This special section in the Journal of Electronic Imaging presents eight select papers from the field of visualization out of 24 papers submitted for consideration. This special section is an outgrowth of the SPIE Conference on Visual Data Exploration and Analysis that we have been fortunate enough to have been involved with over the past several years. This is an opportune time for a special section of Visualization and Data Analysis.

The visualization of vector fields has attracted much attention over the last decade due to the vast variety of applications in science and engineering. Topological methods have been used intensively for global structure extraction and analysis. Recently, there has been a growing interest in local structure analysis due to its connection to automatic feature extraction and speed. We present an algorithm that extracts local topological structure of arbitrary regions in a two-dimensional vector field. It is based on a mathematical analysis of the topological vector field structure in these regions. The algorithm deals with piecewise linear vector fields and arbitrary polygonal regions. We have tested the algorithm for well known analytic vector fields and data sets resulting from computational fluid dynamics.

Software visualization involves creation and display of graphical representations of software characteristics or behavior, often to assist in the discovery of software problems or unanticipated behavior. This paper presents a new approach for threedimensional (3D) visualization (including extraction and display) of design phase object-oriented software metrics. The visualizations are designed to foster developers’ intuitive understanding of the metrics and to aid in the early understanding of structural characteristics of design components. The approach utilizes the extensibility interface of a popular computer aided software engineering tool to access and automatically extract eight key design metrics from Unified Modeling Language class diagrams. Following metric extraction, 3D visualizations of these metrics are generated for each class using perceptually meaningful glyphs. The approach has been implemented by a tool that emits and displays Virtual Reality Modeling Language representations of a design’s glyphs.

An important issue in data visualization is the recognition of complex dependencies between attributes. Past techniques for identifying attribute dependence include correlation coefficients, scatterplots, and equi-width histograms. These techniques are sensitive to outliers, and often are not sufficiently informative to identify the kind of attribute dependence present. We propose a new approach, which we call independence diagrams. We divide each attribute into ranges; for each pair of attributes, the combination of these ranges defines a two-dimensional grid. For each cell of this grid, we store the number of data items in it. We display the grid, scaling each attribute axis so that the displayed width of a range is proportional to the total number of data items within that range. The brightness of a cell is proportional to the density of data items in it. As a result, both attributes are independently normalized by frequency, ensuring insensitivity to outliers and skew, and allowing specific focus on attribute dependencies. Furthermore, independence diagrams provide quantitative measures of the interaction between two attributes, and allow formal reasoning about issues such as statistical significance. We have addressed several technical challenges in making independence diagrams work, ranging from the treatment of categorical attributes to visual artifacts of cell-topixel mapping. Our experimental evaluation, using both AT&T and synthetic data, shows that independence diagrams allow the easy identification of various kinds of attribute dependence that would be difficult to identify using conventional techniques.

Although visualization in the field of dentistry has some of the same requirements as the medicine field, the differences in goal demand specific approaches. This paper reports on the implementation of two fundamentally different approaches to reconstruction of structures from planar cross sections and their application to dentistry data. One of the approaches was an implementation of a distance-based sampling technique, and the other is a new algorithm, based on the Delaunay triangulation. Both were tested using contour data of teeth and the results are compared here in the light of the target applications, which are teaching and training dentistry, as well as simulation of dental procedures and illnesses. Widely mentioned problems encountered in local reconstruction methods such as marching cubes for these cases are clearly illustrated in this paper, and a very satisfactory alternative is given.

Previously, we presented a system called Alignment Viewer that uses information visualization techniques to visualize similarities between a single DNA sequence and a large database of other sequences [Chi et al., IEEE Visualization ’95, pp. 44–51, IEEE CS (1995); Chi et al., IEEE Visualization ’96, pp. 133–140 and 477, IEEE CS (1996)]. In this paper, we extend, summarize, and describe the system using several interesting case studies. We present our comb glyph technique for visualizing alignments between sequences. In this paper, we also extend the original system by incorporating computational steering, and the visualization of differences between data sets. The case studies and the new extended system present our approach of extracting significant relationships in the biological data set.

Most existing visualization systems stress either the original data visualization or the discovered knowledge visualization, such as decision tree, neural network, rules, etc., but lack the abilities to visualize the entire process of knowledge discovery. We propose an interactive model, RuleViz, for visualizing the process of knowledge discovery and data mining. The RuleViz model consists of five components, each of which can be interacted and visualized by using different visualization techniques. According to this model, two interactive systems, AViz and CViz, for visualizing the process of discovering numerical association rules and the process of learning classification rules have been implemented, respectively. To preprocess the data, each system provides users with three approaches for discretizing numerical attributes and the corresponding rule discovery algorithms. The discretization approaches and the algorithms for discovering association rules and learning classification rules are presented, and the approaches to visualizing discretized data and discovered rules are developed. The discovery of numerical association rules in AViz is based on image-based mining algorithm, while, in CViz, the classification rules are learned in terms of a logical rule induction algorithm. We also demonstrate our experimental results with AViz and CViz on the census data sets, UCI data sets, and artificial data sets.

This paper describes two systems for managing threedimensional and anthropometric databases, namely Alexandria and Cleopatra. Each system is made out of three parts: the crawler, the analyzer, and the search engine. The crawler retrieves the content from the network while the analyzer describes automatically the shape, scale, and color of each retrieved object and writes down a compact descriptor. The search engine applies the query by example paradigm to find and retrieve similar or related objects from the database based on different aspects of three-dimensional shape, scale, and color distribution. The descriptors are defined and the implementation of the system is detailed. The application of the system to the CAESAR anthropometric survey is discussed. Experimental results from the CAESAR database and from generic databases are presented.

A colorimetric modeling technique is proposed to give a computational model associated with colorimetry so that the representation of color acquired from camera imaging is accurate and meaningful. First of all, the camera spectral responses are estimated and the colorimetric quality is evaluated to reveal the feasibility of this work. In the modeling process, we present a spectral matching method and an approach of determining a reference-white luminance. As a result, the acquired color and the true (or measured) color can be well coordinated, with the strength of a global illumination or display white, in a perceptually uniform color space, e.g., in CIE 1976 L*a*b* space (abbreviated as CIELAB). Then, lower-degree polynomial regression is employed to eliminate color errors due to the mismatch between spectral response functions. Experimental results indicate that the root-mean-square ?E*ab value (i.e., color error) from the degree-3 polynomial regression is less than a just-noticeable difference (about 2.3) in CIELAB. It appears that the proposed technique can establish an accurate colorimetric model for vision systems.

Since the early 1990s a number of papers on ‘‘robust’’ digital watermarking systems have been presented but none of them uses the same robustness criteria. This is not practical at all for comparison and slows down progress in this area. To address this issue, we present an evaluation procedure of image watermarking systems. First we identify all necessary parameters for proper benchmarking and investigate how to quantitatively describe the image degradation introduced by the watermarking process. For this, we show the weaknesses of usual image quality measures in the context watermarking and propose a novel measure adapted to the human visual system. Then we show how to efficiently evaluate the watermark performance in such a way that fair comparisons between different methods are possible. The usefulness of three graphs: ‘‘attack versus visual-quality,’’ ‘‘bit error versus visual quality,’’ and ‘‘bit error versus attack’’ are investigated. In addition the receiver operating characteristic graphs are reviewed and proposed to describe statistical detection behavior of watermarking methods. Finally we review a number of attacks that any system should survive to be really useful and propose a benchmark and a set of different suitable images.

Digital watermarking is the enabling technology to prove ownership on copyrighted material, detect originators of illegally made copies, monitor the usage of the copyrighted multimedia data, and analyze the spread spectrum of the data over networks and servers. Embedding of unique customer identification as a watermark into data is called fingerprinting to identify illegal copies of documents. Basically, watermarks embedded into multimedia data for enforcing copyrights must uniquely identify the data and must be difficult to remove, even after various media transformation processes. Digital fingerprinting raises the additional problem that we produce different copies for each customer. Attackers can compare several fingerprinted copies to find and destroy the embedded identification string by altering the data in those places where a difference was detected. In our paper we present a technology for combining a collusion-secure fingerprinting scheme based on finite geometries and a watermarking mechanism with special marking points for digital images. The only marking positions the pirates cannot detect are those positions which contain the same letter in all the compared documents, called intersection of different fingerprints. The proposed technology for a maximal number d of pirates, puts enough information in the intersection of up to d fingerprints to uniquely identify all the pirates.

Current ‘‘invisible’’ watermarking techniques aim at producing watermarked data that suffer no or little quality degradation and are perceptually identical to the original versions. The most common utility of a watermarked image is (1) for image viewing and display, and (2) for extracting the embedded watermark in subsequent copy protection applications. The issue is often centered on the robustness of the watermark for detection and extraction. In addition to robustness studies, a fundamental question will center on the utilization value of the watermarked images beyond perceptual quality evaluation. Essentially we have to study how the watermarks inserted affect the subsequent processing and utility of images, and what watermarking schemes we can develop that will cater to these processing tasks. This work focuses on the study of watermarking on images used in automatic personal identification technology based on fingerprints. We investigate the effects of watermarking fingerprint images on the recognition and retrieval accuracy using a proposed invisible fragile watermarking technique for image verification applications on a specific fingerprint recognition system. We shall also describe the watermarking scheme, fingerprint recognition and feature extraction techniques used. We believe that watermarking of images will provide value-added protection, as well as copyright notification capability, to the fingerprint data collection processes and subsequent usage.

One method of attacking an imbedded invisible watermark is to create a derivative image that is geometrically distorted relative to the original. One attack, developed at Cambridge University, is called ‘‘StirMark.’’ 1 Image-distorting methods modify images so subtly that the changes are essentially unnoticeable to a viewer. However, their effect on invisible watermarks can be devastating, rendering them unextractable. In this paper, an automated countermeasure to image-distorting attacks will be described. Employing an unmarked copy of the original image as a reference, the possible distortion in a suspect image is first detected by the method, then measured, and finally reversed, producing a restored image approximately geometrically aligned with the original. Using a robust invisible watermarking method described previously by one of the authors to produce a watermarked image, ‘‘StirMark’’ to distort the watermarked image, and a copy of the original unmarked image for reference, the restoration method is demonstrated to be sufficient by showing successful extraction of the imbedded watermark from a restored image.

Various nonlinear, fixed-neighborhood techniques based on local statistics have been proposed in the literature for filtering noise in color images. We present adaptive-neighborhood filtering (ANF) techniques for noise removal in color images. The main idea is to find for each pixel (called the ‘‘seed’’ when being processed) a variable-shaped, variable-sized neighborhood that contains only pixels that are similar to the seed. Then, statistics computed within the adaptive neighborhood are used to derive the filter output. Results of the ANF techniques are compared with those given by a few multivariate fixed-neighborhood filters: the double-window modified trimmed-mean filter, the generalized vector directional filter— double-window—?-trimmed mean filter, the adaptive hybrid multivariate filter, and the adaptive nonparametric filter with Gaussian kernel. It is shown that the ANF techniques provide better visual results, effectively suppressing noise while not blurring the edges; the results are also better in terms of objective measures (such as normalized mean-squared error and normalized color difference) than the results of the other methods.

This paper presents entropy-constrained learning vector quantization (ECLVQ) algorithms and their application in image compression. The development of these algorithms relies on reformulation, which is a powerful new methodology that essentially establishes a link between learning vector quantization and clustering algorithms developed using alternating optimization. ECLVQ algorithms are developed in this paper by reformulating entropyconstrained fuzzy clustering (ECFC) algorithms, which were developed by minimizing an objective function incorporating the partition entropy and the average distortion between the feature vectors and their prototypes. The proposed algorithms allow the gradual transition from a maximally fuzzy partition to a nearly crisp partition of the feature vectors during the learning process. This paper presents two alternative implementations of the proposed algorithms, which differ in terms of the strategy employed for updating the prototypes during learning. The proposed algorithms are tested and evaluated on the design of codebooks used for image data compression.

This paper introduces two new codebook design algorithms for image data compression applications. The proposed algorithms design codebooks by splitting and merging the prototypes of the training vectors, which play the role of code-vectors in the encoding/decoding process. The split-1 (S1) algorithm begins with a codebook containing a single prototype and designs a codebook of the desired size by splitting one prototype at each iteration. The split-1-merge-2 (S1M2) algorithm begins with a codebook of the desired size and improves its quality in an iterative fashion by splitting one prototype and merging two prototypes at each iteration. The proposed algorithms are used to design codebooks for image data compression based on vector quantization. The experimental results indicate that the proposed algorithms are strong competitors to the popular Linde-Buzo-Gray algorithm in terms of their computational requirements and the quality of the designed codebooks.

An adaptive motion-compensated video coding scheme, that is based on structural video component segmentation and coding complexity analysis, is proposed in this paper. The bits are allocated more efficiently among different frame types and variant video components. A novel scene cut detection algorithm is proposed for partitioning the input video sequences into a set of shots and each shot may be encoded as one or multiple GOPs according to its length. Moreover, the positions of the reference frames (I and P frames) in a video shot are adapted to improve the temporal predictability among frames and provide high coding efficiency, thus high picture quality with the same bit rate. More bits are allocated for these reference frames for providing high quality of the reconstructed pictures. The residue frames in a video shot are encoded as the bidirectional interpolation frames (B frames) and can be also quantized more coarsely because they have high temporal predictability and are not used as references. The bits, that have been allocated for the three different frame types (I , P, B frames), can be further distributed more efficiently among variant video components to avoid the coding artifacts. Experimental results show that this proposed adaptive video coding scheme is more efficient than the traditional fixed GOP coding algorithms and may be an efficient development of the present adaptive coding techniques.

Printing applications using classical halftoning need to resample the original image to a screen lattice. This resampling can cause undesirable moiré artifacts in the screened image. Some printing techniques, e.g., gravure printing, are highly susceptible to moiré , not only because of the low resolution screen lattices they employ but also because the degree of freedom in constructing halftone dots is limited by the physical constraints of the engraving mechanism. Current resampling methods compute new samples by simple interpolation techniques that cannot prevent sampling moiré very well. Therefore precautions against moiré have to be made in the prepress phase, which is not practical and sometimes not feasible. A novel technique is presented to adaptively resample an image on the screen lattice using a local estimate of the risk of aliasing. The purpose is to suppress moire´ while maintaining the sharpness of the image. Experimental results demonstrate the feasibility of the proposed approach.

Watermarking techniques are primarily used for copyright protection. In this paper, we propose a digital watermarking scheme, which is based on neural network, cryptography, and image processing techniques. Our scheme can achieve the following two goals. The first is that illegal users do not know the location of an embedded watermark in the image. The second is that a legal user can retrieve the embedded watermark from an altered (filtering, lossy compression, and scaling) image. The security and robustness of the proposed scheme are improved using neural network technology.

A copy-protection system was developed in order to prevent the unauthorized reproduction of photographic prints at a digital print station. The copy-protection system consists of two parts: (1) copy-protected photographic paper manufactured with a latent image of an array of yellow microdots; and (2) digital print stations equipped with software that detects the presence of the microdots in the digital image of a photographic print made on copy-protected photographic paper. The system has a detection rate of 78% and a negligible rate of false detection.

A crucial aspect of shape similarity estimation is the identification of perceptually significant image elements. In order to understand more about the process of human segmentation of abstract images, a sample of 63 trademark images was shown to several groups of students in two experiments. Students were first presented with printed versions of a number of abstract trademark images, and invited to sketch their preferred segmentation of each image. A second group of students was then shown each image, plus its set of alternative segmentations, and invited to rank each alternative in order of preference. Our results suggest that most participants used a relatively small number of segmentation strategies, reflecting well-known psychological principles. Agreement between human image segmentations and those generated by the ARTISAN trademark retrieval system was quite limited; the most common causes of discrepancy were failure to handle texture and incorrect grouping of components into regions. Ways of improving ARTISAN’s ability to model human segmentation behavior are discussed.

Andrea Kutics received her MS degree in chemical engineering from the University of Veszprem, Hungary, in 1981, in mathematics and computer science from Eotvos Lorand Science University, Hungary, in 1986, and her Doctor Universitas degree in computer science from University of Veszprem in 1991.