The influence of camera placement on the situation awareness of the remote supervisor of an autonomous agricultural machine
Loading...
Date
2021-06-10
Authors
Folorunsho, Olayinka
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Work is ongoing to develop fully autonomous agricultural machines where a remotely-located supervisor will replace an onboard operator. A high-quality automation interface will be required to achieve this significant development in agricultural machinery.
Previous researchers have determined that real-time visual information is needed in an
automation interface to instill the necessary high degree of trust. This thesis reports on
our investigation on the impact of camera placement on the ability of the supervisor to
obtain information from real-time video. We pre-recorded video clips for nine
combinations of camera tilt angle (20° 30° and 40°) and height (1.5m, 1.75m, and 2.0m)
to study the camera's location impact yielding nine distinct image velocities. Participants,
recruited from the student population, completed two tasks. First to watch two side-by-side videos and indicate their preferred camera view. Second, watch a video of a sprayer
moving across a field plot and assess the difficulty of identifying and interpreting
randomly placed cues visible in the video clips. This study aims to determine whether
camera placement influences the situation awareness of the supervisor and, if there is an
influence, to identify the camera placement that maximizes the supervisor's situation
awareness. Based on the experimental results from part one's protocol, we conclude that a look-ahead associated with a camera angle of 30º and a camera height of 1.5m was the
preferred placement combination. No significant difference was present between the
various camera angles for the degree of difficulty in detecting and interpreting frisbees.
The test subjects equally rated the 2.0m camera height as preferred in seeing field
information. By comparison, 1.5m camera height was the preferred camera height for
interpreting field information
Description
Keywords
Autonomous agricultural machine, Remote supervision, Real-time visual information