Semi-Global Depth Up-sampling for 3D Video Acquisition

In this paper, we propose a new method to upsample a depth map for 3D video acquisition. The proposed method is based on Markov random fields. The low-resolution depth values are used for data term and the high-resolution color data are exploited for smoothness term. Especially, we apply semi-global energy minimization to reduce computational time. Experimental results show that the proposed method outperforms conventional methods. Keywords—semi-global depth up-sampling, 3D video, depthimage-based rendering.