Unilateral View Synthesis Prediction for Even View Rate Overhead Reduce in MVC

Abstract View synthesis is introduced into MVC for improving odd views’ coding performance. By synthesizing virtual odd views as reference views, view synthesis prediction (VSP) wins considerable coding gains for odd views. This kind bilateral view synthesis for odd views needs two adjacent views at both sides. Even views have only one reference view according to MVC coding order, and the bilateral view synthesis prediction technique is not suitable then. However, we may try using the only adjacent view to synthesize a virtual view and use it as one additional reference view to improve even views’ coding performance. In this paper, we first present a unilateral view synthesis method based on only one adjacent view. To raise the virtual view's quality, we modify our unilateral view synthesis method with Gaussian background modeling. Eventually, we take it into MVC and try to improve even views’ coding performance. Experimental results show that our scheme raises the even depth views’ coding quality by 0.6 to 1.2 dB, and even texture views less than 0.5 dB comparing with MVC.

[1]  Peter H. N. de With,et al.  Depth-Image Compression Based on an R-D Optimized Quadtree Decomposition for the Transmission of Multiview Images , 2007, 2007 IEEE International Conference on Image Processing.

[2]  P. KaewTrakulPong,et al.  An Improved Adaptive Background Mixture Model for Real-time Tracking with Shadow Detection , 2002 .

[3]  Aljoscha Smolic,et al.  Multi-View Video Plus Depth Representation and Coding , 2007, 2007 IEEE International Conference on Image Processing.

[4]  T. Wiegand,et al.  The Effect of Depth Compression on Multiview Rendering Quality , 2008, 2008 3DTV Conference: The True Vision - Capture, Transmission and Display of 3D Video.

[5]  Hai Tao,et al.  Compression and transmission of depth maps for image-based rendering , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[6]  Anthony Vetro,et al.  Extensions of H.264/AVC for Multiview Video Compression , 2006, 2006 International Conference on Image Processing.