Hole filling using multiple frames and iterative texture synthesis with illumination compensation

In this paper, we propose a novel freeview generation system. We first discuss the problem of misalignment of the edges between color frame and depth frame, and propose an optimization method to solve it. Then an original frame is separated into two parts, depending on the sizes of the resulting dis-occluded areas in the virtual view. Illumination changes between current frame and the information from previous frames are modeled and compensated. In the virtual view, the missing pixels are recovered by the proposed inpainting method. Using depth information, we compute the priority of the recovering order among the missing pixels. The recovering process accounts for pixel consistency, and the process iterates between virtual color frame and depth frame. Experimental results show that, compared with a state-of-the-art work, the proposed method has better subjective and objective performances; the PSNR gain can be up to 2.9575 dB.

[1]  Pau Gargallo,et al.  Stereoscopic Image Inpainting: Distinct Depth Maps and Images Inpainting , 2010, 2010 20th International Conference on Pattern Recognition.

[2]  Thomas Wiegand,et al.  3D Video and Free Viewpoint Video - Technologies, Applications and MPEG Standards , 2006, 2006 IEEE International Conference on Multimedia and Expo.

[3]  Christoph Fehn,et al.  Depth-image-based rendering (DIBR), compression, and transmission for a new approach on 3D-TV , 2004, IS&T/SPIE Electronic Imaging.

[4]  Guillermo Sapiro,et al.  Image inpainting , 2000, SIGGRAPH.

[5]  Michael Schmeing,et al.  Depth Image Based Rendering: A faithful approach for the disocclusion problem , 2010, 2010 3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video.

[6]  Aljoscha Smolic,et al.  Intermediate view interpolation based on multiview video plus depth for advanced 3D video systems , 2008, 2008 15th IEEE International Conference on Image Processing.

[7]  Patrick Pérez,et al.  Region filling and object removal by exemplar-based image inpainting , 2004, IEEE Transactions on Image Processing.

[8]  Ghassan Al-Regib,et al.  Hierarchical Hole-Filling For Depth-Based View Synthesis in FTV and 3D Video , 2012, IEEE Journal of Selected Topics in Signal Processing.

[9]  Ruigang Yang,et al.  Stereoscopic inpainting: Joint color and depth completion from stereo images , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[10]  Liang Zhang,et al.  Stereoscopic image generation based on depth images for 3D TV , 2005, IEEE Transactions on Broadcasting.

[11]  Dong Tian,et al.  View synthesis techniques for 3D video , 2009, Optical Engineering + Applications.

[12]  N. Atzpadin,et al.  Depth map creation and image-based rendering for advanced 3DTV services providing interoperability and scalability , 2007, Signal Process. Image Commun..