Joint multi-video compression for robot explorations

Abstract

To assist human shared control for robotic explorations in complicated environments, it is essential to transmit multiple videos of monitoring wheel-land interactions to human control stations. Since multiple video data is very huge, in this research, we have developed an effective joint compression scheme for multiple robot videos to assist human shared control. The proposed scheme is flexible which can work for multiple videos with both the same and different resolutions. To jointly regulate the overall compression rate to meet the available network bandwidth while obtaining optimal encoding quality, we propose a new joint rate control approach which can intelligently perform target bit allocation among multiple videos and dynamically determine encoding parameters. Further, our scheme also allows flexible priority adjustment among multiple videos to meet diverse needs from human controllers. Our experimental results have demonstrated that the proposed scheme can effectively compress multiple robot videos and regulate overall encoding rates meet diverse channel bandwidths.