-
Jun GOTO, Kiminobu MAKINO, Yuka TAKEI, Taro MIYAZAKI, Hideki SUMI ...
Session ID: 13B-7
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We developed a system which can automatically generate news manuscripts for supporting news production. The system utilizes news archives in broadcasting company and sensor data distributed from public organization for generation.
View full abstract
-
Pages
13C-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Kazuhisa Kuriyama
Session ID: 13C-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We can perform the effective electric wave measurement by using a drone.
View full abstract
-
Hiroaki SUDO, Kazuhiro KOSAKA, Hideki KANEMOTO, Noriyuki GEJOH, Ta ...
Session ID: 13C-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
Time Division Duplex (TDD)-based bidirectional digital field pick-up unit (FPU) system is now being studied. Hence this new digital FPU system shares its frequency band with other wireless systems, digital FPU system should detect and avoid the interference from other systems. We studied the frequency sensing scheme using received power within preamble signals. From the results of indoor experiment, it was confirmed that the proposed scheme achieved an excellent performance of interference detection.
View full abstract
-
Satoshi Takahashi
Session ID: 13C-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In general, receiving several radiowave signals interferes each other at a receiver. Under certain conditions, however, a signal having the largest magnitude controls the output at the demodulator in the receiver. The report analyses the capture effect of two digital modulation signals in terms of Rayleigh fading power fluctuation with the characteristic function method.
View full abstract
-
Pages
14B-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Naoya HIROTA, Kazuki NISHI
Session ID: 14B-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We have developed an original system for measuring shutter vibrations from the camera image. In this study, we focus on the Phase-Only Correlation method and present the modification for realizing a subpixel accuracy Some experimental results show the validity of the proposed method.
View full abstract
-
Niai YANO, Seiya UMEDA, Hiroshi WATANABE, Tomohiro IKAI, Takeshi ...
Session ID: 14B-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We show that in A+, ScSR and SRCNN, which are the state-of-the-art Super-Resolution technologies, the level of image quality improvement varies according to the “directionality” within the patch of the input image. Based on this observation, we propose a method to further improve quality of the image employing the “directionality”.
View full abstract
-
Keigo Kano, Tomio Goto, Satoshi Hirano
Session ID: 14B-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We propose a facial image processing method utilizing super-resolution to obtain beautiful facial images. To improve image quality of super-resolution processing for facial images, we will divide skin and non-skin parts in facial images. Then, super-resolution processing and facial correction processing are performed on non-skin parts and skin parts, respectively.
View full abstract
-
Tomoyuki MATSUMOTO, Satoshi ITO
Session ID: 14B-4
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
Compressed sensing MRI can reduce the data acquisition time. ADMM is becoming popular algorithm for solving the L1-L2 minimization problem. we applied the multi-scale eFREBAS transform based CS to ADMM for 3D image reconstruction. Experiment showed that ADMM can provide higher signal-to-noise ratio than other algorithms.
View full abstract
-
Takehito MIYAZAWA, Satoshi SUZUKI, Kazuya HAYASE, Tokinobu MITASAK ...
Session ID: 14B-5
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In order to reduce inspection cost in photovoltaic power plants, a method for automatically detecting a panel failure is expected. In this research, we focused on detecting defective substrings, and investigated a method using drone-based aerial thermography.
View full abstract
-
Masaki KITAHARA, Xiaojun WU, Atsushi Shimizu
Session ID: 14B-6
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Pages
14C-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Yuya NOBE, Akira NAKAMURA, Kohji ITOH, Makoto ITAMI
Session ID: 14C-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In the satellite transmission, it is difficult to apply OFDM scheme because of the high peak power. MC-CDMC
(Multi Carrier – Code Division Multiplexing Code) scheme could reduce it and control out-of-band emission. In this paper, MCCDMC scheme in the satellite channel is evaluated by computer simulations.
View full abstract
-
Akira NAKAMURA, Takuya FUSE, Hiroaki Otsubo, Makoto ITAMI
Session ID: 14C-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In the next-generation DTTV(Digital Terrestrial Television Broadcasting), mobile reception is considered. In this paper, reception characteristics using time interleavers under mobile reception are evaluated by computer simulations. As the results of computer simulations, it is possible to improve reception characteristics.
View full abstract
-
Hiroaki MIYASAKA, Tomoaki TAKEUCHI, Madoka NAKAMURA, Kenichi TSUCH ...
Session ID: 14C-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
Toward the advancement of broadcasting services, the authors are conducting research on the next generation terrestrial broadcasting. The proposed specification is incorporated the new signal structure and the latest technologies in order to get spectrum utilization efficiency higher than ISDB-T. Furthermore, the proposed specification has been designed based on segmented OFDM to transmit the signal for both a fixed reception service and a mobile reception service. This system is able to provide mobile reception service by partial reception such as one-segment service of ISDB-T. This paper describes the evaluation results of partial reception performance of the proposed specification.
View full abstract
-
Yasutaka MATSUO, Kazuhisa IGUCHI, Kikufumi KANDA
Session ID: 14C-4
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We develop a band-limitation equipment to improve coding efficiency for advanced digital terrestrial 8K UHDTV broadcasting system. The band-limitation is processed in prior to 8K codec to improve image quality of encoded 8K video with high compression. This equipment has high-speed interfaces considering connectivity to 8K recorder and 8K codec such as MPO (SNAP12) and U-SDI (ARIB STD-B58). In addition, filter parameters of the band-limitation are set by control PC. Coding efficiency of critical 8K UHDTV sequence is improved by using the equipment.
View full abstract
-
Ken NAKAMURA, Daisuke KOBAYASHI, Yasuyuki NAKAJIMA, Takayuki ONISH ...
Session ID: 14C-5
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
High Image Quality Real-time 8K HEVC encoder system has developed. This system encodes 4 slices of 8K
59.94p images in parallel by four encoder LSIs and achieves high image quality by sharing coding data and self-monitoring of image quality.
View full abstract
-
Keisuke HASEGAWA, Kenichi NOGUCHI, Toshiharu MORIZUMI, Junichi NAK ...
Session ID: 14C-6
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We devised CLEAR (Conditionally Lossless Encoding under Allowed Rates) method to transmit high-quality audio signal with low-delay under limited network capacity, and a prototype of low-delay audio codec appliance was developed. This paper presents the result of sound quality evaluation of the prototype codec on a commercial line.
View full abstract
-
Pages
15B-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Makoto TAKAMATSU, Ayaka NAKAMURA, Eizaburo IWATA, Seiji ISHIHARA, ...
Session ID: 15B-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
A pseudo stereoscopic vision system using a piece of human picture is discussed. Faces from a group photo are detected using a face recognition system, and the distance from camera lens (depth) of each person is estimated using each face size. Each person in the picture is clipped out by super pixels and k-means classification. To generate our pseudo stereoscopic vision, we create two images for right-eye and left-eye as each region of the person is moved according to its depth, and convert to an anaglyph image.
View full abstract
-
Hirokazu KAKINUMA, Hiromu MIYASHITA, Yoshihide TONOMURA, Hidenobu NAGA ...
Session ID: 15B-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper, we propose a new real-time image segmentation framework using machine learning techniques. Our framework maps input feature vectors to suitable new feature vectors in neural network. The experimental results show that the proposed framework improves segmentation accuracy and its performance is 60 fps for 4K video.
View full abstract
-
Yoko ISHII, Tetsuro TOKUNAGA, Yoshinobu TONOMURA, Kota HIDAKA
Session ID: 15B-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper proposes that a new real time Camera-LiDAR fusion system for tracking of people. Our system provides object labels and the 3D position of people using the deep neural network model called YOLO and LiDAR. By using our system, pseudo-3D-images are recognized sifting back and forward on “Kirari!” stage.
View full abstract
-
Kazuma SUZUKI, Keita USHIDA
Session ID: 15B-4
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
Long-time works on PC will get users tired. The authors propose a system which encourages the users to stretch. In the prototype system the users operate their PCs with stretching and they will get rid of tiredness while lightweight PC works.
View full abstract
-
Jianfeng XU, Kazuyuki TASAKA, Hiromasa YANAGIHARA
Session ID: 15B-5
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
With the development of human pose estimation technology from an image/video, 2D skeleton pose becomes ubiquitously available. Different from the traditional 3D poses, it is more challenging to understand 2D poses such as action recognition because joint positions in 2D poses change a lot when the camera changes its capturing angle. Thus, it is essential to generate a front view pose from a non-front view one to avoid the camera angle problem. In this paper, we propose a novel algorithm for front view generation of 2D poses using homography transformation in an adaptive window. Our experiments show that the algorithm can greatly reduce the errors when computing the parameters of homography transformation.
View full abstract
-
Sou UCHIDA, Yusuke TAKEZAWA, Eizaburo IWATA, Seiji ISHIHARA, Makot ...
Session ID: 15B-6
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
A method of sticky note annotation for three-dimensional objects using image local feature is proposed. Threedimensional objects are browsing using a web camera, and we can perform our sticky note annotation to the three-dimensional objects using key-points of local feature on the frame images of the browsing video. Our method is simulated, and our sticky notes are tracked even if the objects move.
View full abstract
-
Shuhei TARASHIMA, Takashi HOSONO, Yukito WATANABE, Takayuki KUROZU ...
Session ID: 15B-7
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper we propose a novel object verification method based on mining local visual patterns lying in a image pair. Our method is robust to object deformation because (i) any number of local visual patterns can be mined independently of their locations and (ii) each pattern is extracted via clustering of the correspondence graph that embeds geometric similarities of keypoint quadruplets. We validate our approach in an experiment of detecting deformed planar objects in a noisy image collection.
View full abstract
-
Pages
15C-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Kei KAWAMURA, Akio KIMURA, Hideomi TAKIZAWA, Eiji SAWAZAKI, Minoru ...
Session ID: 15C-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
4K content distribution scheme over HFC cable network by 4K Hybridcast browser is proposed. All HC contents
is transmitted by FLUTE, UDP/IP, and ULE/TS capsulation. Both higher bitrate and more stable playback are confirmed.
View full abstract
-
Hiroki ENDO, Shuhei TAGUCHI, Shinya TAKEUCHI, Kazuya FUJISAWA, Ken ...
Session ID: 15C-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
To simplify user’s operation considering multiple distribution media, we have proposed Media-Unifying
Platform which automatically selects the appropriate delivery media according to the device function and the user’s situation. This paper describes a distribution status management system compatible with personal recording content and temporarily provided contents in a restricted space.
View full abstract
-
Shuhei TAGUCHI, Hiroki ENDO, Shinya TAKEUCHI, Kazuya FUJISAWA, Ken ...
Session ID: 15C-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We have proposed the Media-Unifying Platform that enables viewers to receive contents in each appropriate delivery medium according to device functions and a viewers’ situation. This paper describes a method to manage distribution status of multiple content providers by coordinating their servers.
View full abstract
-
Ryosuke ABO, Yoshihiro FUJITA
Session ID: 15C-4
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We propose an efficient method of distributing text data.by subtitle channel using markup languages. An experimental results showed better performances by applying the hybrid markup language we have already developed.
View full abstract
-
Kazushige HIRAMATSU, Hiromu OGAWA, Masaya IKEO, Hiroshi FUJISAWA
Session ID: 15C-5
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We are developing “Hybridcast Connect X”. Because it makes it possible to seamlessly contact the contents of the broadcasting station in various situations of daily life. In this paper, we report the verification of the system that TV control, using voice assistant and “Hybridcast Connect X”.
View full abstract
-
Kazuto OGAWA, Sakurako TAMURA, Goichiro HANAOKA
Session ID: 15C-6
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
To make it easier for subscribers to obtain TV services outside their homes, we propose a model using a mobile terminal. It uses a temporary key and the mobile terminal has a role to take secret data related to the temporary key.
View full abstract
-
Pages
21A-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Wataru YONEDA, Yoshitsugu MANABE, Noriko YATA
Session ID: 21A-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We proposed a method of 3D measurement for inside surface using fish-eye camera and concentric circle patterns. However we cannot measure automatically with the use of the proposed method yet. This paper proposes a calibration method of the proposed system in order to realize the automatic measurement.
View full abstract
-
Daiki MIYAZAKI, Naoki HASHIMOTO
Session ID: 21A-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In marker-based deformation detection, taking account of only the detected marker limits the available positional information to only the marker, and causes excess or deficiency of the projected images. By using the information other than the marker, we find out adequate projection regions and generate correctly projected images.
View full abstract
-
Yuka NAKAMURA, Naoki HASHIMOTO
Session ID: 21A-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
Recently, image projection has attracted a lot of attention, and a simple procedure of the projection in a wide area has been strongly requested. In this research, we improve our geometric correction method combining a fish-eye lens camera with a standard lens camera, and evaluate with some practical projections.
View full abstract
-
Yuki MORIKUBO, Naoki HASHIMOTO
Session ID: 21A-4
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We improve the accuracy of edge-based posture estimation by using contour of the target and that of continuity without disturbing processing speed. This method realizes 3D posture estimation for dynamic projection mapping that can be applied to various target objects and usable under circumstances containing noises such as user’s hand.
View full abstract
-
Akito ARAI, Ryoji HATTORI, Junya MIYAGI, Seiji OKUMURA
Session ID: 21A-5
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
It is difficult to estimate spatial people density using single- viewpoint image when the occlusions are generated. In this report, we propose a method for estimating spatial people density using single-viewpoint image.
View full abstract
-
Pages
21C-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Keiichirou Uozumi, Sumio Yano
Session ID: 21C-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
The method of the sub-pixel detection was adopted to detect the center of the pupil in the eye-ball. As results, we found this method was available to measure the precise evaluation of the gaze point.
View full abstract
-
Hiroyuki KAWAKITA, Hiroshi FUJISAWA, Tobias Höllerer
Session ID: 21C-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper, we report the characteristics and the research tasks about expressiveness of 2D moving picture on augmented reality(AR) glasses as a new TV. Since the TV has high degree of freedom of the position, we focused on reality, visibility, interactivity in terms of user experience(UX).
View full abstract
-
Takuya HANDA, Makiko AZUMA, Toshihiro SHIMIZU, Satoru KONDO
Session ID: 21C-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We have developed a novel haptic interface. This device renders the surface shapes and hardness of a virtual object that is represented in an absolute position in real space. In this paper, we describe an overview of a spatial position tracking method that improve utility of our device in using aerial space.
View full abstract
-
Tomoki Furukawa, Hirofumi Mori, Hatsuo Yamasaki, Mueo Yamade, Tomo ...
Session ID: 21C-4
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
We developed a prototype system to evaluate the cognitive function during driving. We evaluate the cognitive function of elderly drivers with this system. We coach to older drivers and promote awareness of their safe driving. We plan to examine the effects of elderly awareness.
View full abstract
-
Naoto KATO, Tsubasa UCHIDA, Makiko AZUMA, Taro MIYAZAKI, Shuichi ...
Session ID: 21C-5
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper describes the computer graphics (CG)-represented mouth patterns for place names by our developed synthesis system in Japanese Sign Language. In our synthesis system, the mouth patterns are made by combining the basic CG animations such as vowels, and by synthesizing the CG animations. About 130 CG mouth patterns of place names are generated by using our synthesis system and 20 mouth patterns randomly sampled from them are evaluated to be understandable by two of three native signers.
View full abstract
-
Pages
22A-0-
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Natsumi SUZUKI, Yohei NAKADA
Session ID: 22A-1
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper proposes an effects selection tool via a visual saliency map to improve visual attraction of an object in a target image. This tool can improve visual attraction of the target object with suppressing a large sense of incongruity to viewers in reasonable computational time.
View full abstract
-
Satoru Ushijima, Yohei Nakada
Session ID: 22A-2
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
This research proposes a novel dynamic saliency map to predict movements of humans’ eyes while driving. This dynamic saliency map is obtained as a mixture of multiple characteristic maps with mixing weights, which are estimated by using observed fixation point data for in-vehicle camera images.
View full abstract
-
Akari MIYAMOTO, Naho YAGI, Yohei NAKADA
Session ID: 22A-3
Published: 2017
Released on J-STAGE: October 22, 2022
CONFERENCE PROCEEDINGS
FREE ACCESS
This research proposes a visualization method for textile images of apparel products based on consumers’ sensitivities. The proposed method extracts sensitivity-based features of textile images from non-metric similarity data obtained from consumers. Then, this method constructs a prediction model of the sensitivity-based features from image features.
View full abstract