STEREOSCOPIC CONTENT PRODUCTION OF COMPLEX DYNAMIC SCENES USING AWIDE-BASELINE MONOSCOPIC CAMERA SET-UP

Conventional stereoscopic video content production requires use of dedicated stereo camera rigs which is both costly and lacking video editing flexibility. In this paper, we propose a novel approach which only requires a small number of stan-dard cameras sparsely located around a scene to automatically convert the monocular inputs into stereoscopic streams. The approach combines a probabilistic spatio-temporal segmenta-tion framework with a state-of-the-art multi-view graph-cut reconstruction algorithm, thus providing full control of the stereoscopic settings at render time. Results with studio se-quences of complex human motion demonstrate the suitabil-ity of the method for high quality stereoscopic content gener-ation with minimum user interaction. Index Terms — Stereoscopic rendering, 3D video, multi-ple view reconstruction, segmentation. 1