In this paper, a synthesizing method is proposed for the design of discrete-time cellular neural networks for binary image processing. Based on the theory of digital-logical design paradigm of threshold logic, the template parameters of the discrete-time cellular neural network for a prescribed binary image processing problem are calculated. Application examples including edge detection, connected component detection, and hole filling are given to demonstrate the merits and limitations of the proposed method. For a given realization of the parameters of the cloning template, a guideline for the selection of the offset Ic for maximum error tolerance is also considered.

A low-voltage, high-speed 4-bit CMOS single chip microprocessor, with instruction execution time of 1.0µs at a power supply voltage of 1.8V, has been developed. A single chip processor generally includes crystal oscillation circuits to generate a system clock or a time-base clock. But when the operating voltage is lowered, it becomes difficult to get oscillations to start reliably and to continue stably. This paper describes a low voltage circuit design method for built-in crystal oscillators. Simple design equations for oscillation starting voltage and oscillation starting time are introduced. Then effects of the circuit device parameters, such as power supply voltage, loop gain values, and subthreshold swing S, on the low voltage performance of the crystal oscillators are considered. It is shown that the crystal oscillators operate in a tailing (subthreshold) region at voltages lower than about 1.8 V. Subthreshold swing, threshold voltage, and open loop gain have a significant influence on low voltage oscillation capability. This design method can be applied to crystal oscillators for a wide range of operating voltages.

In this article a Neural Network learning scheme is described, which is a generalization of VQ (Vector Quantization) and ART2a (a simplified version of Adaptive Resonance Theory 2). The basic differences between VQ and ART2a will be exhibited and it will be shown how these differences are covered by the generalized scheme. The generalized scheme enables a rich set of variations on VQ and ART2a. One such variation uses the expression ||I||2+||zj||2/||zj||sin(I,zj), as the distance measure between input vector I and weight vector zj. This variation tends to be more robust to noise than ART2a, as is shown by experiments we performed. These experiments use the same data-set as the ART2a experiments in Ref.(3).

Link conditions of a low-earth orbit (LEO) satellite communications system were evaluated, to provide the information necessary for designing a broadband LEO-SAT communications system. The study was made both for optical intersatellite and user/satellite links. For the optical intersatellite link (ISL), we examined several ISL configurations in a circular polar orbit, and found that when the satellites are in the same orbital plane, the link parameters are quite stable, that is, the link between adjacent satellites can be regarded as fixed and, therefore, suitable for broadband transmission via an optical link. However, the link conditions between adjacent orbits change very quickly and over a wide range. To overcome this and extend the network path between satellites in adjacent orbital planes, we proposed intermittent use of the link between satellites in co-rotating adjacent orbital planes at the low latitude region, i.e., only during the period of stable conditions. The optical intersatellite link budget also sets link parameters that are realistic, given present optoelectronic technologies. From quantitative evaluations of the user/satellite link, we believe that both the satellite altitude and minimum elevation angle are critical, both in defining the quality of the service of the LEO-SAT system and in their impact on the other transmission parameters. The link loss, the visible period and the required number of satellites vs. satellite altitude and elevation angle are also indicated. These are important considerations for future system design.

Abrupt variations of attractors caused by argumental discreteness in non-Hermitian complex-valued neural networks are reported. When we apply the complex-valued associative memories to dynamical processing, the weighting matrices are constructed as non-Hermitian in general so that they have motive force to the signal vectors. It is observed that competitions between argumental rotation force and noise-suppression ability of associative memories lead to trajectory distortions and abrupt variations of the attractors.

This paper proposes a superposed SSMA (Spread Spectrum Multiple Access)-QPSK (Quadrature Phase Shift Keying) signal transmission scheme over high speed QPSK signals to achieve higher frequency utilization efficiency and to facilitate lower power transmitters for SSMA-QPSK signal transmission. Experimental results show that the proposed scheme which employs the coding-rate of one-half FEC (Forward Error Correction) and a newly proposed co-channel interference cancellation scheme for SSMA-QPSK signals can transmit twenty SSMA-QPSK channels simultaneously over a nonlinearly amplified high speed QPSK signal transmission channel and achieve as ten times SSMA channels transmission as that without co-channel interference cancellation when the SSMA-QPSK signal power to the high speed QPSK signal power ratio equals -30dB. Moreover, cancellation feasibility generation of the interference signals replica through practical hardware implementation is clarified.

In the PICASSO, a system for the facial caricature generation, as the basic mechanisms to extract the individuality features of faces and to deform the features have been already introduced, it is expected to realize an autonomous mechanism to evaluate facial caricatures. The evaluation should be based on the framework of human visual cognition. In the PICASSO, some visual illusions such as the Wundt-Fick illusion and the Ponzo illusion for example, are applied to evaluate the shapes of the facial parts such as eyebrows, nose, mouth and face contour, in the deformation process. In many cases, as well-deformed caricatures are evaluated to be successful, it is confirmed that the utilization of the visual illusion is effective to evaluate the results of caricatures. In this paper, some experimental results are presented together with the definition of the evaluation measures and the further subjects.

We describe an approach to describe moving pictures in terms of their structural properties for video editing, video indexing, and video coding. The description contains 2D shape, motion, spatial relation, and relative depth of each region. To obtain the description, we develop the incremental segmentation scheme which includes dynamic occlusion analysis to determine relative depths of several objects. The scheme has been designed along the analysis-by-synthesis" approach, and uses a sequence of images to estimate object boundaries and motion information successively/incrementally. The scheme consists of three components: motion estimation, prediction with dynamic occlusion analysis, and update of the segmentation results. By combining the information from extended (longer) image sequences, and also by treating the segmentation and dynamic occlusion analysis simultaneously, the scheme attempts to improve successively over time the accuracy of the object boundary and motion estimation.

A new redundancy technique especially suitable for ultra-high-speed static RAMs (SRAMs) has been developed. This technique is based on a decoding-method that uses two kinds of fuses without introducing any additional delay time. One fuse is initially ON and can be turned OFF afterwards, if necessary, by a cutting process using a focused ion beam (FIB). The other is initially OFF and can be turned ON afterwards by a connecting process using laser chemical vapor deposition (L-CVD). This technique is applied to a 64 kbit SRAM having a 1.5-ns access time. The experimental results obtained through an SRAM chip repaired using this redundancy technique show that this technique does not introduce any increase in the access time and does not reduce the operational margin of the SRAM.

By optimizing the structure of erbium-doped fibers, high efficiency such as a gain coefficient of 6.3dB/mW, or a slope efficiency of 92.6% have been realized with very flat wavelength dependence. Though the optimized structure has high NA, the splice loss with standard fibers can be lowered by the additional arc technique. The carbon coated fiber with a fatigue parameter over 150 guarantees the reliability, even when wounded on a small coil. In-line isolators and WDM couplers have been also developed. An amplifier module has been assembled, resulting in an output power more than +16dBm owing to the high performance of each component.

This paper focuses on methodology underlying the application to fault tolerant computer systems with "no down communication" capability of stochastic Petri nets with general firing times. Based on a formal specification of the stochastic Petri net, we provide criteria for the marking process to be a regenerative process in continuous time with finite cycle-length moments. These results lead to strongly consistent point estimates and asymptotic confidence intervals for limiting system availability indices. We also show how the building blocks of stochastic Petri nets with general firing times facilitate the modeling of non-deterministic transition firing and illustrate the use of "interrupter input places" for graphical representation of transition interruptions.

This paper describes the novel relaxation-based algorithm for the harmonic analysis of nonlinear circuits. First, we present Iterated Spectrum Analysis based on harmonic balance method, where the harmonic balance method is applied to every node independently. As a result, we can avoid dealing with large scale Jacobian matrices and reduce the total simulation time, compared with the conventional method based on Galerkin's procedure or the harmonic balance method. Next, we define the frequency domain latency. Furthermore, we refer to the possibility for exploitation of three types of latency, i.e., relaxation iteration latency, frequency domain latency and Newton iteration latency. And we propose the multirate-sampling technique based on the consideration of the frequency domain latency. Finally, we apply the present technique to the simple analog circuit simulation and verify its availability for the harmonic analysis.

This paper describes a novel but simple method of implementing waveform relaxation technique for bipolar circuits involving ECL gates. This method performs gate level partitioning of ECL circuits not only during the cutoff state of the input transistor but also when the input transistor is in its active state. Partitioning at all times has become possible due to the favorable property of input and output stages of ECL gates. It is shown that this method is faster than direct method even when the circuits containing only few gates is simulated. Further, it is shown that the present method is applicable to the case where the interconnections between the ECL gates is treated as lossy transmission lines.

Formal verification has become an increasing prominent technique towards establishing the correctness of hardware designs. We present a framework to specifying and verifying the design of systolic architectures. Our approach allows users to represent systolic arrays in Z specification language and to justify the design semi-automatically using the verifier. Z is a notation based on typed set theory and enriched by a schema calculus. We describe how a systolic array for matrix-vector multiplication can be specified and justified with respect to its algorithm.

The three layer neural network (TLNN) is treated, where the nonlinearity of a neuron is of signum. First we propose an expression of the discriminant function of the TLNN, which is called a linear-homogeneous expression. This expression allows the differentiation in spite of the signum property of the neuron. Subsequently a learning algorithm is proposed based on the linear-homogeneous form. The algorithm is an error-correction procedure, which gives a mathematical foundation to heuristic error-correction learnings described in various literatures.

Gradient-based methods for the computation of the velocity from image sequences assume that the velocity field varies smoothly over image. This creates difficulties at regions where the image intensity changes abruptly such as the occluding contours or region boundaries. In this letter, we propose a method to overcome these difficulties by incorporating the information of discontinuities in image intensity into a standard local optimization method. The presented method is applied to the synthetic and real images. The results show that the velocity field computed by the proposed method is less blurred at region boundaries than that of the standard method.

In formulating the motion constraint equation, we implicitly take it for granted that the spatial and temporal sampling intervals are very small. In real situations, since the intervals cannot be considered sufficiently small, an error will be introduced into the constraint equation and consequently the velocity estimate will be subject to an error due to inaccuracy of the constraint equation. We perform some experiments to analyze the effect of sampling interval on motion estimation. The understanding of experimental results will provide an insight into necessity and amount of image filtering prior to the application of motion estimation.

Correspondence based on regions rather than lines seems to be effective, as regions are usually fewer than other image features and provide global information such as size, color, adjacency, etc. In this paper, we present a region matching approach for solving the correspondence problem. Images are segmented into regions and are individually described by classification tables using region adjacencies. From the structural description of the two images, the region matching process based on color and structural similarity is carried out. First, a small number of significant regions are selected and matched by using color, and then they are used as handles for constraint propagation to match the remaining regions by using structures. Our technique was implemented by using an efficient selection and propagation algorithm and was tested with a variety of scenes.

The authors recently proposed a design method of stable IIR digital filters based on the interpolation by rational characteristic functions of filters, for a set of values of these characteristic function and, in addition, their higher derivatives prescribed at a number of frequency. This method can be further extended so that, despite usage of a less number of interpolation points, almost the same filter characteristics as one obtained by the former method can be realized. This paper presents an improved design method for making the transfer function meet strict magnitude specifications. The method proposed in this paper is especially efficient for designing a filter whose characteristics is specified not only in the passband but also in the transition band with relatively narrow bandwidth.

In this letter, we propose an algorithm to estimate the optical flow fields based on a hierarchical structure composed of spatio-temporal image pyramids obtained from repetitive application of the Gaussian filtering and decimation in both the spatial and temporal domain. In our approach, an inter-level motion smoothness constraint between adjacent pyramid levels is introduced to estimate a unique optical flow field. We show that the pyramid structure allows us to employ the multigrid algorithm, which is known to accelerate the convergence rate. The multigrid algorithm provides a scheme for efficient combination of local and global information to estimate the optical flow field. The experimental results reveal that the combination of local and global information yields a fast convergence behavior and accurate motion estimation results.