Japanese semiconductor industry has been losing competitive advantage during the last 20 years. It is crucial to enhance capability to create products with potential to be de-facto standards or products used as a platform. This paper discusses an innovation process utilizing a case of two layer polyimide substrate development and business development.

View full abstract

The repayment deadline when the full amount borrowed new opening funds of the medium-scale hospital, is inspected and more than 15 years are needed.

View full abstract

A Regional Development Base on Resources and Economic Revitalization

View full abstract

I will announce the details "Manufacturing Value Innovation" from the failure and the success stories of Komatsu entrepreneurship".

View full abstract

Dome theaters are suitable to provide enhanced sense of presence to viewers. Three-dimensional sound is also essential for the enhancement of presence. Acoustic measurement was carried out to study issues in terms of application of three-dimensional multichannel sound system in a dome theater and possible solutions are discussed.

View full abstract

We describe a calculation method of computer-generated-hologram(CGH) using approximate Fresnel integral. The method can generate a CGH from a 3D object composed of polygons to express the polygons as many planer patches. Propagation of the planer patches to a CGH is calculated by approximate Fresnel integral.

View full abstract

In this paper, we propose a disparity estimation method using hierarchical belief propagation (BP) and Zero-mean Normalized Cross-Correlation (ZNCC). Experimental results show that the proposed method is robust against luminance level difference among stereo camera images. This method is useful for generation of 3D models and their display on 3DTV using Integral Photography (IP).

View full abstract

We are investigating a converting method of 3D models into integral 3D TV. In this paper, we propose high speed-up and image quality improvement techniques for the conversion and experimental results show the effectiveness.

View full abstract

We have been proposed a pixel-offset error measurement method using a lens array, which is adopted to integral three-dimensional imaging. In the method the error of pixel-offset distance between G1 and G2 panels was expanded by the lens array and can be easily detected in high accuracy. However it was not clarified about optimum spatial frequency of pattern and magnification ratio for measurement. In this paper, we discussed about MTFs theoretically and made clear the optimum magnification parameter experimentally.

View full abstract

This paper describes a method of producing scenes for TV programs, using "dynamic 3D human models" which are generated from multi-view cameras. They can provide highly realistic images as well as natural motions. In addition, they allow production to be more flexible, such as enabling the positions of models and cameras to be changed or multiplying the number of models. We succeeded in producing crowd scenes with sufficient quality for a TV program, which was aired in December, 2009. We describe the production flow and some scenes in the program.

View full abstract

Model-based camera tracking methods basically require visual cues to be correctly tracked as possible. However, the requirement is not likely to be satisfied in real environments. Incorrectly tracked visual cues result in jitter which annoys users in augmented reality. Therefore, we propose a jitter reduction method which is based on the comparison of the magnitude of the motion of visual cues with that of a roughly estimated camera motion. Through experimental results with real and synthetic images, we verify the effectiveness of the proposed method.

View full abstract

The authors' group has been developing a geometric modeling system which generates four-dimensional mesh models using a time sequence of three dimensional voxel models captured from OpenGL scene animation. Each four-dimensional mesh model includes object's dynamic information like movement and deformation.

View full abstract

Structural color has strong dependence on the position of light source and viewpoint. In the industrial field, the structural color is reproduced by such as the Chromaflare pigments have a structure of multi-layered interfering thin film. This article shows how to render the structural color of the multi-layered thin film.

View full abstract

In this paper, we propose a method of estimating illumination distribution from a single real image including a human face. First, we extract the domain of the human face from an image. Next, we estimate the light source position by comparing it with the 3D face model about shading.

View full abstract

In this paper, we describe Camera Blocking Simulator based on TVML. By using this system, program directors can do camera blocking easily and efficiently on personal computers.

View full abstract

We are studying a software camera for TV program production. We have developed the software camera experimentally by using GPU. In this report, we describe the algorithm of the image processing and the result of the experiments.

View full abstract

We propose a new tone-mapping method which is aimed at HDTV video processing, intended for forthcoming High Dynamic Range camera we develop. Our method has a feature of interactive tonal control especially of lower luminance level, utilizing the information both of mean luminance and local contrast.

View full abstract

We propose a method for measuring a facial change degree by image processing. The degree is calculated by using SURF (Speed-Up Robust Features) detected in the facial image. Our method is evaluated by comparing images input to the measure system, which we developed, with values output from the system.

View full abstract

This research aims at constructing a content-based image retrieval system by using gradient-based and color-based local features: scale-invariant feature transform and color index local auto correlation. We utilize the canonical correlation analysis to compare class vectors and features in a common space. Experimental results show the effectiveness of the proposed method.

View full abstract

Abstract. In this paper, we propose a refinement process for a visual hull reconstructed by shape from silhouettes algorithm. Our proposed method utilizes geometric consistency as well as photo consistency, and determines unnecessary voxels to be removed based on minimization of energy functions in a voxel space. Experimental results show the effectiveness of the proposed method.

View full abstract

This paper investigates solutions to improvement of realistic rendering quality of CG (Computer Graphics) character face. In order to manipulate face skin feature distribution, based on bio-anatomic structure, skin surface is classified into 3 layers(Internal, Regular and Local layer) to control base color, surface cellular venation level and noise distribution like freckle, respectively. By doing this, a specific transformation from real human face skin data to virtual character is proposed which is meaningful for the synthesis of the CG texture.

View full abstract

Rate-compatible asymmetric Slepian-Wolf coding with non-binary low-density parity-check (LDPC) codes of moderate code length is presented. The proposed encoder and decoder are rate-compatible with only a single mother LDPC code. With the proposed scheme, better compressed rate and lower error rate than those of conventional scheme with optimized binary LDPC codes are achieved with even smaller source length.

View full abstract

Distribution of code length of MPEG-2 DCT Coefficients-VLC (variable length code) is analyzed for DRVLD (Dynamically Reconfigurable VLD) circuit. Some of code set are taken from MPEG-2 test sequence with MPEG-2 reference software.

View full abstract

Low-complexity algorithm for designing Adaptive Interpolation Filter has been proposed in the past work. However, estimated filter coefficient was not necessarily optimal especially when the residual signal after motion-compensated prediction was not significant. From this perspective, in order to improve the coding efficiency, we propose an enhanced low-complexity filter designing scheme which controls filter characteristics dependent on the fractional pixel position. Experimental result showed that bit reduction ratio under low bitrate coding improved 5.3 points in the maximum case against our previous work for low-complexity AIF scheme.

View full abstract

We developed a novel intra prediction method using Block-based Pyramid Prediction(BPP). In BPP, prediction is performed by upsampling the reconstructed block of downsampled input. Our proposed method outperforms conventional intra prediction used in MPEG-4 AVC/H.264. The simulation results show up to 30% improvement at very low bitrate coding.

View full abstract

The hybrid coding which is used in standard H.264 is casual prediction. Thus, it only makes use of the correlation between encoded pixels and encoding pixels, while not encoded pixels can not be used. On the other hand, the non-causal technique is submitted as a new method of prediction. This paper compares the two prediction methods and analyzes the results.

View full abstract

We propose a novel video coding method using wavelet super resolution (SR) for high compression ultrahigh video coding. The proposed method encodes reduced-size images and carries out SR reconstruction after decoding procedure. The SR reconstruction uses some parameters which are encoded as auxiliary information. The auxiliary information is determined by comparing a local-decoded SR image with the original image in encoder side.

View full abstract

We proposed a region-based adaptive interpolation filter optimizing the filter coefficients according to the image characteristics to enhance a frame-based separable adaptive interpolation filter. In this paper, we applied a new region-dividing method to our algorithm and confirmed the performance.

View full abstract

For re-transmission of high speed streams that exceeds the transmission capacity of a career, we propose a method to divide various bitrate transport streams, which can achieve highly efficient multiplexing and low power consumption for a receiver. We also evaluated an experimental system.

View full abstract

We have done our experiment for digital CATV signals transmission in 23GHz band to settle the technical standards to transmit digital signals such as terrestrial digital broadcasting using 23GHz band radio systems of CATV broadcasting. This paper describes the results of this experiment.

View full abstract

We have developed a prototype of dual polarized Yagi antenna for the next-generation digital terrestrial broadcasting. This paper shows fundamental characteristics (radiation patterns and cross-polar discrimination (XPD)) of the prototype and the result of field experiments. As a result, superior XPD characteristics were confirmed.

View full abstract

We propose an method of selecting alternative terrestrial broadcaster channel for automobile digital TV tuner. By this method, Alternative Channel can be searched more rapidly when the reception getting worse.

View full abstract

Along with the installation of a large number of relay stations for digital terrestrial broadcasting, co-channel interference occur in part of service area. As a countermeasure against co-channel interference, we have developed an interference canceller using adaptive array technology and have experimental omparisons of a Pre-FFT type canceller and a Post-FFT type one in a real field.

View full abstract

A high-precision network ID estimation method for ISDB-T has been attained for low reception CN ratio. This paper describes its estimation algorithm and results of indoor experiments. This method can be an effective way of specifying unidentified interfering stations of ISDB-T under anomalous propagation.

View full abstract

In order to investigate the TV receiving circumstance, we have observed the wave intensities on this UHF TV frequency band at the Tsubata town since May 2005. As a result, the receiving antennas sometimes measured the broadcasting waves of the Matsue city over the Japan Sea.

View full abstract

Toward the closedown of terrestrial analogue TV broadcast on 24th July 2011, the construction of the terrestrial digital broadcasting relay stations is becoming the final stage, and it is still necessary for its costs to reduce further. As for mini-transmitting system, we are promoting the low-cost type using a gap filler system. We examined the advantage and the disadvantage in which introduction gap filler system as substitution of mini-transmitting system are examined, as result it became an introduction this time.

View full abstract

Toward the closedown of terrestrial analogue TV broadcast on 24th July 2011, the construction of the terrestrial digital broadcasting relay stations is becoming the final stage, and it is still necessary for its costs to reduce further. As for mini-transmitting system, we are promoting the low-cost type using a gap filler system. However, there are many mini-transmitting stations that should keep frequency precision to avoid SFN interference between the other stations. So, we could not adopt conventional gap filler system, because it has only a crystal oscillator or an OCXO oscillator. Therefore we developed a gap filler system to report here, which can input 10MHz signal from an outside standard oscillator.

View full abstract

We developed image processing technology, which improves smoothness of human skin picture and is based on gradation correcting method. It provides compatibility to other gradation correction and lower cost for implementation.

View full abstract

Among a lot of proposals for super resolution image reconstruction, the total variation (TV) regularization method seems to be the most successful approach with sharp edge preservation and no artifact. But, TV regularization method still has a problem. In this Paper ,we propose improved TV regularization method.

View full abstract

Moving pictures in the one-segment broadcasting are highly compressed by using the H.264 international standard for moving picture compression because of the narrow transmission band. As a result of the compression, blocky noise includes in the compressed pictures. We have proposed a method of improving image quality using wevelet transform.

View full abstract

In one of the noise removable methods for blocky noise and mosquito noise of moving pictures compressed by MPEG-II, the total variation regularization method, which is possible to divide blocky noise and mosquito noise into a texture component, was proposed. In this paper, denoising performance is improved.

View full abstract

Among a lot of proposals for super resolution image reconstruction, the learning method seems to be successful approach. But, the learning method still has several problems. In this Paper ,we propose improved learning method.

View full abstract

We previously proposed a gamut mapping method for extended visual cryptography which expands the dynamic range within a certain predefined noise level. This paper performs subjective evaluations to find out that the 55% gain in the dynamic range compared to the noiseless case is achieved when parameters are optimally defined.

View full abstract

In this paper, we propose the technique that subspace method of mammographic masses automated classification.

View full abstract

In this study, there is it for the purpose of the improvement of the rate of detection of the breast cancer and decrease of the false positive (FP) for developing CAD of high performance.

View full abstract

It is demanded the ratio of a heart and lungs with a CT image quantitatively three dimensions by finding the volume of a heart and lungs. Therefore, it can judge a condition precisely than before.

View full abstract

Cell culture for the implementation of efficient and safe, improving the image of the cell, it is necessary to help the image processing and recognition. The improvement of the quality report.

View full abstract

The liver area is extracted by using two input images with different taking a picture time, and marking of the metastatic liver tumor is done from the extracted liver area by the automatic operation.

View full abstract

The RS code is usually decoded using a hard-decision decoding algorithm in HDD. In this study, the soft-decision iterative decoding for the RS code is studied as the sub-optimal decoding in the perpendicular magnetic recording channel.

View full abstract

In 2008, we investigated the PR2ML system for two track perpendicular recording model where the recording code was the constructive intertrack interference (CITI) code which converts the six user bits into a pair of four channel bits. In this paper, we propose a new CITI code with the high coding rate which can improve the performance.

View full abstract