Recently, 3-D displays used in the medicine is becoming increasingly important for accuracy judges and novel operations. We propose a 3-D CT system using holography and enhancement method its processing performance.

View full abstract

Flat-Panel Multi-view 3D Displays have been developed actively in recent years. But these displays have some pseudoscopic regions that give an unpleasant feeling to viewers. We suggest a new method to eliminate the pseudoscopic regions based on the applying visual characteristics of binocular rivalry. We confirm the validity by the experiments and show that the displeasure in pseudoscopic regions can be decreased effectively.

View full abstract

We have implemented the cubic autostereoscopic display "gCubik" by employing a wide field of view lens array. Although the lens array produces sufficient rays of virtual objects, some of rays are unnecessary. This paper studies another method for producing only the necessary rays by employing a prism array.

View full abstract

We found that the depth of image formation in a head-up display did not affect the perceived depth of a virtual sign presented in the far depth by artificial motion parallax. This finding is the evidence that we can use the small head-up display for such virtual sign.

View full abstract

In this study, the authors evaluated stereoscopic image quality for mobile devices using IBQ (Interpretation-based quality) methodology. In the experiment, paired comparisons and interviews were integrated for understanding user experience under various viewing conditions. This paper describes the results of the experiment and discusses the characteristics in the mobile stereoscopic imaging.

View full abstract

We are researching on design of screen for VR(Virtual Reality) that space recognition is close to real space is possible. In this paper, it reports on examined person's depth perception in real space by experiment with lamps.

View full abstract

This paper presents a novel interactive advertising system employing a camera that can interact with the public audiences. Since proposed system can interact with contents of advertisements without any physical contact, it's more effective than the conventional advertisements media.

View full abstract

"KaGuRa" is a musical performance method, that changes a motion of the body into audio and visual. A sound is created by the position that dancer moved, and dancer's picture which changed is reflected. We did the performance by using this in the National Museum of Nature and Science.

View full abstract

We have developed Virtual Scope, a system which uses a magnifying glass interface that is operated in front of an object for display. By increasing the display magnification so that the displayed information is unrecognizable, we were able to apply this system in the entertainment field of quiz program production.

View full abstract

Recently, techniques for free viewpoint image reconstruction are often discussed. In this paper, at first we test the FFT performance of CELL Broadband Engine. Then, in consideration of the results, we discuss the CELL-based distributed processing for fast free viewpoint image reconstruction from 3-D multi-focus imaging sequences.

View full abstract

A view-dependent polygon texture mapping is the one of efficient texture mapping method. However it causes discontinuity of textures between polygons, since different texture image which is closest from the view point is mapped for each polygon. We propose the view-dependent polygon texture mapping method which blend three closest textures from the view and present the improvement of the image quality comparing with the conventional non-blending methods.

View full abstract

We have proposed a generation method of free viewpoint video to realize "walk-through" experience using divided local regions. In this paper, we extend the method to utilize a zenith camera as well as horizontal arrayed cameras in order to realize 3D movement of virtual viewpoint such as vertical rotation. A computer simulation shows the proposed method can generate natural images from the obliquely upward virtual viewpoints.

View full abstract

In this paper, we propose a geometry based IBR method for sparse multicamera system using multiple local ray-space representation and 3D model generation. Error caused by inaccurate camera parameters and 3D models are further eliminated using the block-matching; so that free viewpoint rendering quality is improved significantly for walk-through generation.

View full abstract

We proposed a new method with foreground/background segmentation for generating time-lapse videos, which samples foreground and background frames individually. In this paper, we show that our method realizes synthesized yet natural time-lapse videos.

View full abstract

We propose a way to reproduce the color of high intensity objects using "KNEE" in capturing by multiband cameras. The characterization method of KNEE and noise characteristics in reproduced images are investigated. Experiments using a six-band HDTV video camera show that glossy objects can be reproduced naturally by the proposed method.

View full abstract

We extend a contrast enhancing image fusion method by using bilateral filters to videos and propose a technique for producing contrast enhancing corss-dissolve videos by using spatio-temporal bilateral filters.

View full abstract

We propose an acceleration technique for iterating the cross bilateral filter by fixing its weights and decomposing them into horizontally and vertically.

View full abstract

This paper presents a spatially-adaptive wavelet denoising method using the L1-L2 type color shrinkage scheme that is derived from the L1-L2 type variational problem in the wavelet transform domain. The L1-L2 type color shrinkage scheme efficiently suppresses noise in the wavelet transform domain by utilizing inter-channel color cross-correlations. The two-stage method, particularly, removes signal-dependent noise in DSC efficiently, while preserving fine image textures.

View full abstract

This paper presents a new IP-pipeline based on nonlinear image decomposition for DSC. Previously, we have presented the IP-pipeline where the multiplicative BV-G image-decomposition is applied to color channel of observed raw color data mosaicked with the Bayer CFA and then each separated component is demosaicked with an interpolation method suitable to it. This paper studies another IP-pipeline where the image-decomposition is applied to color channel of the demosaicked image by the TV-regularization deblurring-demosaicing method, and compares it with our previous IP-pipeline.

View full abstract

We propose a new TV viewing style "CurioView", that an audience satisfies various interests about viewing contents by automatic information retrieval technology. Curio View finds out videos and information related to a TV program which a user watches and Curio View shows them to the user when demanded. This paper described a concept of Curio View and the system which we produced experimentally.

View full abstract

The paper proposes a shot boundary detection method to decrease the insertion errors, while maintaining the insertion leakage at zero. In this method, a similar frame is connected by using frame-based vector space, and the shot is detected. The shot boundary can be detected successfully without insertion leakage.

View full abstract

We developed a learning assistance system with Rich Internet Application (RIA) technology. The student who is a user can search different types of information such as videos or texts by natural language question from the Web page. The system also has video report contents production function that realizes simple and easy operation by RIA technology. The user can make video contents from searched information and publish on the Web easily. These functions assist pleasant video communication among students of distant places.

View full abstract

In this paper, Spherical harmonics are used in 2D image matching. Because it is a 3D descriptor, the 2D image should be reconstructed to 3D shape. Thus, we use two different methods: direct mapping and mapping based on the Logpolar Transform. For modeling, the effectiveness of our method is confirmed in experiments.

View full abstract

In recent years, because of the rapid progress in wireless Technology, microwave absorbers have become necessary for a safe-electromagnetic-wave-level environment. In this study, based on the simulated and experimental values, we explain characteristic evaluation of microwave absorber for wireless LAN (local area network) using a resistive-film, dielectric materials and a variable pattern sheet. And the results of their reflection losses, in free space, using TE and TM waves, were -20dB or more around both the central frequencies of 2.45 GHz and 5.2 GHz, satisfying the conditions for wireless LAN, and it compares experimental results with the EEM-FDM simulation results.

View full abstract

In broadcasting-wave relay station on SFN, the developed distance meter can support effectively counter measure to improve DU ratio to which it deteriorates due to Coupling Loop Interference because of reflection.

View full abstract

Transmission characteristics of pillar-type two dimensional, single and straight photonic crystal waveguides with three types of width are measured by making use of models for microwaves around 4 GHz.

View full abstract

It's difficult to find diffraction points on two edges which make arbitrary angle with different heights. This paper shows the method that can easily find these diffraction points.

View full abstract

We have been studying the advanced satellite broadcasting system which is aimed for the improvement of the Integrated Services Digital Broadcasting for the Satellite (ISDB-S) system. Making use of LDPC codes and reduced Roll-Off factor, this advanced system can increase its transmission channel capacity about 30% larger than the ISDB-S system The LDPC code in this system is designed for PSK and APSK modulation schemes, and furthermore this code lengths is applicable to any other modulation schemes. As a example, it can be used for 16/64/256/1024 QAM modulations. In this report, we describe the multi-level QAM modulation schemes using this LDPC codes and the computer simulation results of performances under the AWGN channel model.

View full abstract

This paper is concerned with an FVTD simulation for relationship between spectrum of random rough surface and propagation characteristic. Based on the direct DFT method, we generate rough surfaces with some spectra, Gaussian, n-th order power law, exponential spectrum. We numerically compute electromagnetic fields above random rough surfaces by the FVTD.

View full abstract

This paper is concerned with the comparison between the urban propagation loss derived from the Hata method based on Okumura curve in mobile telecommunications and our approximation for predicting electromagnetic field. Two parameters, distance and order, are modified so that our results may agree with the results obtained by the Hata method.

View full abstract

Discrete ray tracing method (DRTM) is proposed to numerically analyze electromagnetic wave propagation along 2D rough surfaces. In the DRTM analysis, we make some assumptions in order to simplify the ray serching algorithm as well as to save computaion time. Numerical examples are given for 1D and 2D random rough surfaces and electric field distributions along them.

View full abstract

This study investigated on the voice quality words of normal utterance. We carried out subjective evaluation experiments for normal voices. Finally, we extracted the ten words for evaluating voice quality using cluster analysis and factor analysis.

View full abstract

In this paper, we propose the realtime evaluation system of speech coding method for cochlear implant which investigates not only subjective assessment but also objective one. the system provides displaying stimulus level of electrode, generating the simulated sound, and "transmitting pulse code to cochlear implant" (" "in future).

View full abstract

Color assimilation observed with stripes against a dark background was consistently stronger with peripheral view than with foveal view. The observation suggests that color assimilation is caused by wider intercone spacing in the peripheral area of the retina as well as by the ocular chromatic aberration.

View full abstract

We have developed a shot supporting system for a beginner on Billiards. It recognizes each ball and player's cue-stick with a camera over the table and projects the direction of his effective shot and information for correcting his stance. In this paper, we describe our experimental result of evaluating it.

View full abstract

We use digital maps of different scales in various situation. However, the rough map based on the detailed map is made by manual labor now. In this paper, we propose the method of making the line with river data in order to make maps of different scales automatically.

View full abstract

The purpose of this paper is to investigate the relationship between understandability and display conditions of sign-animation. Results showed that resolution and frame-rate influenced greatly in reading of sign-animation in large size display. Also, we clarified the display condition that was most suitable for each display size.

View full abstract

We have been developing a portable environment verification system that supports the low visions. We call this system V-aid. The V-aid consists of a μ-CMOS camera and PDA. This time we propose the image processing algorithm that confirms obstacle of front in walking with this V-aid system.

View full abstract

Web 2.0 facilitates the construction of the autonomous decentralized network "outside" enterprises. This network is constituted by each autonomous customers and supplier. In this research, we describe a new wave of organization model "inside" enterprises. Yumenomachi Souzou Iinkai is one successful enterprise on Internet portal business in Japan. The detail of the management and business model of the case were described in this research. The case indicates that a new enterprise model based on an autonomous decentralized organization is emerging in IT industry in Japan.

View full abstract

We discuss on fab-less and license models for venture startups of cutting-edge technologies with analysis on needs for partnering and its environment, and propose license model to provide a flexibility to combine technologies of licenser's and licensee's as a successful role model of technology partnering in Japan.

View full abstract

Recently, improvement of corporate image is demanded in marketing. In this research, We grouped retail trade by corporate scale and profit. For each group, We investigated corporate image. And we analyzed feature of corporate image in every group.

View full abstract

Recently, people take a strong interest in evaluation methods about the company and corporate performance. In addition to evaluate these factors by balance sheet and profit-and-loss statement, we propose a new evaluation method of company using the new quantitative process which combined the new qualitative one on viewpoint of KANSEI engineering.

View full abstract

Increasing number of industries are becoming highly linked with science. Many scholars are interested in the meanings of the science for the industry. I analyze the proceedings of the SID (Society for Information Display) and discuss the relationships between them.

View full abstract