(A) single image representation model for efficient multi-view stereoscopic content creation

Abstract

Computer graphics is one of the most efficient ways to create a stereoscopic content. The process of stereoscopic CG generation is, however, still very inefficient compared to that of monoscopic CG generation. Despite that stereo images are very similar to each other, they are rendered independently. Additional requirements for disparity control specific to stereo images lead to even greater inefficiency. Another inefficiency lies in the manipulation of stereo images entailing repetition of the same task to handle multiple images. This dissertation proposed an efficient framework of stereoscopic content creation utilizing a novel single image representation model. The system automatically generates an optimized single image representation of the entire visible area from the stereo epipolar line. The single image sequence can be easily manipulated with conventional techniques, as it is smooth and maintains the original shapes of scene objects spatio-temporally. A sequence of stereo images pair can be easily generated with an arbitrary disparity setting. These convenient and efficient features are achieved by the automatic generation of a stereo camera pair, robust occlusion detection with a pair of Z-buffers and with slit camera model, an optimization method for spatio-temporal smoothness while automatically balacing both and stereo image pair generation with linear and non-linear disparity adjustment. Experiments show that our framework with the extended technique dramatically improve the efficiency of stereoscopic content creation and editing process.