Author(s): |
Jian Qian, Faculty of Information Science and Engineering, Ningbo University, Ningbo, China Mei Yu, National Key Lab of Software New Technology, Nanjing University, Nanjing, China Feng Shao, Faculty of Information Science and Engineering, Ningbo University, Ningbo, China Zongju Peng, Faculty of Information Science and Engineering, Ningbo University, Ningbo, China Zhidi Jiang, Faculty of Information Science and Engineering, Ningbo University, Ningbo, China Fucui Li, Faculty of Information Science and Engineering, Ningbo University, Ningbo, China |
Abstract: |
Free viewpoint video system can satisfy the need of watching objects from any viewpoint, and it is the hot spots in video signal processing area. However, many current image-based rendering methods are often complex and inefficient, which is not suitable for fast rendering of the occasion required. So this article proposes a virtual viewpoint rendering method with assistant of encode-mode information from H.264/AVC compressed streams. Firstly, the information of the various modes is separated from the encode stream. Secondly, the image is divided into three regions, motion, static, and the boundary, through the analysis of these modes, and the different regions use different way to render. Finally, it is necessary to process post-processing for rendered image to solve the edge noise result from inaccurate depth. The experiment show that this proposed approach effectively improves the speed of the virtual viewpoint rendering with the close Power Signal-to-Noise Ratio (PSNR), compared with traditional methods, and saving of about 35% to 50% of the rendering time.
|