Depth generation method for 2D to 3D conversion

Abstract

An efficient depth map generation method is presented for static scenes with moving objects. Firstly, static background scene is reconstructed. Depth map of the reconstructed static background scene is extracted by linear perspective. Then, moving objects are segmented precisely. Depth values are assigned to the segmented moving objects according to their positions in the static scene. Finally, the depth values of the static background scene and the moving objects are integrated into one depth map. Experimental results show that the proposed method can generate smooth and reliable depth maps.

Semi-automatic 2D-to-3D conversion of human-centered videos enhanced by age and gender estimation

“…In recent years, many 2D-to-3D video conversion techniques emerged to compensate for the lack of depth information in monocular video scenes. The most popular monocular depth cues used in such systems are motion vectors [4], vanishing point [5], edge information [6] and depth from focus/defocus [7]. Although fully automatic 2D-to-3D video conversion techniques can be considered as an option for extracting depth from a single view, issues such as temporal inconsistency, conflicting depth cues, depth map errors or motion estimation errors impair the perceived 3-D visual quality [8].…”