A multi-camera person tracking system for robotic applications in virtual reality TV studio

Abstract

In this paper, an integrated multi-camera person tracking system for virtual reality television studios (VR-TV) is presented. The system robustly tracks the moderator while freely moving, sitting or walking around the studio, and the estimation result can be used in order to drive the main broadcasting camera mounted on a large robotic arm. Application of the proposed tracking system to real-time VR-TV results in an autonomous robot cameraman, able to keep the moderator inside the screen with jitter-free viewpoint adjustments, as required by the scene rendering engine.