您的位置：山东大学 -> 科技期刊社 -> 《山东大学学报（理学版）》

山东大学学报（理学版） ›› 2016, Vol. 51 ›› Issue (3): 122-131.doi: 10.6040/j.issn.1671-9352.0.2015.431

• • 上一篇    下一篇

一种基于两视点DIBR的改进虚拟视点合成方法

王路1,赖春露2*   

  1. 1.山东大学(威海)网络与信息管理中心, 山东 威海 264209;2.哈尔滨工业大学(威海)信息与电气工程学院, 山东 威海 264209
  • 收稿日期:2015-09-04 出版日期:2016-03-20 发布日期:2016-04-07
  • 通讯作者: 赖春露(1989— ),女,助理工程师,硕士研究生,研究方向为通信与信息处理. E-mail:laicl_ruby@hitwh.edu.cn E-mail:wanglu@sdu.edu.cn
  • 作者简介:王路(1988— ),男,助理工程师,硕士研究生,研究方向为通信与信息处理. E-mail:wanglu@sdu.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(61573123,61304142)

A modifiedtwo-viewpoint DIBR based virtual view synthesis method

WANG Lu1, LAI Chun-lu2*   

  1. 1. Network and information management center, Shandong University(Weihai), Weihai 264209, Shandong, China;
    2. School of Information and Electrical Engineering, Harbin Institute of Technology at Weihai, Weihai 264209, Shandong, China
  • Received:2015-09-04 Online:2016-03-20 Published:2016-04-07

PDF (PC)

739

摘要： 虚拟视点合成是裸眼3D电视的关键技术,而基于深度的图像渲染(depth-image-based rendering,DIBR)则是一种经典的虚拟视点合成方法。为解决经典DIBR方法存在的空洞问题,提高虚拟视点图像质量,提出了一种基于DIBR的改进虚拟视点合成方法。首先,只在深度边缘处进行深度图的预处理,这可减少空洞数量且不产生大的几何失真;其次,为得到既没有伪影也没有大空洞的虚拟视点图像,提出了基于距离与深度的图像融合方法,并在融合左右两侧参考视点目标图像前,对其空洞掩膜图进行膨胀;最后,利用逐层收缩的融合后空洞填充方法对剩余小空洞进行迭代填充。实验的主观质量与客观评价都显示本文方法能够取得令人满意的结果。

关键词: 虚拟视点, 图像融合, 图像渲染

Abstract: The depth-image-based rendering(DIBR)is a classical virtual view synthesis technique which has been regarded as one of the most crucial techniques for the naked eye three-dimensional television. In order to fill the holes existed in the virtual images obtained by DIBR and thus to improve the quality of the images, a modified two-viewpoint DIBR based virtual view synthesis method is wasproposed. To prevent the image from severe geometry distortions while the number of holes being reduced, only those areas of the depth image with sudden changes on the depth value were preprocessed. In order to obtain virtual images without ghosting errors and large holes, the two virtual images generated from different viewpoints were dilated at first and then merged into one using a distance-and-depth based image blending algorithm proposed in this paper. To fill the residual holes after the image fusion, a shell-by-shell filling method was also presented.Experimental results show that the proposed method can obtain satisfactory performance in both subjective quality and objective evaluation.

Key words: virtual view, image blending, image rendering

中图分类号: 

  • TN911.73
[1] TANIMOTO M, TEHRANI M P, FUJII T, et al. Free-viewpoint TV[J]. IEEE Signal Processing Magazine, 2011, 28(1):67-76.
[2] FEHN C. Depth-image-based rendering(DIBR), compression and transmission for a new approach on 3D-TV[J]. SPIE Stereoscopic Displays and Virtual Reality Systems XI, 2004, 5291:93-104.
[3] AHN I, KIM C. A novel depth-based virtual view synthesis method for free viewpoint video[J]. IEEE Transactions on Broadcasting, 2013, 59(59):614-626.
[4] ZHU Shiping, LI Zheng, YU Yang. Virtual view synthesis using stereo vision based on the sum of absolute difference[J]. Computers and Electrical Engineering, 2014, 40(8):236-246.
[5] YANG Xiaohui, LIU Ju, SUN Jiande, et al. DIBR based view synthesis for free-viewpoint television[C] // Proceedings of 3DTV Conference: the True Vision-capture.Washington: IEEE, 2011, 47(10):1-4.
[6] KRISHNAMURTHY S, RAMAKRISHNAN K R. Virtual view synthesis by Non-local means filtering using temporal data[C] // Proceedings of 3dtv-Conference: the True Vision-capture. Washington: IEEE, 2014: 1-4
[7] KIM B, HONG M. An adaptive interpolation algorithm for hole-filling in free viewpoint video[J]. Journal of Measurement Science and Instrumentation, 2013, 4(4):343-345.
[8] KWAN-JUNG O, SEHOON Y, ANTHONY V, et al. Virtual view synthesis method and self-evaluation metrics for viewpoint television and 3D video[J]. International Journal of Imaging Systems and Technology, 2010, 20(4):378-390.
[9] DARIBO I, TILLIER C, PESQUET-POPESCU B. Distance dependent depth filtering in 3D warping for 3DTV[C] // Proceedings of IEEE 9th International Workshop on Multimedia Signal Processing(MMSP 2007). Washington: IEEE, 2007: 312-315.
[10] QIAO Yiguo, JUNG C. Dictionary based hole filling with assistance of depth[C] // Proceedings of IEEE International Conference on Multimedia and Expo(ICME 2014). Washington: IEEE, 2014.
[11] HABIGT J, DIEPOLD K. Image completion for view synthesis using Markovrandom fields and efficient belief propagation[C] // Proceedings of IEEE International Conference on Image Processing(ICIP 2013). Washington: IEEE, 2013: 2131-2134.
[12] ZHONG Lin, PARK M, METAXAS D. Automatic stereoscopic video generation based on virtual viewsynthesis[J]. Neurocomputing, 2015, 151(P2):654-662.
[13] ZITNICK C L, KANG S B, UYTTENDAELE M, et al. High-quality video view interpolation using a layered representation[J]. ACM SIGGRAPH and ACM Transcations on Graphics, 2004, 23(4):600-608.
[14] WANG Lu, LIU Ju, SUN Jiande, et al. Virtual view synthesis without preprocessing depth image for depth image based rendering[C] // Proceedings of 3DTV-Conference: the True Vision-Capture, Transmission and Display of 3D Video(3DTV-CON 2011).Washington: IEEE, 2011: 1-4.
[15] GREENE N, KASS M, MILLER G. Hierarchical Z-buffer visibility[C] // Proceedings of the ACM SIGGRAPH’93 Conference on Computer Graphics. New York: ACM, 1993: 231-238.
[16] GONZALEZ R C, WOODS R E. Digital image processing[M]. 2nd ed. USA: Prentice Hall, 2002.
[17] WANG Z, BOVIK A C, SHEIKH H R, et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Transactions on Image Processing, 2004, 13(4):600-612.
[18] PINSON M H, WOLF S. A new standardized method for objectively measuring video quality[J]. IEEE Transactions on Broadcasting, 2004, 50(3):312-322.
[1] 方志军,刘心韵,伍世虔,郑文娟. 基于子带加权融合的多尺度
Retinex图像增强算法[J]. J4, 2013, 48(11): 93-98.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2018 《山东大学学报(理学版)》编辑部　
地址: 山东省济南市山大南路27号山东大学中心校区（250100）　电话: +86-0531-88366917　E-mail: xblxb@sdu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发