Stereoscopic Video Generation Method using Motion Analysis
Donghyun Kim, Dongbo Min, Kwanghoon Sohn · 2007
Stereoscopic video generation method can produce stereoscopic contents from conventional video filmed from monoscopic camera. We propose stereoscopic video generation method using motion-to-disparity conversion considering multiuser condition and characteristics of display device. Field of view, maximum and minimum values of disparity are calculated through initialization process in order to apply various types of 3D display. After motion estimation, we propose three cues to decide the scale factor of motion-to-disparity conversion which are magnitude of motion, camera movement and scene complexity. Subjective evaluation is performed by comparing videos captured from stereoscopic camera and generated from one view of stereoscopic video.