In this paper, the LVQ (Learning Vector Quantization) model and its variants are regarded as the clustering tools to discriminate the natural seismic events (earthquakes) from the artificial ones (nuclear explosions). The study is based on the six spectral features of the P-wave spectra computed from the short period teleseismic recordings. The conventional LVQ proposed by Kohenen and also the Fuzzy LVQ (FLVQ) models proposed by Sakuraba and Bezdek are all tested on a set of 26 earthquakes and 24 nuclear explosions using the leave-one-out testing strategy. The primary experimental results have shown that the shapes, the number and also the overlaps of the clusters play an important role in seismic classification. The results also showed how an improper feature space partitioning would strongly weaken both the clustering and recognition phases. To improve the numerical results, a new combined FLVQ algorithm is employed in this paper. The algorithm is composed of two nested sub-algorithms. The inner sub-algorithm tries to generate a well-defined fuzzy partitioning with the fuzzy reference vectors in the feature space. To achieve this goal, a cost function is defined as a function of the number, the shapes and also the overlaps of the fuzzy reference vectors. The update rule tries to minimize this cost function in a stepwise learning algorithm. On the other hand, the outer sub-algorithm tries to find an optimum value for the number of the clusters, in each step. For this optimization in the outer loop, we have used two different criteria. In the first criterion, the newly defined "fuzzy entropy" is used while in the second criterion, a performance index is employed by generalizing the Huntsberger formula for the learning rate, using the concept of fuzzy distance. The experimental results of the new model show a promising improvement in the error rate, an acceptable convergence time, and also more flexibility in boundary decision making.

To implement soft handoff in cellular communication systems that employ code division multiple access (CDMA), it is necessary to establish communication lines between the switch and multiple base stations and distribute the communication data via these multi-connections to the base stations simultaneously. This means that, when soft handoff is performed with the same amount of communication line resources as hard handoff, the blocking probability is higher than for hard handoff, and service quality is thus worse. Furthermore, handoffs occur more frequently as the size of cells becomes smaller, and this increases the probability of forced terminations. Switches must be endowed with greater processing capacity to accommodate the more frequent handoffs. The use of the queuing handoff method can be expected, in general, to mitigate forced termination probability compared with the immediate handoff method. In this regard, we propose a prioritized queuing handoff method that gives priority to fast-moving mobile stations (MSs) as a way to mitigate forced terminations even more than the non-priority queuing method without appreciably increasing the processing load. We then compare the traffic characteristics of our proposed method with these of three other methods in micro cell systems--immediate method, non-priority queuing method, and conventional hard handoff method without multi-connections--by computer simulation. Here, considering that the proposed method gives priority to fast-moving calls, traffic characteristics for these methods were evaluated separately for slow- and fast-moving MSs. The results reveal that proposed method can reduce the forced termination probability and total call failure probability more than non-priority queuing method without having an appreciable impact on slow-moving calls.

Channel-state-dependent (CSD) radio-resource scheduling algorithms for wireless message transport using a framed ALOHA-reservation access protocol are presented. In future wireless systems that provide Mbps-class high-speed wireless links using high frequencies, burst packet errors, which last a certain number of packets in time, would cause serious performance degradation. CSD resource scheduling algorithms utilize channel-state information for increasing overall throughput. These algorithms were comparatively evaluated in terms of average allocation plus transfer delay, average throughput, variance in throughput, and utilization of resources. Computer simulation results showed that the CSD mechanism has a good effect, especially on equal sharing (ES)-based algorithms, and also CSD-ES provides low allocation plus transfer delay, high average throughput, low variance in throughput, and efficient utilization of radio resources.

Given a set of still images taken from a hand-held camera, we present a fast method for mosaicing them into a single blended picture. We design time- and memory- efficient still image mosaicing algorithms based on geometric point feature matchings that can handle both arbitrary rotations and large zoom factors. We discuss extensions of the methodology to related problems like the recovering of the epipolar geometry for 3d reconstruction and object recognition tasks.

Mobile radio and Internet communications services are penetrating our society at an exponential rate of growth. The Internet is the most important driving force towards establishing a multimedia society. Mobile communication systems add an important capability to our communications society, i. e. , mobility. The third generation mobile communications system called IMT-2000 is expected to play an important role in this soon-to- arrive multimedia society. Wideband mobile services based on IMT-2000 will soon become a reality (early 2001). First, we look at the trends of wireless access technologies, centering on IMT-2000. Wideband direct sequence code division multiple access (W-CDMA) will be a major component of a global IMT-2000 standard. Then, we address advanced wireless techniques, i. e. , interference cancellation and employing an adaptive antenna array, which can enhance W-CDMA at a later date. Finally, requirements are discussed for future wireless techniques that will support a fully mobile multimedia communications society.

In this paper, we propose a peak-weighted cepstral lifter (PWL) for enhancing the spectral peaks of an all-pole model spectrum in the cepstral domain. The design parameter of the PWL is the degree of pole enhancement or pole shifting toward the unit circle. The optimal pole shifting factor is chosen by considering the sensitivity to spectral resonance peaks, the variability of cepstral variances, and the recognition accuracy. Next, we generalize the PWL so that the optimal shifting factor is adaptively determined in frame-by-frame basis. Compared with other cepstral lifters, a speech recognizer employing the frame-adaptive PWL provides better recognition performance.

Saltiness elicited by salt is one of the basic tastes. However, components of salt on the market differ depending on manufacturing processes and its taste as well. Salt manufactured by ion-exchange membrane process is composed of more than 99% pure sodium chloride, while bay salt contains trace coexisting components. Despite reports on sensory evaluation, the differences in taste are still uncertain because of a small amount of coexisting components. We studied the taste of salt with trace coexisting components; the bittern ("nigari" in Japanese) was evaluated objectively and quantitatively using a multichannel taste sensor with lipid/polymer membranes. A taste sensor is comprised of several types of lipid/polymer membranes for transforming information of taste substances into electric signals. The model samples were composed of sodium chloride and trace coexisting components such as magnesium sulfate, magnesium chloride, calcium chloride and sodium chloride. The taste sensor clearly discriminated each sample according to the response patterns. Based on the sensor outputs, we evaluated the taste by means of the combination of principal component analysis and ionic strength. The results show the taste of salt with nigari has a correlation with ionic strength.

In this paper, a modified neural network approach called the Guided Neural Network is proposed for the longitudinal dynamics identification of a vehicle using the well-known gradient descent algorithm. The main contribution of this paper is to take account of the known information about the system in identification and to enhance the convergence of the identification errors. In this approach, the identification is performed in two stages. First, the Guiding Network is utilized to obtain an approximate dynamic characteristics from the known information such as nonlinear models or expert's experiences. Then the errors between the plant and Guiding Network are compensated using the Compensating Network with the gradient descent algorithm. With this approach, the convergence speed of the identification error can be enhanced and more accurate dynamic model can be obtained. The proposed approach is applied to the longitudinal dynamics identification of a vehicle and the resultant performance enhancement is given.

Polymeric thin films can be prepared by physical vapor deposition in several manners such as direct evaporation of the polymer, co-evaporation of two monomers followed by polyaddition or polycondensation reaction, or evaporation of single monomer followed by chain polymerization. The ionization-assisted deposition (IAD) was proposed as a new method of polymer deposition that has special features such as activation of polymerization reaction and aligning of the dipole orientation. These mechanisms were utilized for the formation of vinyl polymer and polyurea thin films aiming for such applications as organic light emitting diodes and piezoelectric devices.

A novel chemical sensor for sulfate detection was proposed in this study, utilizing sulfate binding protein (SBP) derived from Escherichia coli as sulfate recognition element. Purified SBP was immobilized on a gold electrode modified with cysteamine and glutaraldehyde. In this study the surface potential change of the SBP modified electrode to sulfate and various ions were investigated. In order to evaluate nonspecific interaction with ionic species, proteins with various isoelectric point were immobilized on the surface of gold electrode and response to ions were measured and compared to sulfate binding protein modified electrode. We made clear that the protein modified electrode shows the potential change to ions and these potential change was effected by the isoelectric point of the protein molecule, and BSA, whose isoelectric point is closest to that of SBP, showed the similar response to ions except sulfate. With use BSA modified electrode as a reference electrode, this sensing system showed selective response to sulfate, probably because of the selective binding sulfate by SBP. This potential change difference between the SBP modified electrode and the BSA modified electrode depended on the concentration of sulfate with in the range of 5 - 150 mM.

We developed a new water repellent coating consisting of PTFE particles dispersed in PVDF resin. This coating exhibited a contact angle of 150 degrees. By ice accreting test, the intensity of reflected microwave on the water-repellent coated plate did not decrease, whereas that on uncoated one decreased.

In this paper, a method of fusing ground-based laser range image and CCD images for the reconstruction of textured 3D urban object is proposed. An acquisition system is developed to capture laser range image and CCD images simultaneously from the same platform. A registration method is developed using both laser range and CCD images in a coarse-to-fine process. Laser range images are registered with an assumption on sensor's setup, which aims at robustly detecting an initial configuration between the sensor's coordinate system of two views. CCD images are matched to refine the accuracy of the initial transformation, which might be degraded by improper sensor setup, unreliable feature extraction, or limited by low spatial resolution of laser range image. Textured 3D model is generated using planar faces for vertical walls and triangular cells for ground surface, trees and bushes. Through an outdoor experiment of reconstructing a building using six views of laser range and CCD images, it is demonstrated that textured 3D model of urban objects can be generated in an automated manner.

A new method for measuring the scattering coefficient using a metal-plate reflector was developed in order to provide a non-destructive way for the assessment of microwave materials in free space. By displacing the position of the metal-plate reflector on the specimen to be tested, the incident wave and the scattered wave from the measured area were determined without the influence of extraneous waves such as the direct coupling between transmitting and receiving antennas and scattered waves from background objects. Because the behavior of a metal-plate reflector is similar to that of an optical shutter in optics, our new scattering measurement system enables us to measure both backward- and forward-scattering coefficients of small regions of the specimen for various types of materials in a non-destructive manner. Our study examined the metal-plate size dependence of the complex reflection and transmission coefficients of some dielectric sheet samples. The measured data indicated that the reflection and transmission coefficients of a Bakelite flat plate and Styrofoam sheet were constant for various sizes of metal plates at the X-band.

Since odor sensing system is required in many fields, we have developed the system using QCM (Quartz Crystal Microbalance) sensor array and neural-network pattern recognition. In the present study, the mixed sensing films of two kinds of liquid-phase materials were characterized. As a result, it was found that the variety of sensing films were obtained by mixing two kinds of liquid-phase materials. The relative remnant of sensing film after repeated exposures was examined, and mixed films of two kinds of liquid-phase materials were found stable for the sensing materials.

Short-circuit photocurrents (ISC) utilizing surface plasmon polariton (SPP) excitation were investigated for the merocyanine (MC) LB film photoelectric device. The device has a prism/MgF2/Al/MC LB film/Ag structure. In the attenuated total reflection (ATR) configuration, SPPs were resonantly excited at the interfaces between MgF2 and Al (MgF2/Al) and between Ag and air (Ag/air). The thickness and the dielectric constants of the layers were evaluated from the ATR measurements. Short-circuit photocurrents, ISCs, as a function of the incident angle of the laser beam were observed simultaneously during the ATR measurements. In the ISC curves, large and small peaks were observed, and the peak angles of the ISC almost corresponded to the dip angles of the ATR curves due to the SPP excitations. Electric fields and optical absorptions in the cell were calculated using the dielectric constants and the film thickness obtained from the ATR measurements. The calculated absorption in the MC layer as a function of the incident angle corresponded to the ISC curve. It was thought that the optical absorption in the MC layer affected directly to the profile of the ISC. Furthermore, the calculated absorption in the cell with the prism and the MgF2 layer exhibited much larger than that of the cell without them. It was estimated that the photocurrents were enhanced by the excitation of SPPs in the ATR configuration.

We have developed a new type electrical probing system based on an atomic force microscope. This method enables us to measure simultaneously the surface topography and surface potential of thin films containing the crystal grains. The obtained local potential changes give an insight into conduction through the grains and their boundaries.

In this paper, we propose a fast algorithm to realize parallel median filter for processing 1-D and 2-D signal. In the proposed pipelined architecture, m-passes are employed for filtering signal while word resolution is m bits. One pass employs one processing element (PE), and the number of PEs is independent of the number of samples. Therefore, we only need m PEs for real-time operation. With 8-bits resolution, the system gate-count is less than 5 k. Moreover, this median architecture could be easily modified to consist of the programmable feature that may choose the better sampling number to filter signal. It should be also noted that our proposed processing flow has a progressive property, which is very suitable for bandwidth-limited channel application.

The k-mutual exclusion problem is the problem of guaranteeing that no more than k computing nodes enter a critical section simultaneously. The use of a k-coterie, which is a special set of node groups, is known as a robust approach to this problem. In general, k-coteries are classified as either dominated or nondominated, and a mutual exclusion mechanism has maximal availability when it employs a nondominated k-coterie. In this paper, we propose two new schemes called VOT and D-VOT for constructing nondominated k-coteries. We conduct a comparative evaluation of the proposed schemes and well-known previous schemes. The results clearly show the superiority of the proposed schemes.

The visual inspection of printed circuit boards (PCBs) at the final production stage is necessary for quality assurance and the requirements for an automated inspection system are very high. However, consistent inspection of patterns on these PCBs is very difficult due to pattern complexity. Most of the previously developed techniques are not sensitive enough to detect defects in complex patterns. To solve this problem, we propose a new optical system that discriminates pattern types existing on a PCB, such as copper, solder resist and silk-screen printing. We have also developed a hybrid defect detection technique to inspect discriminated patterns. This technique is based on shape measurement and features extraction methods. We used the proposed techniques in an actual automated inspection system, realizing real time transactions with a combination of hardware equipped with image processing LSIs and PC software. Evaluation with this inspection system ensures a 100% defect detection rate and a fairly low false alarm rate (0.06%). The present paper describes the inspection algorithm and briefly explains the automated inspection system.

A novel method is proposed to retrieve image sequences with the goal of forecasting complex and time-varying natural patterns. To that end, we introduce a framework called Memory-Based Forecasting; it provides forecast information based on the temporal development of past retrieved sequences. This paper targets the radar echo patterns in weather radar images, and aims to realize an image retrieval method that supports weather forecasters in predicting local precipitation. To characterize the radar echo patterns, an appearance-based representation of the echo pattern, and its velocity field are employed. Temporal texture features are introduced to represent local pattern features including non-rigid complex motion. Furthermore, the temporal development of a sequence is represented as paths in eigenspaces of the image features, and a normalized distance between two sequences in the eigenspace is proposed as a dissimilarity measure that is used in retrieving similar sequences. Several experiments confirm the good performance of the proposed retrieval scheme, and indicate the predictability of the image sequence.