ヤノウ　アキラ   矢納　陽    所属   川崎医療福祉大学  医療技術学部 診療放射線技術学科    職種   教授
言語種別	英語
発表タイトル	Robustness of Visual-Servo against Air Bubble Disturbance of Underwater Vehicle System Using Three-Dimensional Marker and Dual-Eye Cameras
会議名	MTS/IEEE OCEANS'15
学会区分	国際学会及び海外の学会
発表形式	口頭
講演区分	一般
発表者・共同発表者	◎Myo Myint, Kenta YONEMORI, Akira YANOU, Shintaro ISHIYAMA and Mamoru MINAMI
発表年月日	2015/10
開催地 （都市, 国名）	Washington, DC, America
概要	After the Tohoku Earthquake on March 11, 2011, FUKUSHIMA prefecture has been afflicted by disasters like earthquake, tsunami and nuclear power accident, then decontamination work has been needed to be done in radioactive contamination area. A visual-servo type underwater vehicle system with binocular wide-angle lens has been developed and it has expanded the sphere for surveying submarine resources and decontaminating radiation from mud in dam lake, river, regulation pond, sea in FUKUSHIMA. This paper studies the robustness of visual-servo type underwater vehicle system using three-dimensional (3D) marker under air bubble disturbance on real-time pose (position and orientation) tracking for visual servo. The recognition of vehicle's pose through 3D marker is executed by Genetic Algorithms (GA). The proposed system does not merely calculate the pose information, but can recognize the target pose information through GA while visual servo, because the system utilizes a 3D marker shape and color to recognize the marker. In our previous research, a performance of the system to regulate the vehicle's pose to the relative desired pose against the 3D marker was explored under the condition that there was no disturbance on images. Therefore this paper studies the robustness of the proposed system for air bubble disturbance on the image, aiming at confirming the robustness of the proposed visual servo system. The following results were derived; (1) The proposed system is robust to time-variant target position in zaxis (front-back direction of the vehicle). (2) Although the fitness value of GA is influenced by disturbance, the system can keep recognizing the pose of 3D marker, and (3) tracking by visual servo could be stably kept under the air bubble disturbances.