected. Based on the formulas derived, we calculate the probability of symbol error for radio links in environments varying from urban settings to office buildings. The results obtained allow the prediction of the performance of wireless systems under a wide range of conditions.

Spatial interpolation is an important technique that is often used to recover an image from its downsampled version, or to simply perform image expansion. Many conventional linear techniques exist, however, these often perform rather poorly in a subjective manner. In this paper, image interpolation is performed using a binary-based Gibbs random field (GRF) model. Images are interpolated from their downsampled versions along with a number of texture parameters that are estimated within smaller image blocks. These iterative GRF methods are subsequently approximated by a non-iterative nonlinear filtering operation, thereby reducing the computational complexity of the interpolation process. Experimental results indicate that the statistical GRF approaches adapt to textured regions as well as the smooth areas within an image, and thus, can achieve better results than the conventional linear schemes.

We present a fast fractal image encoding algorithm which is based on a refinement of the fractal code from an initial coarse level of the pyramid. The pyramid search algorithm is quasi-optimal in terms of minimizing the mean square error. Assuming that the distribution of the matching error is described by an independent, identically distributed (i.i.d.) Laplacian random process, we derive the threshold sequence for the objective function in each pyramidal level. The computational efficiency depends on the depth of the pyramid and the search step size and could be improved up to two orders of magnitude compared with the full search of the original image.

Recognition-based tracking and image coding methods are often based on different techniques. However, these techniques share the necessity of reliable and fast methods for the representation of the information content of the scene. In this paper, a method based on a lossy shape descriptor is presented which can be used for both recognition and coding purposes. The statistical morphological skeleton provides a noise-robust shape descriptor on which a further approximation phase is performed in order to improve the compression ratio. Then, it is possible to estimate the object's pose through a comparison of the shape descriptor with a set of object models stored in a database. An application to surveillance is presented where the obtained description is used to transmit shape information to a remote control center.

We consider the performance of frequency-hopping spread spectrum (FH SS) radio networks in a Poisson field of interfering terminals using the same modulation and power. The problem is relevant to mobile communication systems where the mobility of users requires random modeling of transmitter positions in the network. Assuming logarithmic attenuation of the signal strength over distance between the transmitter and receiver, we show that the interference in the network could be modeled as a mixture of Gaussian and circularly symmetric /spl alpha/-stable noise. Based on this model, we derive union bound approximations for the probability of error for FH systems with M-ary frequency shift keying (FSK). We generalize some of the results of Sousa (see IEEE Trans. I.T., vol.38, no.6, p.1743, 1992), where the focus was limited to Cauchy random variables (RVs), a special subclass of stable distributions, and where the effect of a background (Gaussian) noise was neglected. Numerical calculations and Monte Carlo simulations confirm the accuracy of our analysis. The results obtained allow the prediction of wireless system performance in environments varying from urban settings to office buildings.

A new class of statistical morphological operators for binary image processing is introduced. These operators are based on a digitized version of the mean field approximation. The main advantage of the new operators is provided by the capability of taking into account both noise and shape information. Binary statistical dilation (BSD) and binary statistical erosion (BSE) are considered as a case study. Extensivity properties of BSD and BSE are also discussed.>

One of the most important problems in computer vision and image processing is signal representation and shape description. This usually involves mapping the signal from an original domain to a set of coefficients describing feature contents of the signal. These methods can be classified in two, global and local, categories. Global methods are sensitive to inaccurate segmentation and masking effects. Local methods, on the other hand, are better in partial pattern recognition but are generally application-dependent. Generalized pattern spectrum (GPS) is a local method and it has proved to be a robust shape descriptor which can also be used in partial pattern recognition. We have investigated the classification properties of this mapping on both noisy and partially missed objects. The results show that it is superior to the pecstrum method in the case of partially missed patterns.>

A real-time visual surveillance system is based on three main image processing phases, devoted to extract information about the observed scene: change detection, focus of attention, feature-extraction. In this paper attention is paid to a theory (i.e., binary statistical morphology) which provides a common framework for designing fast and noise-robust methods for the three tasks of interest. The main theoretical novelty is to establish a link between binary statistical morphology and voting methods. An application is presented which deals with intruder detection in a railway-crossing area.>

The hierarchical finite-state vector quantization (HFSVQ) introduced in the paper is an improvement of the finite state vector quantization combined with hierarchical multirate image coding. Based on an understanding of the perception of human eye and the structural features of images, the HFSVQ technique employs different coding rates and different numbers of the predictive states for representative vector selection. The bit rate used to encode images is very low while the reconstructed images can still achieve a satisfactory perceptual quality.>

A method is proposed whereby a color image is treated as a vector field and the edge information carried directly by the vectors is exploited. A class of color edge detectors is defined as the minimum over the magnitudes of linear combinations of the sorted vector samples. From this class, a specific edge detector is obtained and its performance characteristics studied. Results of a quantitative evaluation and comparison to other color edge detectors, using Pratt's (1991) figure of merit and an artificially generated test image, are presented. Edge detection results obtained for real color images demonstrate the efficiency of the detector.

Vector directional filters (VDF) for multichannel image processing are introduced and studied. These filters separate the processing of vector-valued signals into directional processing and magnitude processing. This provides a link between single-channel image processing where only magnitude processing is essentially performed, and multichannel image processing where both the direction and the magnitude of the image vectors play an important role in the resulting (processed) image. VDF find applications in satellite image data processing, color image processing, and multispectral biomedical image processing. Results are presented here for the case of color images, as an important example of multichannel image processing. It is shown that VDF can achieve very good filtering results for various noise source models.

This paper presents the development of several efficient learning algorithms for neural networks (ELEANNE). The ELEANNE 1 and ELEANNE 2 are two recursive least-squares learning algorithms, proposed for training single-layered neural networks with analog output. This paper also proposes a new optimization strategy for training single-layered neural networks, which provides the basis for the development of a variety of efficient learning algorithms. This optimization strategy is the source of the ELEANNE 3, a second-order learning algorithm for training single-layered neural networks with binary output. A simplified version of this algorithm, called ELEANNE 4, is also derived on the basis of some simplifying but reasonable assumptions. The two algorithms developed for single-layered neural networks provide the basis for the derivation of ELEANNE 5 and ELEANNE 6, which are proposed for training multilayered neural networks with binary output. The ELEANNE 7 is an efficient algorithm developed for training multilayered neural networks with either binary or analog output.>

Two interpolative block truncation coding (BTC) schemes with vector quantization (VQ) and median filters as the interpolator are proposed. The first scheme is based on quincunx subsampling and the second one on every-other-row-and-every-other-column subsampling. Compared to the standard BTC, the new schemes yield a significant reduction in bit rate, only a small degradation in quality, and better channel error resisting capabilities. Moreover, the methods are shown to outperform the BTC schemes with pure VQ at the same bit rate and require minimal computations for the interpolation.>

A novel image coding scheme is proposed and studied, in which fine spatial features are separated from an image and encoded using directional decomposition-based coding techniques. Morphological techniques are used for multiresolution feature decomposition and directional filtering in order to preserve the spatial integrity of the features. The experimental results from a preliminary study show that a high compression ratio can be achieved with good reconstruction image quality.>

The method of histogram equalization for grey-level image enhancement is extended to color images in the paper. A method of direct 3-D histogram equalization is proposed which results in a uniform histogram of the RGB values. Due to the correlation between the bands, the principle on which grey-level (1-D) histogram equalization is based is not valid in the case of color images (3-D). This problem is alleviated by employing a histogram specification method where a uniform histogram is specified.>

A family of nonlinear filters based on order statistics is presented. A mathematical tool derived through robust estimation theory, order statistics has allowed engineers to develop nonlinear filters with excellent robustness properties. These filters are well suited to digital image processing because they preserve the edges and the fine details of an image much better than conventional linear filters. The probabilistic and deterministic properties of the best known and most widely used filter in this family, the median filter, are discussed. In addition, the authors consider filters that, while not based on order statistics, are related to them through robust estimation theory. A table that ranks nonlinear filters under a variety of performance criteria is included. Most of the topics treated are very active research areas, and the applications are varied, including HDTV, multichannel signal processing of geophysical and ECG/EEG data, and a variety of telecommunications applications.>

An architecture for the efficient and high-speed realization of morphological filters is presented. Since morphological filtering can be described in terms of erosion and dilation, two basic building units performing these functions are required for the realization of any morphological filter. Dual architectures for erosion and dilation are proposed and their operations are described. Their structure, similar to the systolic array architecture as used in the implementation of linear digital filters, is highly modular and suitable for efficient very-large-scale integration (VLSI) implementation. A decomposition scheme is proposed to facilitate the implementation of two-dimensional morphological filters based on one-dimensional structuring elements constructed using the dual architectures. The proposed architectures, which also allow the processing of gray-scale images, are appropriate for applications where speed, size, and cost are of critical significance.>

Novel types of opening operator (NOP) and closing operator (NCP) are proposed. An adaptive morphological filter is then constructed on the basis of the NOP and NCP. The filter can remove any details consisting of fewer pixels than a given number N, while preserving the other details. Efficient algorithms are also developed for the implementation of the NOP and NCP.

A methodology based on mathematical morphology is proposed for efficient recognition of two-dimensional (2D) objects or shapes. It is based on the introduction a shape descriptor called the morphological autocorrelation transform (MAT). The MAT of an image is composed of a family of geometrical correlation functions (GCFs) which define its morphological covariance in a specific direction. The MAT is translation-, scale-, and rotation-invariant. It is shown that in most situations, a small subset of the MAT suffices for image representation. The characteristics and performance of a shape recognition system based on the MAT are investigated and analyzed. Computational complexity of the proposed morphological-based recognition system is examined. It is shown that shape properties, such as area, perimeter, and orientation, are readily derived from the MAT representation, and that the proposed system is well suited for shape representation and classification.

Learning from examples uses specific instances (examples and counterexamples) to produce general rules. It is a convenient learning scheme in cases where the process of interviewing human experts and analyzing and formalizing their decision is very difficult or time consuming. The system proposed is capable of obtaining the rules that fit a set of examples and counterexamples based on the minimal entropy (ME) criterion. The system proposed can also set various parameters of the rule (e.g., thresholds) in such a way that entropy is minimized. The system can also handle incremental learning from examples. Applications of the proposed system to seismic image analysis are included.>

The problem of modeling of nonGaussian processes generated by linear systems driven by a white nonGaussian process, and nonlinear systems driven by a white Gaussian process is addressed using the Volterra representation of systems. Cumulant-based approaches are developed for identifying the parameters of the proposed model when only a finite sample of received observations is available. It is shown that by using a partial set of the output cumulant samples, the computational complexity required in determining the kernels of the model is considerably reduced. The analysis is not restricted to special forms of the second-order Volterra system.>

A new approach for shape representation is described which provides a general scheme for object description and unifies some of the existing representation techniques (e.g. CSG, skeletons, shape decomposition). It is based on the use of simple geometric objects which are intuitively used by humans in their perception of shapes and on mathematical morphology. Furthermore, its unifying framework can be used both in computer graphics and computer vision, thus providing a tool to close the existing gap between object modeling and object recognition in certain applications, e.g. in robotic vision.>

The joint optimization of different criteria is addressed as a powerful means of incorporating prior information in linear image restoration algorithms. The resolution-to-noise-tradeoff (RNT) approach introduced enables the incorporation of both spatial and spectral information regarding the peculiarities of the problem. Depending on the nature of the spatial information, this approach can be interpreted as either a regularized or an adaptive scheme. As a regularized scheme, it offers an alternative to conventional approaches, in which the ringing artifacts, especially those due to the noise, are efficiently suppressed by means of the regularizing functional. As an adaptive scheme, the RNT approach offers the flexibility of applying either linear Wiener filtering, or inverse filtering, or no filtering at all, depending on the local signal-activity. The capabilities of the RNT approach as both a regularized and an adaptive scheme are demonstrated through restoration examples.>

Input control policies had been proposed previously (e.g. in Lam and Kleinrock, 1975; and Fayolle et al., 1977) to improve the delay-throughput performance of Aloha access schemes. Through simulation results, Lam and Kleinrock observed that the optimal control policies were in general of control limit type. The authors discuss input control policies for one such Aloha access scheme and prove analytically that the optimal policy is indeed of the control limit type. The optimal policy is found to be extremely simple for the access scheme discussed. This policy is to accept the newly arrived packet if the present system state is either zero or one, and to reject the packet otherwise.>

A new descriptor for representing two-dimensional continuous or discrete signals is introduced. The proposed shape descriptor, which is called the geometrical correlation function (GCF), is based on the principle of mathematical morphology. The properties of this shape descriptor are examined. It is shown that the family of GCFs associated with different orientations of a particular shape is translation, scale, and rotation invariant. Geometrical properties such as the area and perimeter of the shape can be derived from the GCF family. The utilization of the GCFs for shape recognition is considered. The GCF family can be computed using the associated morphological correlator which is composed of m parallel computation units and a feature-function selection unit where a small subset of the GCF family is selected for classification. It is shown that with a suitable criterion for selecting the feature function, promising results for successful classification are obtained.>

An adaptive filter structure is proposed which is based on linear combinations of order statistics. An efficient method to update the filter coefficients is presented. It is based on the minimal mean-square-error criterion and is similar to the Widrow algorithm for linear adaptive filters. Another method for coefficient update that is similar to the RLS algorithm and has faster convergence properties is presented. The proposed filter can adapt well to a variety of noise probability distributions, including impulsive noise.>

A technique for decomposing a binary shape into a union of simple binary shapes is presented. The decomposition is shown to be unique and invariant to translation, rotation, and scaling. The techniques used in the decomposition are based on mathematical morphology. The shape description produced can be used in object recognition and in binary image coding.>

The authors present a generalized morphological skeleton transform (MST) algorithm for rotation-invariant vision applications. Several subclasses of MST are derived from the algorithm by using appropriate sets of structuring elements (SE). The rotational property of an MST is determined by the circularity of its SEs. A pseudo-Euclidean morphological skeleton transform, which uses quasi-circular SEs composed through a dilation interpolation procedure, is proposed. Experimental results show that the pseudo-Euclidean MST gives the best rotation-invariant performance among existing MST subclasses. These quasi-circular SEs can also be used in other morphological transformations requiring rotation-invariant performances.>

The nature of the morphological skeleton representation of a binary shape is related to the composition of structuring elements through the distance function defined by morphological set transforms in digital space. Two digital metrics, uniform-step distance and periodically-uniform-step distance, are introduced to provide useful spatial measures for morphological transforms. A natural shape representation by ribbonlike components is accomplished by extraction of skeletal feature primitives from the morphological skeleton of a shape. The hierarchical structure of the representation makes it stable and insensitive to noise disturbance. The matching is a simple top-down process in which the inverses of the skeletal feature primitives at each level are compared. The recognition is based on the similarity measure provided by the matching process.>

In this paper, quadratic filters are designed for seismic deconvolution. These nonlinear filters can be easily obtained from optimizing the D norm for the Minimum Entropy Deconvolution (MED) algorithm. Due to the symmetric structure of the quadratic filter, the actual number of filter coefficients is reduced. In finding the filter coefficients, a system of linear equations is obtained. Recently, efficient methods to realise these quadratic filters have been reported. Simulation results show that low order quadratic filters are usually sufficient and that the filtered output indicates the seismic events even when the seismic wavelet is nonminimum phase. When the data are contaminated with noise, quadratic filters can still perform reasonably well. In the case of a complex reflectivity series, nonlinear filtering is superior to linear filtering. Examples of processing synthetic seismic data by linear and nonlinear filtering are presented.

There are many significant applications of nonlinear adaptive digital filters such as the cancellation of echoes and intersymbol interference, the equalization of transmission channels, adaptive noise cancellation and design of optimal predictors in communication systems. In this paper, we introduce a class of efficient architectures for adaptive quadratic digital filters based on the LMS algorithm and on rank compressed lower-upper (LU) triangular decomposition method. The architectures exhibit high parallelism as well as modularity and regularity. They are mapped into parallel pipeline and systolic array implementations and are evaluated on hardware cost (in bits), and data throughput delay.

Geophysical interpretation of seismic images is part of geophysical oil prospecting. Its aim is the detection of geologic formations in a seismic cross-section of the earth, which are likely to contain oil reservoirs. It is a labour intensive task and it is heavily based on the experience of the interpreter. Therefore it has not been automated as ithas already been done with other tasks of geophysical seismic signal processing. The aim of this work is to construct an expert system which can automate, at least partly, seismic interpretation. The system developed is able to detect several geologic formations of interest (eg. faults, anticlines, unconformities, salt domes, reefs).

Carrier sense multiple-access (CSMA) protocols are widely used in local area networks (LAN's) to control access to a shared communications channel such as a coaxial cable or radio frequency band. CSMA protocols are designed to exploit the property that the signal propagation time across the LAN is much smaller than the packet transmission time. Consequently, their performance depends on the exact timing of a sequence of asynchronous events originating at different points in the network. Previous models of CSMA protocols have not captured this essential space-time characteristic. Instead, the system has been reduced to a 1-dimensional model (i.e., a sequence of events on a single time line) by assuming either that events are synchronous (i.e., "slotted" operation of the protocol) or that all stations are mutually equidistant (so it can be assumed that each station's time line is identical). The novelty in our work is to describe the system state in terms of an (N + 1)-dimensional "ribbon" of space-time, allowing us to faithfully model the exact timings of events for LAN's where the stations are distributed over a fully connected region inN-space. For example, an Ethernet-like "bus" network can be viewed as a 1-dimensional LAN, a terrestrial packet radio network can be viewed as a 2dimensional LAN, and so on. First we highlight some of the key properties exhibited by such a model in the general case, including the notion of an embedded Markovian sequence of "idle points," and the decomposition of the space-time ribbon into cycles (each of which can be further decomposed into a number of "regions"). Then we present extensive results for the interesting special case of a 1-dimensional LAN, i.e., an Ethernet-like "bus" network. Our results show that previous models of bus LAN's can significantly underestimate their performance.

Multiple-access protocols control access to a broadcast communication channel. Tree conflict resolution algorithms are the heart of some distributed multiple-access protocols with nice properties like stability, high capacity, and low delay under light load. We consider a random access protocol based on a tree conflict resolution algorithm similar to one first proposed by Capetanakis, but in which constant size windows on the arrival time axis are used to admit packets into the algorithm instead of the more common "free" or "blocked" access methods. We obtain exact recursive relationships for the (steady-state) distribution of packet delay, and thus, the exact throughput-delay curve for any finite (in general, nonhomogeneous) configuration under a Bernoulli-per-window arrival model. Then, we consider the address assignment problem. We show that by choosing an appropriate addressing scheme, we can improve the performance of the algorithm with respect to both mean delay and maximum throughput.

We present a brief and simple analysis of asynchronous (unslotted) 1-Persistent CSMA protocols in the worst-case star topology. The purpose is two-fold. First, we give a simple approach based on an embedded Markov chain at the beginning of subbusy periods that greatly simplifies the analysis. And second, using the above approach we give the correct analysis for 1-Persistent CSMA with collision detection, since the only available analysis of the present model [1] is in error.

A technique for the design of three-dimensional (3- D) filters is introduced in this paper. The design is based on coefficient transformations of two-dimensional (2-D) circularly symmetric filters. These filters can be designed by cascading 2-D rotated filters. The stability of the 3-D filters designed is discussed and a stabilization procedure based on cepstrum analysis, is proposed. Stable implementation schemes are introduced. Examples of spherically symmetric filters designed on the basis of the technique introduced are presented.

A class of edge detectors is proposed which is obtained from the difference of the outputs of two nonlinear filters. The nonlinear filters used are based on nonlinear means (which contain homomorphic filters) or order statistic (which contain median filters). The performance of these edge detectors in the presence of noise is evaluated and is compared to the performance of well established edge detectors. Their computational requirements are described and some examples are presented.

A filter function suitable for modular implementation of two-dimensional digital filters was determined. Given this transfer function, a two-chip set was implemented in 5 micron NMOS technology. This chip set performs an all-pole or all-zero filtering operation in 4-bit slices (i.e. the wordlength is indefinitely expandable in an expansion chain) on a 512 × 512 pixel array using two's complement arithmetic. The chip set may be used as a building block for implementing general first-quadrant two-dimensional digital filters.

A sufficient condition for the extendibility of two-dimensional (2-d) quadratically symmetric autocorrelation samples is derived based on matrix decomposition approaches. This condition is very simple to test because it boils down to a positive definiteness checking of a set of Toeplitz form matrices. It is shown by means of examples that this sufficient condition can be very useful, in some cases, by making the tedious extendibility test unnecessary. This paper also presents a method which can approximate with small error a given quadratically symmetric autocorrelation function (a.c.f) over a square with a new a.c.f that is guaranteed to be extendible. This method is also based on a matrix decomposition approach and employs an iterative gradient algorithm. Its implications to the estimation of a 2-d power spectrum is demonstrated.

Waves (acoustic or electromagnetic) are used in many areas, such as ultrasonic medicine, geophysical oil prospecting, etc., for non-destructive imaging of the interior of objects, such as the earth and the human body. Wavefield measurements recorded on the surface of an object are inverted according to the wave equation to give an "image" of the interior of the object. This inversion usually enhances the noise present in the measurements and degrades the quality of the final image. Another source of degradation of the image is the insufficient knowledge of some properties of the medium, such as the wave propagation velocity. The aim of this paper is to use estimation theory to determine good estimates of the velocity and of the object image from known wavefield measurements.

In this paper a new two-dimensional (2-D) digital filter structure is presented. The proposed structure has a short clock cycle time or a high data throughput rate, independent of the order of the filter. Parallelism and pipelining are the two features of the proposed filter structure that contribute to its high speed performance. Hardware of the filter utilizing no multipliers is outlined. It has been shown that the filter is capable of filtering images of size up to 512×512 pixels with a TV scan rate of 30 frames/sec in real-time. Finally, the new filter structure is compared to the one implemented by distributed arithmetic and is shown to be superior.

A new approach to the realization of two-dimensional (2-d) FIR and IIR digital filters is introduced based on the so-called "Lower-Upper triangular (LU) decomposition" of matrix coefficients of their two-dimensional polynomials. It is shown that a general two-dimensional transfer function can be exactly realized in terms of polynomials of order one, each one of which is a function of one of the two space variables only. It is demonstrated that the LU realization scheme exhibits high inherent parallelism as well as great modularity, regularity and flexibility. This paper also shows that the computational requirements of the LU realization are much less than these of the Jordan (J), Singular Value (SV) and canonical realization schemes.

The Urn scheme is known to perform better than optimal ALOHA and TDMA for all ranges of traffic rates. In this paper we discuss the dynamic behavior of the Urn scheme to show that it possesses bistable behavior in a manner similar to ALOHA schemes and that dynamic control procedures can be applied to improve the system performance effectively. In particular, an input control procedure (ICP) is presented that gives a delay-throughput characteristic very close to optimal (perfect scheduling) for a wide range of throughput rates. The improvement is obtained at no extra cost in terms of information acquisition and the complexity introduced is minimal. An analytical method is described to calculate the expected delay, throughput, and the probability of packet rejection. Numerical results are shown for various values of user population and compared with corresponding results for other schemes.

Number theoretic transforms using the modulus2^{2_{q}} - 2^{q} + 1, which require no multiplications, are introduced. These transforms have 50% greater sequence lengths for a given wordlength than Fermat transforms, and there is greater flexibility in the choice of wordlength. Furthermore, arithmetic modulo2^{2_{q}} - 2^{q} + 1is not significantly more complex than arithmetic modulo a Fermat number. Complex number theoretic transforms using these moduli are also discussed and techniques for using them to convolve real data are described.

Two dimensional convolution computed using minicomputers is very time-consuming using Fourier Transform techniques because of the number of complex multiplications required. It has been shown that by employing Fermat Number Transforms a hardware butterfly unit becomes relatively simple to implement and significantly increases the speed of the calculation. Such a hardware unit, suitable for interfacing to a minicomputer, is briefly described. An algorithm is given which removes the need for a matrix transposition. This algorithm is useful on small disc based computers which cannot store the complete input matrix in main memory since a matrix transposition is a time-consuming operation.

Necessary and sufficient conditions for a direct sum of local rings to support a generalized discrete Fourier transform are derived. In particular, these conditions can be applied to any finite ring. The function O(N) defined by Agarwal and Burrus for transforms over ZN is extended to any finite ring R as O(R) and it is shown that R supports a length m discrete Fourier transform if and only if m is a divisor of O(R) This result is applied to the homomorphic images of rings-of algebraic integers.

The problems of optimum performance and signal design for on-off signaling over linear, time-varying stochastic channels are considered. The channel is modeled as a nonstationary linear time-varying Gaussian filter with additive white Gaussian noise at its output. A new expression for the probability of error is presented and is minimized under the constraint of fixed received energy. An approximately optimum signal design method that yields the minimum probability of error is proposed.

An upper bound on the signal-to-noise ratio (SNR) at the output of an estimator-correlator receiver is derived, and the nature of the transmitted signals and random channels that realize this bound is investigated.