A 1/3-inch format 1M square-pixel 1T-CCD image sensor has been developed. High saturation signal of 600mV, high sensitivity at 1/15s of 110mV/lx, and low smear level of -90dB have been achieved. The device gives high vertical and horizontal resolution, making the device suitable for various applications.

View full abstract

A 1/3-inch 410k-pixel hyper dynamic range CCD image sensor has been developed. Each pixel has four-phase vertical CCD for the purpose to transfer both the normal signal and highlight signal simultaneously. By using these two type of signals, 64-times wider dynamic range are achieved cpmpared with a conventional CCD. The horizontal CCD which adpots parallelogram gate structure can be driven by 2V clock even at 28.8MHz and can also reduce the power consumption.

View full abstract

We have developed a 1/4-inch format 360,000 effective pixel 1T-CCD image sensor, corresponding to DV format. As high quality is expected for DV format picture, improvement of sensitivity by 2.3dB and expansion of dynamic-range by 2.5dB have been achieved in this device, in comparison with conventional technology.

View full abstract

We succeeded in fabricating a 300K pixel atrificial retina chip with the filtering function in both horizontal and vertical directions. This function was realized by providing equivalent two output circuits, and the pixel shrinkage was attained by putting the current inversion circuit out of the pixel.

View full abstract

The back illumination type photosensitive field emitter was fabricated by depositing p-type a-SiC : H/intrinsic a-Si : H photodiode on the back surface of non-gated n-type cone-shaped Si emitters by using plasma CVD methode. The emission current correspnding with illumination intensity was observed and the quantum efficiency of about 0.7 was obtained at λ=680nm. However, the emission current has tendency to be saturated at high illumination level. To expand the dynamic range of the photodetector, the photosensitive field emitter composed of gated field emitter and p-type a-SiC : H/intrinsic a-Si : H photodiode film was demonstrated.

View full abstract

The photocurrent multiplication of Si : H/a-SiC : H staircase photodiodes with graded-gap multiplication region has been observed. The threshold voltage of photocurrent multiplication was changed by the conductionband offset of the graded-gap band structure. As the band offset becomes larger, the threshold voltage of photocurrent multiplication becomes low. But when the conductionband offset is larger than 0.4eV, the quality of the films is deteriorated, and the threshold voltage of photocurrent multiplication increases. It was found that avalanche multiplication of the photodiodes was influenced by the conductionband offset of the graded-gap super lattice and the quality of the films.

View full abstract

This paper computed the detailed motion of the photo-generated charges, during charge storage time, in a CCD photo-diode with a vertical overflow drain. The computed data of current flow, its vector and potential made clear that photo-generated charges are stored in a center of a cell, accompanying charge overflow to a drain, while keeping considerable empty region at edge.

View full abstract

The photocurrent multiplication has been observed in a-Si;H pin photodiode films. The temperature dependence of the photocurrent and current in this photodiode was investigated. It was found that two origins caused the dark current. In a low electric field, the generation at crystal silicon/a-Si interface is dominant, and the band-to-band tunneling breakdown is occurred in an intrinsic a-Si : H layer in a high electric fields. Moreover, It was found that the photocurrent avalanche multiplication is strongly affected with the geminate-recombination phenomena.

View full abstract

This is a study of new image sensor, FEIS (Field Emitter Image Sensor), which consists of a matrix field emitter array, a mesh electrode, and a HARP target. The fundamental beam characteristics of this device were investigated in a vacuum chamber. Experiment confirmed that the effective beam current and the resolution characteristics were largely affected by the mesh electrode voltage.

View full abstract

It is the first time in Japan that uncooled infrared camera is realized, which employs 256×256 uncooled bolometer focal plane array. Vanadium Qxide is selected for the uncooled bolometer sensor material and its element size is 50μm×50μm. It has a NETD of 0.2K for an F/l optics.

View full abstract

For low light level imaging, a proximity focused electron bombardment CCD tube (EB-CCD) was developed. With a CCD of 512 by 512 pixels, a resolution of 360TV line was obtained. At applied voltage of-8kV to the photocathode the gain was 750,and a single photon could be detectable above a noise floor of the CCD.

View full abstract

This paper analyzed charge storage process to increse the amount of signal charge in the very small cell. It was made clear that the small potential gradient, spreading from the p-well into the deep inside of the storage region, disturbs the large charge storage. A cell structure is proposed which bring about large signal charge.

View full abstract

A wave analysis optics simulator TOCCATA-Wave was developed. It can analyze optical characteristics of image sensors, including diffraction and interference phenomena. The focus conditions of micro-lens with different radius and aperture size and smear characteristics of a IT-CCD were studied by this method.

View full abstract

We have analyzed smear in CCD image sensors based on spectral response measurement using test devices with size-reduced apertures, which are formed at peripheral regions of photodiodes. As a result, the dominant generation mechanism seems to be multi-reflection and diffusion when the aperture is located near channel-stop and transfer-gate, respectively.

View full abstract

White defects on CCD determine the lifetime of TV camera on a space station. CCD is sensitive to cosmic ray compared with other semiconductor devices. However, the results of fight experiment and ground irradiation experiment did noto coincide with each other to generate any defects. The camera had been withdrawn from the space station due to the increasing white defects after 5 years' operation. We will report on the possibility of the long-term use of broadcasting equipment on manned on manned space station MIR.

View full abstract

In this paper, we present our results of experiments about the functionality of video compression sensors. We evaluated the functionality of pixel-parallel image sensor and column-parallel compression sensor. The evaluation includes control of the amount of output pixels from pixel-parallel compression sensor and the reconstruction of the image by using the output pixels and flag signal from column-parallel compression sensor.

View full abstract

In this paper, we propose motion adaptive image sensor that aims reducing random noise and getting wide dynamic range. The proposed sensor is able to control the storage time in each pixel, which results in no motion-blur and no saturation. Thus it is expected to have wide dynamic, high temporal resolution in the moving area and high SNR in the static area.

View full abstract

In this paper, we propose a high speed 2-D motion vector detection sensor with the edge detection, double array digital memory, and block matching of the binary edge information are integrated on a focal plane. The processing unit is composed of analog processors with pixels parallel structure and the digital processors with column paralle structure. The first prototype chip is designed under the process of 1.5μm CMOS technology with a fill factor of 17% and with the die size of 7mm×7mm.

View full abstract

This paper presents sub-band coding techniques suitable for both on-chip CFA(color filter array) for single CCD camera and wired logic. Enough compression ratio and image quality are obtained applying SSKF(symmetric short kernel filter) horizontally and vertically to red and blue color signals from square grid CFA, and two dimensional perfect reconstruction filter to a green color signal from quincunx CFA. Also, a comparative investigation of three image compression methods; DPCM, Hadamard transform, and sub-band coding, was made, which allows an output signal from CCD in a color camera to be compressed directly.

View full abstract

The three CCD imaging method provides high quality pictures because of sufficient color sampling points. However, it requires a color separation prism. On the other hand, single chip CCD imaging has an advantage to reduce camera size but insufficient color sampling points. With the aim of realizing a small size and high quality camera, we have been studying imaging method for NTSC using single chip HDTV CCD. The arrangement of a color filter array is a key issue to improve color image quality. In this paper, we propose a new color filter array, and we examine driving CCD with this filter array and color separation method for both HDTV and NTSC. The proposed filter array has quincunx sampling structure for difference color signal, resulting in improved color resolution.

View full abstract

A color-offset sub-sampling image acqusition method to obtain a super-high definition image is presented. The acquisition is performed with four imagers, two for green, one for red and the remained one for blue, attached on a color-separating prism, forming a "checker" sampling pattern. Optical LPFs are inserted both in vertical and horizontal direction in red and blue channels to prevent the generation of alias components in chrominance signal. This arises another problem, ie the incomplete cancellation of diagonal alias components in luminance signal. The problem is solved by inserting the optical LPFs with the same transfer function in diagonal direction of green channels as that of the LPFs for red and blue. Simulation results are presented to confirm the validity of this method.

View full abstract

To cover the Nagano Winter Olympic Games in HDTV, NHK is now developing several newly-designed HD equipment. As a result of the development, NHK finished ultra compact camera for HDTV. Newly designed high quality CCD which has RGB color stripe filter on it's surface could make it possible to design the camera small enough. This paper describes the specification of the camera and how this camera was designed.

View full abstract

SrGa_2S_4 : Ce films for blue-emitting thin-film electroluminescent devices have been prepared by multi-source deposition. Sr, Ga_2S_3 and CeCl_3 were used as source materials. It was found that the annealing in H_2S atmosphere was considerably effective for the improvement of the crystallinity and luminescent properties of the SrGa_2S_4 : Ce thin films. The SrGa_2S_4 : Ce film that was annealed at best anneal condition of 800℃ for 1〜2 hours in H_2S showed strong blue PL and CL with CIE coordinate of (0.12,0,10).

View full abstract

Surface morphology of TPD thin films has been observed by Atomic Force Microscopy (AFM). It was found that the morphology depends on significantiy the deposition rate of the TPD films, and there exist the optimum region of the deposition rate to obtain flat surface. Furthermore, it was found that maximum luminance has a close relationship to the surface morphology of TPD thin films.

View full abstract

Motion blur is observed when motion pictures are displayed on a TFT-LCD. However, it is not clear how mechanism of viewing arises such picture quality impairment. As a step to clear the mechanism, we investigated the observed frequency amplitude response on various moving velocities. As a result, the minimum points of the obserbed response almost corresponded with estimated values.

View full abstract

Trajections characteristics were calculated for the electrons emitted from the top of cone shaped Field Emitter. It is clarified that the optical cross talk is not occurred when the distance from substrate to anode is 200 micro miters, .

View full abstract

The dependence of the suppression of both the charging-up on phosphor surface and the degradation of luminescent intensity under by low energy electron excitation lower than 500eV on the fraction of In[OCH(CH_3)_2]_3 to ZnS : Ag, Cl phosphor in the sol-gel coating has been investigated to obtain the optimum coating condition. It was found that about 3.7wt% of In[OCH(CH_3)_2]_3 was most effective.

View full abstract

A simplified model of TFT-LCD equivalent circuit for fast and accurate simulation was studied. This model consists of a pixel circuit, delay circuits and a common electrode circuit. The common electrode was divided into smaller areas near the common voltage input points to decrese the number of resistance elements in the common electrode circuit. Simulation using this model gave good agreement corresponding to the measured characteristics of TFT-LCD.

View full abstract

A horizontal deflection circuit using a 1700V VCBO transistor has been studies to reduce CRT display power consumption. The transistor power can be reduced by 4W at an 80kHz horizontal scanning frequency by reducing a deflection yoke current. The transistor power can be related to on-resistance of the transistor.

View full abstract

We have proposed computer generated rainbow hologram (CGRH), which can be reconstructed with white light. In the previous paper, 3-D objects are assumed as collections of self-illuminated points. In this paper, we employ stereoscopic approach to generate full color CGRH of polygon model 3-D objects.

View full abstract

It is important to execute hologram calculation at high speed to realize the real time electro-holography. In other words, a higher-speed computer and a study of calculation algorithm becomes an important subject. This time, we realized to speed-up the calculation of Fresnel-Hologram by networeking technology.

View full abstract

Recently, liquid crystal display(LCD) have been used for reconstruction of holographic images. In holographic display on the LCD, reconstructed true images are often disturbed by transmission beams, conjugate images, and harmonic images, because the LCD consists of picture elements and has not enough resolution to display interference fringes. We studied about spatial filtering as a means of suppressing such disturbance of true images and confirmed its effectiveness in an experiment. We will be able to enlarge the display area by switching this spatial filter time-sequentially.

View full abstract

A new autosteroscopic display is described that uses a LCD combined with a multifocus lens. The properties of the multifocus lens allow the independent establishment of viewing zones for arbitrarily chosen sets of the LCD's pixels. Unlike autosteroscopic display based on lenticular this new display does not suffer from reverse 3D effects and maintains 3D images despite movement of viewers.

View full abstract

This paper proposes the new human face pick up method using a stereo image pair. Proposed method is effective for a human face pick up in free distance and face size measurement.

View full abstract

This paper proposes the occlusion compensation method by 3 camera array. This method uses 2 stereo image pair and get two result disparity map with occlusion. This paper shows unified disparity map is made by each stereo pair results.

View full abstract

Silhouette backprojection is a mapping from an epipolar plane to a polar coordinate parametric slice plane. A cosine curve on the epipolar plane corresponds to a single point on the parametric plane. If we resample the silhouette intensities along the curve in a constant interval, then the topological information on the epipolar plane is lost. We use laplacian backprojection as well as the traditional silhouette backprojection to utilize the coherence on the epipolar plane and to sharpen the silhouette slice image.

View full abstract

We propose a new method of "space coding" to transmit the 3-D space as it is. A multi-dimensional DCT is applied to the 4-D data space in which the ray data is filled densely. Experimental results show the relation between the resolution of 4-D data space and the coding performance.

View full abstract

In this paper we proposed an algorithm for blending between two given key polygonal shapes. Generally speaking, 2-D shape blending requires two mainsteps : one is vertices correspondence and the other is shape interpolation. Some approaches use matrix interpolation which sometimes causes shrinkages and kinks because a transformation common to all vertices may not exist. Other methods try to interpolate the length of edges and the angles formed by two adjacent edges[1]. This article makes an attempt based on center of mass to generate in-betweens more fast.

View full abstract

[in Japanese]

View full abstract

We present a method to synthesize an arbitrary view using images from 5 cameras. The method consists of, first, transforming the depth map into that of a new view point and then, mapping texture by exploiting images from inspection cameras. This paper focuses on detection of uncovered area and texture mapping technique for the area.

View full abstract

This paper present a new detection method of distorted area in motion picture film using Progressive Grobal Motion Megmentation algorithm. PGMS method search for motion parameters and its following areas. When some areas were distorted, these area could not find correspond area in previous or next frame. Using this proparty, distorted areas can be detected as the remainings of PGMS process. We shows the results of the computer simulation using images which artificially distorted with blotches, and shows the capability of PGMS detection method.

View full abstract

We propose a new method of measuring textures for video components of a Desk Top Program Production. This method can measure the textures without specular reflections on the objects using a polarized light, and one who doesn't know a material of objects and position of lights can measure the textures.

View full abstract

IEEE1394 is expected as a new interface to connect degital products such as DVD, DVCR and IRD. Most part of IEEE1394 is supposed to be implemented by software. We estimated how much the software would influence on the MPU performance.

View full abstract

We have developed a digital signal processing LSI for video capture use. In addition to conventional video camera process, video signal process (YC separation, sync signals separation and sub-carrier PLL) for external video input are contained in the LSI in order to decode NTSC/PAL signal or output signal of VTR to digital YUV signals.

View full abstract

A new VCR for digital satellite broadcast system based on the D-VHS format has been developed. It can record digital bit stream data from the receiver via SDAV Bus (Simplified Digital A/V Bus) using the D-VHS tape in addition to the current TV recording.

View full abstract