|
Since the texture of a depth map is easier than that of its corresponding color image, the prediction mode which is used to predict complex texture for color image is less used in depth maps, and could be discarded during the intra coding of depth maps so as to reduce coding complexity. Experimental results demonstrate that not only the coding complexity is reduced, but the coding efficiency is also improved. Compared with H.264, the average coding time reduction is 7.18%, while the maximum and average bit rate savings of depth maps can be achieved as high as -14.79% and -5.92%, respectively, while maintaining the same quality of synthesized virtual views. |
|
Keywords:depth map; 3D video coding; view synthesis |
|