Depth coding using coded boundary patterns

The depth information plays an essential role in the virtual-view (or free-viewpoint) 3D video systems. In this paper, we propose a new algorithm to code a depth map for the purpose of virtual view synthesis. The idea is to use H.264/AVC to represent the rough shape (including depth values) of a depth map and then additional information is transmitted to improve the depth values around the object boundaries. The complete encoding and decoding simulation system was built on the H.264/AVC JM 18.0 platform. In our experiments, three tools can be turned on individually and thus four coding modes are defined and tested. Our data show that these proposed tools offer advantages in either coding efficiency or image quality improvement and some tools work best on simple images while the others work best on complex images. With proper parameter setting, the overall quality of virtual view rendering is noticeably improved.