Spatio-temporal adaptive 2D to 3D video conversion for 3DTV

This paper proposes an efficient scheme to automatically achieve the 2D to 3D video conversion, which exploits different algorithms based on the spatio-temporal relationship of image sequences. For the static sceneries, we directly use geometrical transformation to generate a pair eyes´ frames; while for the motion ones, we firstly extract depth map of the frame, then use the depth image based rendering (DIBR) algorithm to render two viewpoints´ frames. Experimental results show that the proposed method can achieve 2D to 3D conversion effectively and successfully.