Synthesis of Stereoscopic 3D Videos by Limited Resources of Range Images

Recently, depth-image-based rendering has been suggested for 3D video creation, in which a monoscopic video stream is recorded along with a depth (range) video stream and a stereoscopic video stream is then generated at the user end. The fundamental assumption of this framework is the full availability of range images at video rate. In this work we alleviate this hard demand and assume that only limited resources of range images are available, i.e. corresponding range images exist for some, but not all, color images of the monoscopic video stream. We propose to synthesize the missing range images between two consecutive range images. Our approach eases the recording of 3D material by using less expensive range sensors and enables to enhance existing 2D video material with 3D effect by limited manual overhead. Experiments on real videos have demonstrated very encouraging results. Especially, one 3D video was generated from a 2D video without any sensory 3D data available at all. In a quality evaluation using an autostereoscopic 3D display the test viewers have attested similar 3D video quality for our synthesis technique and rendering based on depth ground truth.