In the present study we conducted an experiment in which young and old subjects reach for an LED target presented prior to the movement in a darkness room and the position error was measured. The results show that age-related changes of position accuracy in downward reaching tasks were confirmed.

View full abstract

We conducted experiments to evaluate our prototyped remote for digital broadcasting receivers using the semantic differential method. Two major components - "General evaluation" and "Design" are extracted by the result of the principal component analysis. We find that there are significant differences in the each component score between the prototyped remote and the conventional remote that has many buttons. The result shows the effectiveness of our design processes to develop an easy-to-use remote.

View full abstract

This paper presents some experimental results on understanding Japanese language shown in TV scenes. We chose 38 students. Two kinds of captions are used; Test is done by the people both from the regions where kanji is used and people from the regions where kanji is not used.

View full abstract

This paper presents a preliminary study of an effect of hiragana (Japanese syllabary) above kanji in TV captions on understanding of TV scenes by foreign people who know little Japanese. Comprehension test results together with some ideas to make Japanese TV scenes more understandable for foreign people are given.

View full abstract

We are constructing the educational system for learning swimming. This system works not only at the pools but also at home using 3-dimensional positional information system with computer graphics. The learners can see the swimming model visually by CG in ever angle and get effective information for learning swimming.

View full abstract

Toward an ultra-small high-performance microphone for the next generation, we are studying a silicon microphone. In recent research, we developed a frequency response tuning technique by optimally designing the structure of a microphone capsule. By using this technique high-frequency response can be tuned up without any drop in sensitivity.

View full abstract

We have designed a CMOS image sensor especially designed for in vivo, on-chip neural imaging. We also developed packaging technique of an in vivo imaging module. In vivo fluorescence imaging of a mouse's hippocampus was successfully performed

View full abstract

This newly developed HDTV I.I. camera is the first HDTV camera with built-in ion noise reduction circuit. It has capability to film with low noise at night, and is also highly sensitive at a near-infrared range through the adoption of GaAs-I.I device that has the sensitivity to the near-infrared for the camera's red channel.

View full abstract

In this paper, a CMOS image sensor for confocal microscopes is presented. To increase the sensitivity, the proposed system uses a multiple beam light source and the image sensor can detect selectivity and concurrently the peak intensities of multiple light beams. A low noise signal readout technique suitable for the beam light detection is also presented.

View full abstract

To improve picture quality, we doubled the sensitivity to red light of 15-μm-thick New-Super-HARP film by doping more tellurium into the film. We also found a way to suppress highlight defects by controlling the temperature of the film.

View full abstract

A novel filter-less fluorescence detection sensor was successfully fabricated using CMOS integrated circuit technology. Previously filter-less fluorescence detection sensor was proposed and fabricated, and it was able to sense fluorescence intensity. CMOS image sensor technique was applied for 1-pixel of filter-less fluorescence detection sensor. Consequently, it was confirmed that general CMOS image sensor technique can be applied to the novel filter-less fluorescence detection sensor.

View full abstract

We have fabricated CMOS image sensor overlaid with an organic photoconductive film. The output signal voltage based on signal charges generated by the organic film has been observed.

View full abstract

A 300,000 pixels ultrahigh-speed image sensor at up to 1,000,000 fps was developed. To achieve this sensor, we use the battable method with plural CCD chips. This method requires a special battable layout and a highly precise dicing technique, highly precise die bonding technique. Using this sensor, we can achieve to get the high resolution ultrahigh-speed image which is applicable to the high definition.

View full abstract

We have developed an active matrix HEED (High-efficiency Electron Emission Device) cold cathode to be applicable to the image scanning for a HARP super high sensitivity image sensor. The driver circuits that allow the pixel sequential scanning and the controller circuits were incorporated in an LSI onto which the HEED was fabricated. A prototype HARP image sensor has been built using the active matrix HEED and we have confirmed that the image pick-up was working property.

View full abstract

A HARP image sensor with an active-matrix HEED field emitter array was fabricated and tested. The experimental results show that prototype sensor has both enough resolution for its pixel size and high sensitivity due to the 15 μm-thickness HARP target, thus demonstrating its potential as a next-generation high-sensitivity practical image sensor.

View full abstract

A 256×192 pixel field emitter array (FEA) image sensor with a 25-μm-thick high-gain avalanche rushing amorphous photoconductor (HARP) target was fabricated and tested as a step toward the development of ultrahigh-sensitivity compact image sensors for high-definition TV cameras. Our experimental results revealed that the prototype sensor had both ultrahigh-sensitivity and enough resolution corresponding to the number of pixels, thus demonstrating its potential as a practical image sensor.

View full abstract

We have developed an ultra-high-sensitivity HDTV Hand-held camera that can shoot beautiful picture under low light condition. To improve the picture quality, we have reduced noise especially in the dark part of the shot and color afterimages with the lower electric gain and a uniform electrical current for the red, green, and blue color signals.

View full abstract

An ultra-high-sensitivity HARP HDTV camera can shoot beautiful night time scenes just like images seen with the naked eye. Based on data obtained through the shooting experiment, we're looking for new ways of shooting dramas with the camera and attempt to improve the camera's picture quality.

View full abstract

A column-parallel double-integration-type A/D converter(ADC) for CMOS image sensors is presented. This ADC simultaneously reduces the random noise of readout circuitry and the quantization noise of the ADC. Simulation results show that the resolution of the ADC can be improved by a factor of 5 bits using 1024 times integration.

View full abstract

We demonstrated a staircase avalanche photodiode (APD) film with low operation voltage. The avalanche multiplication on the staircase APD give an extra energy from the band offset. In the result, the low operation voltage type a-Si:HAPD was fabricated. The maximum multiplication gain ratio is 2, the applied voltage is 7V.

View full abstract

Microangiography with spatial resolution in the micrometer range was carried out to depict coronary vascular response in a rat heart under Langendorff perfusion using a high-speed imaging system and a synchrotron radiation source. The imaging system makes possible the direct evaluation of vasodilatation in small coronary arteries with less than 50 μm diameter.

View full abstract

This paper deals with an image-processing method for removing image flickers. The method is based on a logarithmic luminance-correction algorithm mimicking the luminance accommodation of the human vision. We perform experimental simulations using moving image sequences containing artificial image flickers as test image sequences. Their results demonstrate that the method can remove image flickers very efficiently.

View full abstract

This paper presents two different processes of nonlinear diffusion of logarithmic intensity for separating the illumination and the albedo from an observed color image. The one process is the two-dimensional nonlinear diffusion process, and the other process is the one-dimensional nonlinear diffusion process that diffuses the intensity only along the edge-running direction. The one-dimensional nonlinear diffusion process outperforms the two-dimensional nonlinear diffusion process, and separates the illumination and the albedo more successfully.

View full abstract

This paper presents a new scheme to extend our previously presented scalar coupled nonlinear diffusion sharpening method to color image sharpening. The scheme separately processes the norm of a color vector and its unit directional vector with the scalar/vector coupled nonlinear diffusion method, and outperforms the existing methods.

View full abstract

The method for evaluating the degradation degree of out-of-focus images quantitatively using the spectral characteristics of the output from the Measurement Differencing Filter(MDF) is proposed. The proposed evalution value is almost linear to the radius of out-of-focus PSF.

View full abstract

This paper improves the correspondence accuracy of pixels between adjacent frames in colorization for motion pictures by verifying variance of intensity in block matching stage. By considering the variance, texture pattern can be tracked correctly between adjacent frames.

View full abstract

This paper proposes an interactive and user-friendly visualization of the collected data, designing a new standard for local information database.

View full abstract

We propose a hierarchical secret sharing method for image information using JPEG2000 code stream syntax. It can control the quality of the decoded images by the number of the combined share files. We verify the performance of the proposed method and confirm its effectiveness.

View full abstract

The proof of the contents of the talk by the communication image that applied the digital watermark has been studied. The digital image to which a watermark is applied is obtained via the analog terminal that is the external output of communication system. In such a system, analog conversion of the image where the watermark was embedded will be carried out, generally, at this time, a digital watermark disappears and data proof is no longer performed. In this paper, watermarking based on KL transform is reported, and the decoding results from a stego-image transformed by both H.323 transmission and analog conversion are shown.

View full abstract

In this paper, the authors propose an MP3 scrambling scheme which alters scalefactor values. The conventional method to alter quantized spectral coefficients consumes much more computational resources. On the other hand, the proposed scheme can be realized with very low complexity. Moreover, it enables control of audio degradation level by altering with consideration of human perception characteristics.

View full abstract

Images generated by VAC-VSSS (Void-and-Cluster Visual Secret Sharing Scheme), which is a novel VSSS method proposed by authors, are investigated in order to study image characteristic of share images and visibility of a secret image. Compared with a random dithering VSSS and Bayer's matrices VSSS, the proposed method generates higher quality share images in terms of a white noise and artificial dot patterns. Besides, a newly defined metric for visibility of a secret image shows VAC-VSSS provides almost the same visibility of an ideal case.

View full abstract

The accuracy of image search engines is not good compared with the text search engines. The reason why the accuracy is not good is that they use only text information when they search Images. In this paper, I discuss the search system for Images based on Image features.

View full abstract

This paper describes a method of image generation based on transformation integrating multiple differently focused images. First, we combine acquired images and spatial frequency of objects by using a 3-D blurring transformation. Then, we reconstruct an all-focused image from the acquired images based on the spatial frequency analysis.

View full abstract

In this paper, we propose a method of objects tracking and distance estimation under high frame rate such as 1000fps. The real time processing can be done by using not image signal but 1 bit motion-flag signal.

View full abstract