Fig 1.
Flow for the proposed Directional Filter Bank Depth Image based Rendering System (DFB-DIBR).
Fig 2.
(a) Original image (b) Dehazed image.
Fig 3.
2D frequency partitioning.
Fig 4.
(a) Noisy image (b) Noise free image.
Fig 5.
Pixels intensity histograms.
Fig 6.
Image Segmentation.
Fig 7.
Image profile/depth hypothesis determination.
Fig 8.
Image profiles/hypothesis generation base on vp (.a, .b, .c, .d, .e, .f, .g, .h, .i, .j).
Fig 9.
Refined depth map Dmap.
Fig 10.
Depth map results of dataset [37].
(a) Input Images (b) Depthmap produced by Zhuo.et al. [12] (c) Depthmap produced by Yang.et.al. [26] and (d) Depth-map generated by DFB-DIBR.
Fig 11.
Anaglyph results of dataset [37].
(a) Input Images (b) Anaglyph produced by Zhuo.et al. [12] (c) Anaglyph produced by Yang.et.al. [26] and (d) Anaglyph generated by DFB-DIBR.
Fig 12.
MAE and RMSE of the dataset [37].
Table 1.
Comparison parameters PSNR, SSIM and UQI using dataset [37].
Fig 13.
Depth map results of dataset [38].
(a) Input Images (b) Depthmap produced by Zhuo.et al. [12] (c) Depthmap produced by Yang.et.al. [26] and (d) Depth-map generated by DFB-DIBR.
Fig 14.
Anaglyph results of dataset [38].
(a) Input Images (b) Anaglyph produced by Zhuo.et al. [12] (c) Anaglyph produced by Yang.et.al. [26] and (d) Anaglyph generated by DFB-DIBR.
Fig 15.
MAE and RMSE of the dataset [38].
Table 2.
Comparison parameters PSNR, SSIM and UQI using dataset [38].
Table 3.
Comparison parameters PSNR, SSIM and VIF using dataset [39].
Fig 16.
Holes percentage of occluded area.
Table 4.
Holes percentage using block size 8x8 and block size 16x16 tested dataset [37].
Table 5.
Holes percentage using block size 8x8 and block size 16x16 tested dataset [38].