A Hole-Filling Method for DIBR Based on Convolutional Neural Network

There are holes for images generated by Depth-Image-Based Rendering (DIBR)method due to occlusion of foreground. An algorithm to fill holes based on convolutional neural networks is presented. PSNR of the image after filling holes is 32.65dB.

[1]  Gene Cheung,et al.  Disocclusion hole-filling in DIBR-synthesized images using multi-scale template matching , 2014, 2014 IEEE Visual Communications and Image Processing Conference.

[2]  Alexei A. Efros,et al.  Context Encoders: Feature Learning by Inpainting , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  C. Fehn.  A 3D-TV system based on video plus depth information , 2003, The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003.

[4]  John Flynn,et al.  Deep Stereo: Learning to Predict New Views from the World's Imagery , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

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