Yuichi Ohta

Japan's largest platform for academic e-journals: J-STAGE is a full text database for reviewed academic papers published by Japanese societies

We propose to visualize pre-recorded activity of a person on a movable transparent screen for in-situ reviewing of his/her activity in augmented reality fashion. Activity of the target person was taken as a video by a surveillance camera. Viewers can watch the activity in the scene as if it happened there because segmented image of the target person was projected onto a human-size transparent screen and other static objects around the target person can be visible throughout the transparent screen. Our final goal is to move the transparent screen by mounting it on a small robot and moves the robot to follow the target person in the video. On the way to the final goal, we currently assume that the viewers move the screen manually so as to pose the screen at the same location of the target person in the video. A tracking method of the transparent screen in the scene is devised by utilizing projector-camera calibration and simple infrared marker tracking.

In this paper, we introduce a method to estimate 3D position of a badminton shuttle using unsynchronized multiple-view videos. The research of object tracking for sports is conducted as an application of Computer Vision to improve the tactics involved with such sports. This paper proposes a technique to stably estimate object's position by using motion blur that used be considered as observational noise in the ordinary works. Badminton shuttle has a large variation of the moving speed, the motion trajectory is unpredictable and moreover the observation size is very small. Thus, it cannot be grasped correctly with human eyes. We apply our proposed technique to badminton shuttle tracking to confirm the ability of our method to enhance the human vision. We also consider that there is some contribution to augment sports in future.

A lot of efforts aim to realize a society where autonomous vehicles become general transportation in industrial, political and academic fields. In order to make autonomous vehicles more familiar with the public, it is necessary to develop not only advanced auto driving control but also comfortable environments for the passengers. This paper proposes our trial to improve comfort of passengers on autonomous vehicles. We developed an experimental vehicle equipping Mixed Reality (MR) display system which aims to reduce anxiety using visual factors. Our proposed system visualizes a road surface that is out of the passenger's field of view by projecting the see-through image on the dashboard. Moreover, it overlays the computer graphics of the wheel trajectories on the displayed image using MR so that the passengers can easily confirm the auto driving control is working correctly. The displayed images enable passengers to comprehend the road condition and the expected vehicle route out of the passenger's field of view. We investigated change of the mental stress by introducing methods for measuring physiological indices, heart rate variability and sweat information.

Mixed Reality (MR) can merge real and virtual worlds seamlessly. This paper proposes a method to realize smooth collaboration using a remote MR, which makes it possible for geographically distributed users to share the same objects and communicate in real time as if they are at the same place. In this paper, we consider a situation where the users at local and remote sites perform a collaborative work, and real objects to be operated exist only at the local site. It is necessary to share the real objects between the two sites. However, prior studies have shown sharing real objects by duplication is either too costly or unrealistic. Therefore, we propose a method to share the objects by virtualizing the real objects using Computer Vision (CV) and then rendering the virtualized objects using MR. We have proposed a remote collaborative work system to create a smoother user experience for collaborative work with virtualized objects for remote users. Through experiments, we confirmed the effectiveness of our approach.

複数視点から撮影した画像を用いて注目物体の３次元位置を推定する場合，各画像が同期撮影されていることが前提となるが，大規模空間で同期撮影が困難となるケースが頻出する．運動モデルを仮定すれば３次元位置を近似的に推定することができるが，被写体が高速かつ変則的に移動する場合はモデルのあてはめが困難である．本稿では，大規模な空間において非同期撮影された映像を用いて，高速かつ変則的に移動するバドミントンシャトルの３次元位置を推定する手法を提案する．露光時間を長く設定したカメラで撮影した画像を用いてシャトルの３次元軌道を復元し，そこに短い露光時間で撮影した画像で観測されたシャトル領域を投影することにより，正確な３次元位置推定を実現する．実写画像とCGシミュレーション画像を用いた実証実験により，提案手法の有効性を確認する．利用は著作権の範囲内に限

This paper proposes a method to estimate a mobile cameras position and orientation by referring to the corresponding points between aerial-view images from a GIS database and mobile camera images. The mobile camera images are taken from the users viewpoint, and the aerial-view images include the same region. To increase the correspondence accuracy, we generate a virtual top-view image that virtually captures the target region overhead of the user by using the intrinsic parameters of the mobile camera and the inertia (gravity) information. We find corresponding points between the virtual top-view and aerial-view images and estimate a homography matrix that transforms the virtual top-view image into aerial-view image. Finally, the mobile cameras position and orientation are estimated by analyzing the matrix. In some cases, however, it is difficult to obtain a sufficient number of correct corresponding points to estimate the correct homography matrix by capturing only a single virtual top-view image. We solve this problem by stitching virtual top-view images to represent a larger ground region. We experimentally implemented our method on a tablet PC and evaluated its effectiveness