A novel 2D-to-3D scheme by visual attention and occlusion analysis

Abstract

We propose a novel 2D to 3D scheme by considering perceived 3D experience caused by occlusion and visual attention. In this scheme, initial depth model, the saliency map of visual attention and occlusion analysis are integrated in depth calculation. Mean-while, characteristics of human visual system are also considered as weight factors. Then, depth normalization and refining are implemented. The experimental results show that the final depth map presents proper depth order, details of objects and interesting region of 3D perception and the perceived 3D experiences for the converted images and videos are satisfied.