ITE Technical Report
Online ISSN : 2424-1970
Print ISSN : 1342-6893
ISSN-L : 1342-6893
34.25
Showing 1-35 articles out of 35 articles from the selected issue
  • Type: Cover
    Pages Cover1-
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (16K)
  • Type: Index
    Pages Toc1-
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (159K)
  • Kentaro KIMURA, Osamu HOSHUYAMA, Tomohiro TANIKAWA, Michitaka HIROSE
    Type: Article
    Session ID: MVE2010-23
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    This paper proposes a sound image presentation system on large screen using parametric loudspeaker and paraboloid reflector diffusing the directional sound on the screen. The diffusion of an ultrasonic directional sound on a paraboloid forms a sound focus of the paraboloid, and enables a clear sound image at any point on the screen. Evaluation of the proposed system implemented as a large screen with a projector shows that the proposed system has 26% higher sound localization accuracy than a conventional sound localization method, Vector Base Amplitude Panning method.
    Download PDF (1720K)
  • Hiroki NAKAHARA, Michiko OHKURA
    Type: Article
    Session ID: MVE2010-24
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    By examining the effect of sound physical characteristics to the presence of sound, we can show an index to evaluate the presence of sound. Although there are various sound physical characteristics and various words to express sounds impression, we selected sampling frequency as a sound physical characteristic and "powerful" as a word to express impression related to "powerful" for the experiment this time. At audition of sound through headphone, we performed an experiment in which 16 participants heard 4 orchestra music and 4 environmental sounds and answered a questionnaire in order to clarify the difference of sampling frequency of sound affects the sound impression of power. As the result, it was confirmed that a certain kind of change of sampling frequency of sound affect the difference of impression evaluation related to "powerful."
    Download PDF (1432K)
  • Masahiko KUBARA, Takahiro YAMAMOTO, Mamoru ENDO, Makoto ITO
    Type: Article
    Session ID: MVE2010-25
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    To the walking rehabilitation activity, we are working on the measurement and the evaluation with the shoe-type device equipped with two or more sensor groups like the pressure, the acceleration, and the distance, etc. When considering real-time feedback activity evaluation to patients in rehabilitation activities, using the visual information had the problem of obstructing the patient's activity. In this study, we propose to present the activity evaluation by aural, concretely we propose the real-time feedback mechanism that the activity evaluation to the patient by the harmony synthesis, the performance tempo, and the voice guide, etc.
    Download PDF (1365K)
  • Mitsuhiko YAMAZAKI, Kunihiro NISHIMURA, Tomohiro TANIKAWA, Michitaka H ...
    Type: Article
    Session ID: MVE2010-26
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    To resolve problems of information sharing using a head mounted display, we proposed a solution which uses a hand-held projector. In our research, we established a projector-camera system which enables high presence and spatial information sharing, and a virtual large-screen was created by scanning a hand-held projector. We developed two systems and evaluated estimation accuracy, and conducted a user survey. As a result, improvements of high presence and spatial recognition were observed.
    Download PDF (1323K)
  • Yuichiro HAYAKAWA, Yuichi MORI, Ryugo KIJIMA
    Type: Article
    Session ID: MVE2010-27
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Method for solving time delay problem is proposed. It is called Reflex compensation using VRAM. And a Video see-through HMD with Reflex compensation using VRAM is developed. The Reflex compensation using VRAM is realized using the gyroscope sensor to measure the head rotation during the lag period, embedded CPU, FPGA and SRAM to modify the output image. A HMD with this Method provide the rapider response to the user's head motion. A Video see-through HMD with the Reflex compensation usign VRAM was evaluated by tracking test. As a result, it is proved that A Video see-through HMD with the Reflex compensation using VRAM is better than Normal Video see-through HMD.
    Download PDF (919K)
  • Michiko OHKURA, Sittapong SETTAPAT, Tetsuro AOTO, Yuki KAKEGAWA, Tiran ...
    Type: Article
    Session ID: MVE2010-28
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    A trial of impression evaluation of shared virtual spaces was performed between Japanese university and Thai university. Presenting 2-dimensional plane images and 3-dimensional stereoscopic images of the same animation contents, participants evaluated those impressions using questionnaire sheets. Comparing the evaluation results, it was shown that evaluation scores were different in some questionnaire items respectively between Japanese university and Thai university, between 2-dimensional images and 3-dimensional images, and between contents. In addition, we confirmed the effectiveness of interactive content from its evaluation. Based on the result of this trial, we will plan to design an interactive remote teaching using shared virtual space.
    Download PDF (1052K)
  • Wataru MIYAZAKI, Takashi YOSHINAGA, Kohji MASUDA
    Type: Article
    Session ID: MVE2010-29
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Echography has been used in every field of medical diagnosis because of its safety and cost-effectiveness. However, the skill to recognize relationship between ultrasound probe and internal organ is required to acquire echocardiogram. Therefore we have developed an extraction software of internal organs from real-time echogram. Then we have applied to a reconstruction system of 3D shape of internal organs based on AR (Augmented Reality) and VR (Virtual Reality) technology using ARToolKit. Furthermore, GUI interface to visualize the internal organ was developed by OpenGL. We confirmed the usefulness of the system for the reconstruction of the shape of left ventricle and liver to visualize of internal organ.
    Download PDF (1593K)
  • Hiroki MORIGUCHI, Yoshihiro KURODA, Hidekazu TAKIUCHI, Masataka IMURA, ...
    Type: Article
    Session ID: MVE2010-30
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In AR surgery, a surgeon is supported by the 3D organ model that overlaid on the surgical view. When using an oblique-viewing scope, the rotation of cylinder changes the optical axis of the endoscopic camera. Thus, the angle of rotation is required to estimate the viewing direction in addition to the posture of the cylinder. In a previous study, rotary encoder is used for acquisition of the rotational angle of the cylinder. However, calibration is required since the cylinder makes a precession movement against ideal rotation axis. Besides, the rotary encoder attached to the endoscope tends to be big. In this study, we calculate the posture of the cylinder against the camera head by using optical marker and accelerometer. Accelerometer detects the orientation of gravity. The posture of the cylinder is calculated from the change of the orientation of gravity. The proposed method enables to estimate the viewing direction of oblique-viewing endoscope camera without calibration. In this presentation, we report the details of the proposed method and the result of AR images obtained by using our method.
    Download PDF (488K)
  • Yuki HAMADA, Yoshiyuki KAGIYAMA, Yoshihiro KURODA, Masataka IMURA, Osa ...
    Type: Article
    Session ID: MVE2010-31
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In recent years, VR technology is counted on surgery simulator in medicine. The Aim of this study is to develop a rotational acetabular osteotomy (RAO) simulator. In this paper, the impact-breaking of the elastic-plastic body is described. RAO is the surgical operation with cutting hip socket and transrating or rotating it. We assume that bone has elasto-plastic property. Therefore, we simulate to transfer the impact-wave and to break the bone when impacting with elasto-plastic body. We also display the haptic sensation of breaking bone considering user's handling and states of bone. In this experiment, we use spring-mass model, and confirm that this system can display the stress wave propagation and fracture by shock.
    Download PDF (1000K)
  • Kazuyoshi TAGAWA, Hiromi T. TANAKA
    Type: Article
    Session ID: MVE2010-32
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In this report, we state a hybrid dynamic deformation model for haptic interaction with specific portion. This model is capable of haptic interaction and cutting manipulation with large scale dynamically deformable object. We implemented this approach into our prototype system and simulation results by the approach were presented.
    Download PDF (864K)
  • Koichi HIRAGA, Kaoru WAKABAYASHI, Norihiko MATSUURA
    Type: Article
    Session ID: MVE2010-33
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    A virtual touchpad system has been developed that supports basic mouse operations such as moving the cursor, clicking, dragging and clutching simply by having the user gesture as if operating with a touchpad. Two cameras are used to capture images of the user's fingertip movement. The 3D trajectory of the user's fingertip is obtained from these pairs of images, and the intended operation is then extracted by analyzing the 3D trajectory. Preliminary experiments with both skilled and novice users were used to evaluate the system performance, and the results were discussed.
    Download PDF (833K)
  • Koji TERAKAWA, Masahiro URA, Takatoshi NAKA, Masashi YAMADA, Mamoru EN ...
    Type: Article
    Session ID: MVE2010-34
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In CG production, a flat panel display is generally used, and 3D model is manipulated by flat input devices such as a mouse. However, it is not intuitive to manipulate a 3D model by single pointing on a flat surface. In this study, we propose a manipulation interface which users can pose a 3D human model on a flat surface by using a multi-touch tabletop. We define multi touch gestures to manipulate the 3D human model spatially by surface input. We also evaluate the proposed interface by posing the 3D human model by using it.
    Download PDF (715K)
  • Takashi KAJINAMI, Oribe HAYASHI, Takuji NARUMI, Tomohiro TANIKAWA, Mic ...
    Type: Article
    Session ID: MVE2010-35
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In our research, we propose Digital Display Case system designed based on conventional display cases which enables museums to convey the background information about their exhibits. We categorized the background information into two categories, the synchronicity and diachronicity. Synchronicity includes relations among exhibits which were made in the same age, while diachronicity includes history or provenance of exhibits. Then we consider how to convey these kinds of information effectively. Based on this consideration we made the interactive exhibition system composed of the case system and the panel system, and made the exhibitions which convey background information about Doguu, Japanese ancient figure. We got review over the system and exhibition from museum, and consider its usefulness.
    Download PDF (1660K)
  • Yutaka NAKAJIMA, Takao SATO
    Type: Article
    Session ID: MVE2010-36
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    When a square translates in a circle and vertices of the square are hidden by occluders, The physical motion of not-occluded contours of the square is sinusoidal oscillation in the direction normal to contours' orientation. Though we can mainly perceive the moving square, not oscillating contours. This is known as Motion binding. In this study, we investigate effects of depth information on Motion binding. All stimuli had a same 2D diamond structure, and we added binocular disparity to each contour independently. As results, Motion binding was inhibited when all contours had crossed disparity and when each contour had different disparities. While the combination of the disparity adding each contour also affect the perception of Motion binding. These results indicate that depth information would constraint stronger than 2D structure in Motion binding.
    Download PDF (1043K)
  • Wen WEN, Toru ISHIKAWA, Takao SATO
    Type: Article
    Session ID: MVE2010-37
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    This study examined how different components of working memory are involved in spatial knowledge acquisition for good and poor sense-of-direction people. We employed a dual-task method, and asked participants to learn routes from videos with verbal, visual, and spatial interference tasks and without any interference. Results showed that participants with a good sense of direction encoded landmarks and routes verbally and spatially, and integrated knowledge about them into survey knowledge with the support of all three components of working memory. In contrast, participants with a poor sense of direction encoded landmarks only verbally, and tended to rely on the visual component of working memory in the processing of route knowledge. Based on the results, a possible model for explaining the differences in spatial knowledge acquisition and sense of direction was proposed.
    Download PDF (649K)
  • Kenchi HOSOKAWA, Takao SATO
    Type: Article
    Session ID: MVE2010-38
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Generally motion parallax studies assume that there is linear movement between the object with the observer.. However a relative motion stimulus is able to be interpreted as both linear movement and rotation. In this study we investigated perceptual alternation between linear movement and rotation during 5 minutes observation on a relative motion stimulus. As a result, observer perceived mostly linear movement in the first 120 seconds. After that, however, their perception is altered between linear and rotational movement. Additionally, perceptual depth order is also reversed as perceptual motion is altered.
    Download PDF (830K)
  • Yuho YAMAZAKI, Masahiro ISHII, Zheng TANG, Kazuya YAMASHITA
    Type: Article
    Session ID: HI2010-90
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    When the passing speed of the self-projection thing like the mouse pointer changes, inner force sense is perceived. This phenomenon is called Pseudo-Haptics. Without giving power with the motor when Pseudo-Haptics is used, The inner force sense can be felt. As for Pseudo-Haptics, the application to haptic interface is hoped for. To put Pseudo-Haptics to practical use, the consideration of the environment that gives Pseudo-Haptics easily is needed. In this research, the influence that the load that joins straining and the arm of the arm when operating it exerts on the perception of Pseudo-Haptics is investigated.
    Download PDF (510K)
  • Kimiaki Tokumaru
    Type: Article
    Session ID: MVE2010-39
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    The difference between human and non-human animals is simply digital or analog communication methods. Dogs also have concepts and establish concept system. Once concept system is constructed in the mind, it works as the framework of world perception and hides some reality. Isn't it a virtual reality? The human mistake is to have developed its civilization disconnected with and antagonistic to the nature. It is time to create a new civilization harmonious with the nature.
    Download PDF (1089K)
  • Toshiki TAKEUCHI, Takuji NARUMI, Kunihiro NISHIMURA, Tomohiro TANIKAWA ...
    Type: Article
    Session ID: MVE2010-40
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    A purpose of this research is to improve user's balance of income and expenditure by forecasting and presenting future expenditure from past purchase information. First, we logged various purchase information by scanning large number of receipts and applying them to OCR (Optical Character Recognition). We call this system "Receiptlog". From data of Receiptlog, we found out the periodicity with time and place in expenditure behavior. Making use of it, we developed an iPhone application, "Expenditure Forecast", which forecasts an amount and possibility of future expenditure and presents them to users. We evaluated this application, and found out that presentation of forecast affected users' expenditure behaviors.
    Download PDF (1105K)
  • Takahiro KOIZUMI, Yuichi NAKAMURA, Shin'ichi SATOH
    Type: Article
    Session ID: MVE2010-41
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Lifelog, that is a personal experience record in a daily life are expected to be a great help for people who have problem with of memory disorder such as cognitive impairment. In this paper, we propose a novel framework of scene retracing in personal view videos for the purpose of searching for a lost object. For a flexible memory retrieval for finding lost objects as humans do, a variety of clues such as places, time, behaviors, objects, etc. are necessary in reaching to the relevant portion. For example, places where a lost object tends to be left, situations in which the user actually put a similar or identical object, scenes of moving with the object would be good clues for finding the object. Detection and browsing of those scenes are helpful for efficient search for the object. In this paper, we first discuss useful scenes, and then the features for detecting those scenes. Then we show the experiment with the personal eye-view videos how the clues can be detected from those videos.
    Download PDF (983K)
  • Takahiro YAMAMOTO, Masahiko KUBARA, Mamoru ENDO, Makoto ITO
    Type: Article
    Session ID: MVE2010-42
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    The negative aspects of current society claims comfortable sleep. To bringback natural sleep, intelligent alarm clock is proposed with utilize aquision and analysis system of pulse beat based LifeLog. And we propose that to be utilized the music used to to wake up or relax as comfortable life and sleep. It is being developed and experimented. Specifically we developed wearable portable device for logging pulse beat, alarm clock type sound device and PC software to be used for analysis, and we experimented a part of it.
    Download PDF (1759K)
  • Kazuhiro KASADA, Takuji NARUMI, Tomohiro TANIKAWA, Michitaka HIROSE
    Type: Article
    Session ID: MVE2010-43
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In our research, we developed a system which enables people to experience past scene impressively. This system is assumed to be used at a place where past traces exist. Firstly, it estimates a camera position of a past photo or an animation. Secondly, it connects present image captured by web camera with past images naturally at the estimated place. The system estimates a camera position by using 3 points in order to do it at a place with a past trace. We showed the accuracy of the estimation by simulations and experiments. Next, we made sure that the connections between present and past images were done seamlessly. Finally, experiments with some subjects showed that this system express past situation effectively.
    Download PDF (1527K)
  • Xiaoguang Hei, Mamoru Endo, Shigeki Yokoi
    Type: Article
    Session ID: MVE2010-44
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In this paper, a novel identification method based on an image capturing system is proposed to produce Object VR movies. Object VR movies have been successfully used in many applications. In order to spread its range of uses, for more people and companies could use this technique easily, an auto Object VR movie generator has been developed. This system is a regulating system which includes the motorized rig for image capturing and the rig's control and image processing software. Users could make their works even if they don't have any specific knowledge about Object VR movies, and also, be a low cost system. From the result of the experimentation, it verified that our design is fit for the project.
    Download PDF (1702K)
  • Takashi KOBAYASHI, Masafumi NODA, Daisuke DEGUCHI, Tomokazu TAKAHASHI, ...
    Type: Article
    Session ID: MVE2010-45
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    On the social networking and micro-blogging service "Twitter", there has been increasing "on-the-spot comments" while watching a sport game at a stadium or watching TV at home. We aim at the detection of these on-the-spot comments among other comments on Twitter. This paper reports an experiment to detect on-the-spot comments for TV programs, analyzing the frequency of occurrence of words by focusing on both program independent and dependent information.
    Download PDF (357K)
  • Shigeki MORISHIMA, Tomohiro MASHITA, Kiyoshi KIYOKAWA, Haruo TAKEMURA
    Type: Article
    Session ID: MVE2010-46
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    We have previously proposed a remote collaboration system that supports an asymmetric interaction between an instructor and a worker at distant locations. The instructor is able to observe a remote environment from an arbitrary point of view by using a 3D model with projective texture mapping of live video captured by a remote worker's camera. However, the 3D model is manually prepared in advance and not updated during collaboration. To tackle this limitation, we integrate an omnidirectional range sensor into the previous system. A preliminary experiment has shown subjective effectiveness of the proposed system.
    Download PDF (1636K)
  • Masaaki MIYAURA, Takushi NARUMI, Kunihiro NISHIMURA, Tomohiro TANIKAWA ...
    Type: Article
    Session ID: MVE2010-47
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    An olfactory display which emits odors just in front of the nose can control their strength temporally and spatially. It can enhance their benefit by adjusting their strength according to a user's condition. We examined their effects on user's tasks when we control their strength according to user's concentration ratio, which we estimated from his/her heart rate fluctuation (RRV). Miss rates of user's tasks decreased 30% on average when the display emits odors when user's concentration ratio is low compared to when it doesn't. Moreover we found it is more effective to control them than to emit it constantly or off and on.
    Download PDF (1218K)
  • Haruaki TAMADA, Akihiro OGINO, Hirotada UEDA
    Type: Article
    Session ID: MVE2010-48
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In this paper, we propose a learning support framework for programming beginners by the one application of human-robot interaction. The proposed framwork supports each learners and the teacher and does not prevent learning of learners. Specifically, the proposed framework collects compilation errors from each learners and analyzes them. Then, the framework shows resultant data to the teacher for understanding comprehension of learners. When the learner encounters problems while solving an assignment, the framework gives hints for solving problems using an assistant robot. The learner can use the framework without being aware of it.
    Download PDF (1048K)
  • Masakazu YAMAOKA, Shun IDO
    Type: Article
    Session ID: MVE2010-49
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    It is necessary to express the physical behavior of the rheological object such as organ and jelly under the virtual space and the 3D simulation circumstance correctly. In this paper, we propose the system which acquires the surface information of the rheological object from the texture. It enables us to deal with more specific surface information of the rheological object according to the texture size.
    Download PDF (1617K)
  • Yasuyuki SASAKI, Kazuyoshi TAGAWA, Hiromi TANAKA
    Type: Article
    Session ID: MVE2010-50
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    The development of the surgical simulation system attracts attention in a medical field. However, it is difficult to perform the simulation of the physics base in real time. Therefore, in this paper, suggestion and implementation technique to calculate the parallel that I used GPU for for the online re-mesh type flexibility thing which could lower a calculation cost. And we are able to do about 20 times speedup in calculation time when we compare it when we calculate with a CPU.
    Download PDF (728K)
  • Noriyuki UDA, Yoshitaka KAMIYA, Shigeki Yokoi
    Type: Article
    Session ID: MVE2010-51
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Glare Effecter is a computer graphics tool to express and render the brightness of the light source. It is consisted of flare and streak. Flare is diffusing light surround the light source and streak has a light trace. The purpose of this study is to confirm the glare effect of streak with luminance gradients. Our comparative experiments between flare image with diffusing luminance and streak with luminance gradients showed the amplifier effects of brightness. Streak amplified the glare effect of flare in both condition of back light, low luminance (#00) and high luminance (#5A). In former condition the contrast effect between light luminance and back luminance amplified the glare effects. In another case, the self luminosity of glare (Zavagno: 1999) occurred around the adjacent area of streak and flare affected brightness perception and made the glare more bright.
    Download PDF (1571K)
  • Type: Appendix
    Pages App1-
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (102K)
  • Type: Appendix
    Pages App2-
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (102K)
  • Type: Appendix
    Pages App3-
    Published: June 29, 2010
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (102K)
feedback
Top