Content-based Image Retrieval (CBIR) has been an active area of research for retrieving similar images from large repositories, without the prerequisite of manual labeling. Most current CBIR algorithms can faithfully return a list of images that matches the visual perspective of their inventors, who might decide to use a certain combination of image features like edges, colors and textures of regions as well as their spatial distribution during processing. In practice, however, the retrieved images rarely correspond exactly to the results expected by the users, a problem that has come to be known as the semantic gap. In this paper, we propose a novel and extensible multidimensional approach called matrix of visual perspectives as a solution for addressing this semantic gap. Our approach exploits the dynamic cross-interaction (in other words, mix-and-match) of image features and similarity metrics to produce results that attempt to mimic the mental visual picture of the user. Experimental results on retrieving similar Japanese cultural heritage symbols called kamons by a prototype system confirm that the interaction of visual perspectives in the user can be effectively captured and reflected. The benefits of this approach are broader. They can be equally applicable to the development of CBIR systems for other types of images, whether cultural or noncultural, by adapting to different sets of application specific image features.

Image relevance evaluation in conventional Content-Based Image Retrieval (CBIR) researches typically relied on a given criterion. However, it is important that this criterion can be changed according to the use of the image database. This work proposes a framework for tuning the parameters embedded in the relevance evaluation algorithm of a CBIR system, by optimizing them according to the suitability of the retrieved results, using Particle Swarm Optimization (PSO). In the experiments, the parameters that affect the similarity evaluation in a binary shape matching CBIR system were tuned to improve the retrieval ranking score. After tuning the parameters by PSO, it was found that the ranking of the retrieved images were improved according to the given criterion.

Recently, various ways to reduce the size of multimedia data have been developed. However, there is a problem of the tradeoff of data size, and quality. In compressed audio encoding, high frequency components are lost from the original signal. It is impossible to retrieve the frequency components that have been cut using the conventional interpolation method (Shannon’s method). We review the Fluency sampling function, which is a locally supported function of second order. Although the use of this function will not retrieve the cut frequency components, it has a special nature that the reproduced signal includes ultrasonic components of higher frequency. Therefore, we can expect the subjective enhancement of sound quality for encoded audio data. The small support of the function is also an advantage when using on portable devices such as cell phones, as the number of calculation can be drastically saved when compared with the conventional method.

An improvement to the content-based image retrieval (CBIR) system for kaou images which has been developed by the authors group is introduced. Kaous are handwritten monograms found on old Japanese documents in a Chinese character-like shapes with artistic decorations. Kaous play an important role in the research of historical documents, which involve browsing and comparison of numerous samples. In this work, a novel method of kaou image modeling for CBIR is introduced, which incorporates the shade information of a closed kaou region in addition to the conventionally used contour characteristics. Dissimilarity of query and dictionary images were calculated as a weighted sum of elementary differences in the positions, contour shapes and colors of the component regions. These elementary differences were evaluated using relaxation matching and empirically defined distance functions. In the experiments, a set of 2455 kaou images were used. It was found that apparently similar kaou images could be retrieved by the proposed method, improving the retrieval quality. .

The paper proposes a new publishing system using PC and large-sized printer that can produce high quality printed matter comprising text, charts, illustrations, etc. As the core component of this system, a coding method that can ensure the quality of decoded images even on affine-transformed enlargement or reduction is needed. Conventional coding methods, however, cannot claim to have satisfied such a requirement. The paper presents a novel method for encoding and decoding binary images via adaptive approximation of the object contours using fluency functions, in the form of straight lines, arcs, and curved lines, that were proposed by one of the authors and others (see Kamada, M. et al., J. Approx. Theory, vol.55, no.1, p.27-34, 1988). Based on experimental results, we demonstrate that the quality of the decoded images is maintained without jaggy-noise, even on resizing. Furthermore, signboards produced by this publishing system have been used in several live events.

In this paper, a Trademark Application Assistant (TAST) that aims to speed up the process of a successful trademark application is proposed. The core of TAST is a similarity-based image retrieval system that makes use of relaxation matching on extracted contour segments. A novelty of this research lies in the processing architecture introduced for adapting the highly effective but computation expensive relaxation matching algorithm to real-time image retrieval. Preliminary results illustrating the quality of matching and the scalability of the processing architecture in terms of response time are included. The proposed architecture is not limited to using relaxation matching based servers, but is also applicable to other types of matching server as well.

The authors' previous work in image similarity retrieval by relaxation labeling processes (2001) required adding local constraints to obtain an initial set of compatible objects and labels on pairs of images to ensure labeling consistency at convergence. This approach suffers from the problem of over-specified constraints that leads to potentially invaluable information being prematurely removed. To address this problem, we introduce the idea of a Match-all label for objects that failed these constraints. It serves to give them a defined labeling probability as well as allowing them participate in global correspondences through the compatibility model. We show that this enhanced formulation still meets the conditions for a theorem on labeling consistency in Hummel and Zucker (1983) to be satisfied. At convergence, the set of objects and their most consistent labels constitute a best partial labeling for the pair of images.

This paper presents a new multiscale curvature-based shape representation technique with application to curve data compression using B-spline wavelets. The evolution of the curve is implemented in the B-spline scale-space, which enjoys a number of advantages over the classical Gaussian scale-space, for instance, the availability of fast algorithms. The B-spline wavelet transforms are used to efficiently estimate the multiscale curvature functions. Based on the curvature scale-space image, we introduce a coarse-to-fine matching algorithm which automatically detects the dominant points and uses them as knots for curve interpolation.

Classical probabilistic relaxation method has been widely used for solving optimisation problems in various fields, including image processing and pattern recognition. However we realize that there exist cases in which a probability theoretic model is not adequate, especially there exists incompleteness in available information by noises. In order to solve the problem, this paper proposes a relaxation matching method based on Dempster-Shafer theory. Then the update process in probabilistic relaxation method is derived as a special case of Dempster's combination rule (1967) in DS theory.

An observation method for mathematical graphology is proposed. The motivation of mathematical graphology is to realize graphological information processing on a machine. Furthermore, mathematical graphology aims to construct a new information theory for visual communication. In order to construct the mathematical model this paper focuses on the observation process, which is the first stage in the model. Based on the physiological saccade phenomena, feature points are extracted which include the important information. Then, contours between adjacent feature points are approximated using smooth functions. The performance of the proposed method is verified by experiments.

An automatic function-font generating system for brush-written characters has been proposed. The system extracts sharp corners of boundary lines, (so-called joint points), and approximates the boundary lines between adjacent joint points using a piecewise polynomial. The quality of the generated function-fonts is controlled by the location of joint points. An attempt is made to develop an advanced function-font generating system with an algorithm to extract appropriate joint points. The algorithm consists of three stages: (1) extracting joint points by evaluating digital curvature, (2) extracting additional joint points by evaluating the curvature of approximated boundary line, and (3) removing the redundant joint points. The performance of the method is verified by applying it to brush-written characters, printed characters, and figures.>

A training method of dictionary characters is presented for handwritten character recognition by the model based on relaxation approach. The dictionary characters are constructed by analyzing the structure of several training characters using curved line approximation and relaxation matching. The performance of the constructed dictionary characters was verified by the experiment. 956 (chars.) /spl times/ 20 (sets) in the standard database ETL-8 (Japanese KANJI and HIRAGANA) are used as sample characters for training and 10 (sets) are used as test characters (5 known sets and 5 unknown sets). The mean of the number of constructed dictionary characters is 3.81 for each category, and recognition rate is 99.96% on the known sets and 99.73% on the unknown sets.>

A new method for stamped character recognition is proposed. Conventional techniques have used a camera to observe the characters and get the image as a gray-level. However, it has been difficult to get an exact image, because gray-level noise occurs at some distortion around the characters, which makes a shadow on the characters, and also at these parts of stamped surface which are dirty or painted. The approach presented can solve these problems by using a range finder system as the observation system. The method observes the stamped characters as range image. The authors compare the characteristics of a range image to the characteristics of a gray-level image, especially about the noise, and show a noise reduction method for range images. Also, the performance is verified by an experiment.>

Various combinatorial optimization methods have been considered independently. The purpose of this study is to organize theoretically those optimization methods, which differ by their natures, and to indicate clearly what are common and what are different among those methods. The authors deal with a labeling problem in the optimization methods and propose a generalized algorithm to solve the problem. They show that the algorithm can derive the conventional optimization methods, such as the relaxation matching method, neural networks, and the dynamic programming matching method.>

An automatic seal identification method using a relative relation among segments of the seal impression is proposed. Features of a seal impression image are extracted by fluency function approximation, and matching is performed by relaxation matching. 2 (seals) /spl times/ 2 (sets) images are used to verify the performance of the present method. The recognition rate is 88.8%, and the processing speed for observing and matching is about 30 s on middle scale computers.>

The authors propose a compression method for storing the electroencephalogram (EEG) waveforms into digital storage under the condition of keeping the quality sufficient for diagnosis. The method is designed on the basis of features of the EEG waveforms. The data volume is reduced by storing the coefficients of the functions approximating the EEG waveforms. The waveforms are divided into segments, in each of which the difference velocity is similar, and are decomposed into two wave components for efficient compression. The effectiveness of the method has been evaluated by applying it to several EEG waveforms. The quality of the reconstructed waveforms is bandwidth of 0.5 approximately 120 Hz, dynamic range of 60 dB, and maximum error of 2 mu V/sub p-p/. The compression ratio is 0.648 on the average.>

A function generator that converts a discrete time signal to its interpolation by a quadratic spline function with no phase distortions is described. The generator synthesizes the output as a linear combination of B-spline functions generated coherently and continuously. The method reduces the total scale of the circuitry and improves the accuracy in the final output waveform. An implementation of the spline function generator is described.>

A psychophysical experiment was carried out to study the effects of ultrasonic components on perceived tone qualities. The physical attribute of the experiment was the decay rate of the envelope of the ultrasonic components, and its psychological attributes were the distinctness, brightness, and loudness of tones. It is shown that slower decay tends to make brightness higher and loudness lower. It is concluded that the ultrasonic components must be taken into account in acoustic signal processing as well as auditory analysis.>

A method of improving the image quality of video hard copies is proposed. This method performs bivariate quadratic spline interpolation on a digital image. The impulse response of the interpolation is derived and used to design a fast digital filter for its implementation. Simulation for an input image with a resolution of 200 TV lines shows that the method produces an output that is psychovisually equivalent to one with a resolution of 300 TV lines. Thus the image impression can be improved to about 1.5 times that of the input image. It is noted that this method is as effective for image processing tasks such as enlargement in medical image diagnosis.>

The authors derive the biorthonormal basis of the spline sampling basis to complete the biorthonormal expansion formula in the signal space composed of spline functions. This makes it possible to obtain an operator that evaluates the expansion coefficients as inner products of a given signal with the biorthonormal basis. It is also shown that to sample a signal by using the biorthonormal basis is to obtain sample values of a spline function that approximates the signal in the least mean squares sense. It is noted that one of the applications for the biorthonormal basis is to use it as an impulse response for approximation of waveforms or figures.>

A handprinted Chinese character database is introduced. Features of characters in the database are analyzed in order to develop a handprinted Chinese character recognition method based on relaxation matching. An analysis shows that this method, which is one of the most accurate methods for handprinted Kanji, is also successful in recognizing handprinted Chinese characters. The results will be used to modify the method to improve its performance for handprinted characters.>

This paper proposes a method of extracting features of handwritten Japanese characters by using spline approximation. This method represents a character by contours expressed by well-approximating functions and steady breakpoints which characterize the connection of the strokes so that it provides proper features for handwritten Japanese character recognition with relaxation matching. Its effectiveness tested on a Japanese character database is also shown.

Hybrid splines whose bases are constructed by multi-order B-splines are presented. B-splines of order 2 are induced for representing corner, and that of order 3 or 4 for representing curved parts. A presented hybrid spline of order 2 and 4 cannot be represented by conventional splines with multiple knots.

Relaxation method attracts attention as an effective method in image processing and pattern recognition. From the first, the relaxation operation has been said to be capable of parallel processing. But previous hardware implementations were specialized to each application. This paper presents a parallel architecture of a relaxation processor which can be applied to general use.

A new system for automatically analyzing left ventricular cineangiograms (LVC) is presented. The system processes images of LVC to obtain parameters available for assessing left ventricular functions. Special attention is paid on the left ventricular boundary extraction, since it is of crucial importance to all other processes. The method for reconstructing three-dimensional view of left ventricle is also described.

This paper describes a new time-varying digital filter, which is called a spline filter as it makes use of spline functions. We present the principle of the spline filter and its design technique in the time domain. An application of the spline filter to the left ventricular blood pressure measurement is shown.

The process for composing the polynomial spline function of order m which interpolates n data points is roughly divided into two stages: 1) The stage to construct the operator A which transforms B-spline coefficient vector c into sample-value vector s, 2) The stage to calculate the vector c. Then the effects on the number of computations for spline interpolation with respect to location of knots, boundary conditions, location of sampling points, and order of spline functions are evaluated.

This paper presents a new systematic series of digital filter windows which are given by linear combinations of B-splines. Then a method to design a window in the series is presented in which the specifications are given by shape of the main-lobe and its neighborhood, and damping ratio of the side-lobes.