提出了一种深度图失真引入的虚拟视失真估计方法. 该方法依据能量一致性假设要求, 将深度图宏块划分为平坦宏块和非平坦宏块. 平坦宏块使用频域块级方法, 非平坦宏块使用时域像素级方法估计由深度图失真造成的绘制视失真. 虚拟视失真采用左右绘制视融合计算, 并计入虚拟视中空洞修复引入的失真. 该方法在宏块分类判据中综合了相机参数, 对于不同拍摄条件下的视频序列可以使用相同的判别阈值获得最佳区域划分, 通用性较强. 实验结果证明, 该方法兼顾了估计准确性和复杂度, 对于在有限码率预算条件下提高深度视频编码效率、优化三维视频码率分配具有重要的指导意义.
A method for estimating the virtual view image distortion caused by the depth map coding distortion is proposed. The depth map is divided into macro blocks according to the coding unit. These macroblocks are then classified into uniform macroblocks and nonuniform macroblocks according to the energy density consistency constraint. The rendered view distortion in the uniform macroblocks is estimated by a block-level frequency domain method, while that in the non-uniform macroblocks by a pixel-level time domain method. Finally, the virtual view distortion is combined with both rendered view distortions. The distortion introduced by inpainting is also computed. With a shooting parameter, it can correctly classify the macroblocks of depth map sequences acquired from different shooting conditions. Simulation results show that the proposed method can accurately estimate virtual view distortion caused by depth map distortion. It achieves effective balance between accuracy and algorithm complexity, which is important to guide 3D video (3DV) coding and rate allocation under the limited bit rate budget.
[1] 张兆扬, 安平, 张之江, 等. 二维和三维视频处理及立体显示技术[M]. 北京: 科学出版社, 2010: 1-100.
[2] MPEG. Applications and requirements on 3D video coding [C]//ISO/IEC JTC1/SC29/WG11. 2009: 13-29.
[3] Fehn C. Depth-image-based rendering (DIBR), compression and transmission for a new approach on 3D-TV [C]//Proceedings of SPIE Stereoscopic Displays and Virtual Reality Systems Ⅺ. 2004: 93-104.
[4] Rusanovskyy D, Zhang L, Chen F C, et al. 3D-AVC test model 8 [C]//Joint Collaborative Team on 3D Video Coding Extension Development of ITU-T SG 16 WP 3 and ISO/IEC JTC
1/SC 29/WG 11 JCT3V-F1003. 2013: 1-62.
[5] Zhang L, Tech G, Wegner K, et al. Test model 7 of 3D-HEVC and MV-HEVC [C]//Joint Collaborative Team on 3D Video Coding Extension Development of ITU-T SG 16 WP 3 and
ISO/IEC JTC 1/SC 29/WG 11 JCT3V-G1005. 2014: 3-40.
[6] Liu Y W, Huang Q M, Ma S W, et al. Joint video/depth rate allocation for 3D video coding based on view synthesis distortion model [J]. Signal Proc Image Comm, 2009, 24(8): 666-681.
[7] Tech G, Schwerz H, Muller K, et al. 3D video coding using the synthesized view distortion change [C]//PCS. 2012: 25-28.
[8] Merkle P, Morvan Y, Smolic A, et al. The effects of multiview depth video compression on multiview rendering [J]. Singal Proc Image Comm, 2009, 24(1/2): 73-88.
[9] Lai P, Ortega A, Dore A, et al. Improving view rendering quality and coding efficiency by suppressing compression artifacts in depth-image coding [C]//Proc of Visual Communic and
Image. 2009: 894-897.
[10] Wang Q F, Ji X Y, Dai Q H, et al. Free viewpoint video coding with rate-distortion analysis [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2012, 22(6):
875-889.
[11] Kim W S, Ortega A, Lai P, et al. Depth map distortion analysis for view rendering and depth coding [C]//ICIP. 2009: 721-724.
[12] Kim W S, Ortega A, Lai P, et al. Depth map coding with distortion estimation of rendered view [C]//Proc of SPIE Visual Information Processing and Communication. 2010: 3534-3545.
[13] Oh B T, Lee J, Park D. Depth map coding based on synthesized view distortion function [J]. IEEE Journal of Selected Topics in Signal Processing, 2011, 5(7): 1344-1352.
[14] Oh B T, Lee J, Dusik P. Fast joint bit-allocation between texture and depth maps for 3D video coding [C]. ICCE. 2013: 193-194.
