Title: A study on quantitative analysis of the relationship between camera image characteristics and operability of rescue robots

Authors: Noritaka Sato; Masayoshi Koiji; Yoshifumi Morita

Addresses: Department of Electrical and Mechanical Engineering, Nagoya Institute of Technology,Nagoya, Aichi, Japan ' Department of Electrical and Mechanical Engineering, Nagoya Institute of Technology,Nagoya, Aichi, Japan ' Department of Electrical and Mechanical Engineering, Nagoya Institute of Technology,Nagoya, Aichi, Japan

Abstract: Many rescue robots have been developed that are teleoperated using images from a camera mounted on the robots. This study examined how the conditions of the images from the camera affect the operability of a rescue robot used for a search task in a 2D maze. Two experiments with subject were performed using a robot simulator USARSim, whereby the first one was a preliminary experiment. As a first, this study compares 100 combinations of the position and posture of a camera mounted on a robot. In the second experiment, 16 images with various indices were used in the main experiment. The indices xmin and h use the images captured by the camera to provide information relating to the robot and the environment around the robot, respectively. These were defined from the results of the preliminary experiment. The tasks performed and scoring method used for the experiments were determined by referring to the DHS-NIST-ASTM International Standard Test Methods for Rescue Robots. The results of the main experiment showed that the operability of the robot has an extreme value when the indices (xmin, h) were (100, 3,000).

Keywords: mobile robot; rescue robot; teleoperation; user interface; remote control; situation awareness; camera image; quantitative analysis; simulator; search and rescue.

DOI: 10.1504/IJAMECHS.2017.099313

International Journal of Advanced Mechatronic Systems, 2017 Vol.7 No.6, pp.350 - 358

Received: 13 Jul 2017
Accepted: 30 May 2018

Published online: 26 Apr 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article