User experience tests over a Mixed Reality interface of a SAR system with UAVs

Authors

Abstract

In this paper, two main aspects of a Search and Rescue (SAR) operations system with Unmanned Aerial Vehicles (UAV) are addressed. Firstly, an architecture that contains modules with specific tasks, providing to non-expert pilots an easier mission planning of UAVs over Mixed Reality (MR) interfaces. Secondly, a deep convolutional network trained through transfer learning techniques and pictures taken from the air, used to improve human detection in natural disaster zones. Results of a subjective testing approach by expert and non- expert pilots of the mixed reality interface of the system are presented. The results, based on a Mean Opinion Score (MOS) metric of Quality of Experience, show that users have a good perception of the video transmission with human detection.

Conference

RECPAD 2019 - Portuguese Conference on Pattern Recognition, October 2019, October 2019