We previously proposed a lossy image coding scheme based on recursive-type intra prediction which can use both local and non-local reference pels at the same time. In this paper, we extend the prediction technique to make use of multiple non-local pels collected by template matching. To reduce side information on prediction coefficients, a weighted average of the non-local pels is used as the reference one.

抄録全体を表示

The 4K/8K BS trial broadcasting will start from August 2016. The BS has so wide bandwidth that it can transmit 4K HEVC with 44Mbps. A cable network restricts the bandwidth to 29Mbps/38Mbps by the 64/256QAM specification. In this manuscript, a BS 4K transcoding performance is reported for the feasibility study of a retransmission over the cable network.

抄録全体を表示

In recent years, video image resolution is greatly enhanced, and it is necessary to satisfy the requirement of high speed processing. To accelerate the encoding process, we propose a fast PU mode selection method that utilizes the variance of CU and enables to reduce encoding time by more than 30%.

抄録全体を表示

This paper shows study on an application of Augmented TV with ExPixel. Using ExPixel, we solved a problem of viewing conditions for Augmented TV, and we confirmed the effectiveness through an experiment.

抄録全体を表示

We developed the m ixed r eality system that visualize the pianist's finger motion and her hear tbeat while playing the piano. T he viewer is able to watch her fingering and her emotional ups and downs in free - viewpoint with a HMD.

抄録全体を表示

It becomes often difficult to grasp controllers in the real space through the VR-HMD. However, it would lose interest to go back and forth between the real world and a virtual world, and damage immersive presence. We propose a simple method for acquiring the information of the real space through the VR-HMD.

抄録全体を表示

C hemical light (psyllium) color transition at a real l ive stage is one of the important direction technique to heap up audiences . We have developed an automatic system of psyllium color transition for the l ive karaoke system . A simple evaluation result shows that our system effectively realizes psyllium color transition

抄録全体を表示

Our technology allows encoders on the distributor side to change transmission rates precisely according to the reception situation. This can reduce sudden variations in the reproduced image quality.

抄録全体を表示

IPDC using an area broadcast was performed in the golf course as service for galleries of THE CROWNS Golf.

抄録全体を表示

We transmitted 4K-HDR(HLG Method)Live Video by an Area Limited Broadcast Wave,and We received 4K-HDR Live Video signal by television receiver. And We checked 4K-HDR HEVC image quality.

抄録全体を表示

High Dynamic Range (HDR) Video is expected to be a high picture quality technology, but HDR Video is concerned about increase in transmission rate compared with SDR (Standard Dynamic Range) Video. Here we report analyze the results of increment transmission rate in 4K HDR Video for cable network.

抄録全体を表示

We are developing a new video codec system toward 8K broadcasting with 120Hz frame frequency video signal. In this paper, we report the results of a study on video coding parameters. By choosing suitable parameters, the encoding time can be reduced by 17% without large loss in the coding efficiency.

抄録全体を表示

We previously proposed a block - adaptive post filter ing method for reconstructed image s corrupted by DCT - based coding. In this method, quantization constraint of DCT coefficients was exploited . In this p aper, we optimize the filter coefficients in DCT domain to reflect the quantization constraint for filter design

抄録全体を表示

In motion compensated prediction using multiple reference frames, arithmetic mean on a block- by-block basis is a common way to generate predicted values . In this paper, we introduce a kind of robust estimation technique to improve the prediction accuracy.

抄録全体を表示

We previously proposed a block adaptative prediction method for RGB 4:4:4 color video coding. In this paper, we combine the method with a pixel-wise motion estimation technique which exploits previously decoded color signals to improve accuracy of motion compensation performed in the prediction.

抄録全体を表示

We have developed Laser Backlighting Liquid Crystal Display Television using the laser diode for one of the light sources. The 4K laser backlight liquid crystal television has been achieved the 80%, or higher, coverage with respect to the gamut defined in ITU-R Recommendation BT.2020 for the 4K pictures in studio production.

抄録全体を表示

We have developed a RGB Laser Backlight that uses three laser of red, green, and blue. The color gamut of the developed display nearly corresponds to that of BT.2020 using the laser for the light source.

抄録全体を表示

Flexible displays using high-mobility oxide TFTs and inverted OLEDs were fabricated and evaluated. Moving images with a writing time of about 3.8 s were reproduced on the fabricated displays. This result indicates that these devices are suitable for sheet-type displays in an ultra high-definition broadcasting such as 8K Super Hi-Vision.

抄録全体を表示

We present design of optical beam steering devices using electro-optic (EO) polymer arrayed waveguides. EO polymer phase shifters have been numerically investigated and applied to 16 channels optical phased array devices. We confirmed basic characteristics of optical beam steering devices by using far-field pattern analysis.

抄録全体を表示

This report presents a method of receiving TMCC on ISDB-T broadcasting signals. TMCC signals contain the emergency wake-up trigger and the efficient reception would be useful for a prompt disaster alert. For decoding multiple TMCC subcarriers, the matrix representation of the low intermediate frequency receiver is proposed.

抄録全体を表示

In this paper, carrier choice technique in cooperative reception is proposed. In the evaluation of carrier choice each the location probability in the area that we assumed can be also improved by the proposed carrier choice technique in the Full-segment service.

抄録全体を表示

This paper describes a high dynamic range measurement method of a delay profile.

抄録全体を表示

W hile audio signa l is transferred via IP network, delay time of received audio is varied and fluctuated by IP Codec and network equipment. We have successfully developed a new synchronizing unit which lets those delay time constant by inserting unique signal into transmitt ing audio band. This unit can make it possible to utilize IP networks for transmitting audio signal from studio to more than two transmitters which compose SFN (single Frequency Network) of AM and FM Radio broadcasting.

抄録全体を表示

We propose a new segmentation process of answer section for the automatic grading system. The segmentation process is based on marginal distributions of black pixels on an answer sheet image. As a result of experiments, the correct segmentation rate of answer sectio n is 97.2% for 71 handwritten answer sheets

抄録全体を表示

We present an iterative method for refining weave patterns in color tape weaving images by minimizing errors between input colors and weaved images smoothed with a cross bilateral filter. Color palettes are adaptively derived from each input image and they are set to the color of tapes. A procedure for restricting floating tape length is incorporated into this weaving method for ensuring mechanical strength of weaved tapes.

抄録全体を表示

We propose a method for color image enhancement based on the Neugebauer model, which expresses an arbitrary color as a weighted combination of primary colors. The proposed method changes the weights to enhance colors. Experimental results show that the proposed method can enhance the colors in images naturally without gamut problem.

抄録全体を表示

Recently, many pap er documents used in the hospital have been computerized thanks to spreading of Hospital Information Systems (HIS). In the se documents, some line drawings called “Body D iagram ” are som e- times used to indicate organs , bones and so on . The authors discuss a b ody diagram recognition method for me d- ical document retrieval. In this paper, we proposed a method to detect misrecognition cases for the conventional method.

抄録全体を表示

Face detection is one of the most important techniques for video content analysis. The joint cascade method[2] enables fast and accurate face detection by combining face alignment with detection. We train and evaluate a classifier of the joint cascade to investigate an ability for TV video analysis. In experiments, we used 20000 face images to train the classifier and evaluated its effectiveness.

抄録全体を表示

Toward realization of advanced Hybridcast services, we are studying technologies to synchronize internet delivered content with broadcast programs. We confirmed our proposed method is enough capable by validation experiment as shown in the synchronized presentation of real-time tracking data delivered over the internet with broadcast live footage.

抄録全体を表示

We propose three different methods for timestamp conversion that achieve precise synchronization of broadcast and on-demand video streams on Hybridcast. This paper describes design, prototype and comparative evaluation of these methods.

抄録全体を表示

We have proposed a concept that connects TV contents and everyday life activities. In order to make the proposed concept more specific, we develop a prototype system that provides TV contents and topics in accordance with user’s behavior or situation. This paper describes a system model of our proposed concept.

抄録全体を表示

TBS Television has developed the Hybridcast ap plication which leads the TV viewer to a real action.

抄録全体を表示

We have developed an experimental application program for Android TV, which can make mult iple video streams synchronized . Experimental results showed that e ach streaming delay s can be adjusted and viewers can see the content streams simultaneously via the internet. .

抄録全体を表示

In recent years, various new services are proposed to improve the convenience of e - comics. In order to commercialize such services, there is need for the technique of autom atically metadata extraction from comic images. In this paper we studied extraction of character, text, b alloon, and panel using Faster R - CNN.

抄録全体を表示

In the field of computer vision, haze lead to degrade captured image quality, obtained image will lose some useful information. For autonomous cars in hazy environment, it cannot correctly recognize obstacles such as cars or pedestrian s. We proposed a fast dehazing method to solve this problem, our method have achieved some amazing results.

抄録全体を表示

Technologies to segment videos into “scene” units consisting of the shot sequence captured in a same place have been useful for program video retrieval and video summarization. We proposed scene segmentation methods based on the change of color formation produced from color histogram , and the change of block-image-kinds formation considering not only color but also textural properties. However, these methods have problems that slight color noises and drastic changes of camera angle could cause false positives. In this paper, we report on our new proposal method of scene segmentation adopting the color formation based on “superpixel” as an unit of image segmentation to solve problems in our conventional methods.

抄録全体を表示

The spatial property of video signals is different from their temporal property. If the 3-D DFT is applied to video signals, mainly phases of specific DFT coefficients will be affected by apparent motions of moving objects. On the other hand, the 3-D DCT is equivalent to the following transform, that is to say, video signals are first converted into even-symmetric signals about the origin, and then the symmetric signals undergo 3-D frequency analysis. Object’s motion produces many significant 3-D DCT coefficients. Obviously, the 3-D DFT is superior to the 3-D DCT in energy compaction. This paper shows the superiority of the 3-D DFT over the 3-D DCT, through theoretical analysis and experimental simulations.

抄録全体を表示

This Device makes it possible to play video files taken by consumer video camera .

抄録全体を表示

It's effective in realization of a plain broadcast to indicate the track of flight objects like a ball, a javelin etc., in sports programs.. So, this time, we developed the javelin throw tracking system. It’s possible to display the curvature and the angle of the javelin throw, only to crick several points of javelin in the TV image.

抄録全体を表示

Construction of the Tully transmission system using the inter communication system in the TVstation. We can perform Tully transmission with a smartphone even if we don't use a terminal of the inter communication system.

抄録全体を表示