TA04 SICE-ICROS Joint Session: Service Robot Technology and Its Application in Various Fields
Time : 09:10~10:40
Room : 207B
Chair : Takashi YoshimiMin Cheol Lee (Shibaura Inst. of Tech., )
09:10~09:25        TA04-1
Efficient face recognition based on GAP feature

jisu kim(Yonsei University, Republic of Korea), Jeonghyun Baek, Euntai Kim(Yonsei Univerity, Republic of Korea)

In this paper, we propose efficient face recognition based on Grayscale Arranging Pairs (GAP) feature. GAP is a robust holistic feature considering the intensity relationships about all of pixels. Therefore, it has good performance for face recognition. However, GAP feature consider all of pixels and it takes high computational time. In order to reduce computational time, we uses GAP feature in terms of dominant parts such as eyes, nose, mouth and so on. Considering dominant pixels reduces computational time as well as maintains recognition performance.
09:25~09:40        TA04-2
Navigation system development of the underwater vehicles using the GPS / INS sensor fusion

In Uk Lee, Hang LI, Nhat Minh Hoang, Jang Myung Lee(Pusan national university, Republic of Korea)

Sensor fusion of GPS/INS using the Kalman filter design is proposed in this paper. GPS/INS data are utilized for estimating position of AUV (Autonomous Underwater Vehicle) and Kalman filter elaborate the position estimation. The received GPS signals are stable most of times, but they are not accurate. So the INS data (gyroscope sensor, accelerometer, magnetic compass) compensate for the inaccuracy of the GPS data. The noise in the acceleration data from INS data is reduced by Kalman filter in this paper the system performance is confirmed by experiments.
09:40~09:55        TA04-3
A Model-Based System Development of a Walking Support Wheelchair Robot

Yoshinobu Akimoto, Eri Sato Shimokawara, Yasunari Fujimoto, Toru Yamaguchi(Tokyo Metropolitan University, Japan)

This paper proposes a model-based development method with user model in consideration of user. The user model in this paper is an independent model which can behave on behalf of actual human in simulation environments. The user model is effective for developing products used by a variety of users, and is effective also for evaluating the product having some risks damaging the human. In order to evaluate the effectiveness of this method, we have applied this method to the development of automatic stop functions of electric wheelchair.
09:55~10:10        TA04-4
Study on Online 3D Environment Construction for Teleoperation

Masamitsu Kurisu(Tokyo Denki University, Japan)

This paper presents an online 3D map construction by using a mobile robot operated remotely. The 3D map is constructed incrementally from environment data acquired with a stereo camera which is mounted on the robot. In order to decentralize processing loads, the construction system consists of some process modules to be executed in parallel. Details of system architecture and implementation of the system to a small mobile robot are described. Feasibility of the system is discussed based on experimental results of 3D map construction in an imitated environment.
10:10~10:25        TA04-5
A study of towel folding by robot arm -Spreading and vertex detection using image processing-

Tomoya Oshima, Takashi Yoshimi, Makoto Mizukawa, Yoshinobu Ando(Shibaura Institute of Technology, Japan)

We aim to construct a sequence of towel folding operations by a robot arm. We propose the methods for spreading a towel and detecting the grasp positions. We propose how to put the towel on the table. The method lifts one vertex of the towel, and detects the shadow part by using a camera image for recognizing the overlapped condition. For the folding motion, we set the vertices of the towel edges as grasping positions. So, we propose the way to detect the vertices of the towel correctly by generating a rectangle which includes all feature points by using image processing.

<<   1   >>