抄録
The rapid expansion of multimedia techniques has made stereo display systems more prevalent than ever. Due to the insufficiency of stereo sources, we propose a framework for generating visual stereo image pairs based on the common 2D images. The framework consists of two parts: rendering method by exploiting the pseudo 3D structure and synthesis method by exploiting the motion information between image frames. The tentative experimental results illustrates that these approaches can help viewer to perceive and understand the three-dimensionality of the world.
