DIBR synthesized image quality assessment based on morphological wavelets

Abstract

International audienceMost of the Depth Image Based Rendering (DIBR) techniques produce synthesized images which contain nonuniform geometric distortions affecting edges coherency. This type of distortions are challenging for common image quality metrics. Morphological filters maintain important geometric information such as edges across different resolution levels. In this paper, morphological wavelet peak signal-to-noise ratio measure, MW-PSNR, based on morphological wavelet decomposition is proposed to tackle the evaluation of DIBR synthesized images. It is shown that MW-PSNR achieves much higher correlation with human judgment compared to the state-of-the-art image quality measures in this context

Multi–Scale Synthesized View Assessment Based on Morphological Pyramids

“…Selected versions of the proposed metrics using morphological pyramid decompositions presented in Table 3 are: PSNR calculated on scale 5 of the MBP ED pyramid using SE of size 3 × 3 ; reduced and full versions of MP-PSNR using SE of size 5 × 5 in pyramid MBP ED decomposition. Selected versions of the metrics based on morphological wavelet decompositions [10] are: PSNR calculated on scale 6 between vertical [6] 0.5267 0.6411 0.5320 UQI [27] 0.5199 0.6529 0.5708 SSIM [28][29]…”

“…Introduced non-linear morphological filters maintain important geometric information such as edges across different resolution levels [9]. In the previous work [10], we explored morphological wavelet decompositions for the multiscale metric MW-PSNR, which achieves much higher correlation with human judgment compared to the state-of-the-art image quality measures. We also explored morphological pyramid decomposition in the multiscale metric MP-PSNR [11] which uses mean squared errors of all pyramids' sub-bands.…”

DIBR-Synthesized Image Quality Assessment via Statistics of Edge Intensity and Orientation

“…The weighted distortion maps were pooled to generate the quality score. In [5], the Morphological Wavelet Peak Signal-to-Noise Ratio (MW-PSNR) method was proposed. The multi-scale Mean Squared Error (MSE) was calculated based on the morphological wavelet decomposition.…”

“…The MSEs on all scales were pooled to produce the score. The reduced version of MW-PSNR, namely RMW-PSNR [5] was presented, which was obtained by only using MSEs on higher scales. In [6], MW-PSNR was improved by using morphological pyramids, producing the Morphological Pyramid Peak Signalto-Noise Ratio (MP-PSNR) method.…”

“…The reduced version of MW-PSNR, namely RMW-PSNR [5] was presented, which was obtained by only using MSEs on higher scales. In [6], MW-PSNR was improved by using morphological pyramids, producing the Morphological Pyramid Peak Signalto-Noise Ratio (MP-PSNR) method. The reduced version of MP-PSNR (RMP-PSNR) [7] was also presented, which only used MSEs on higher scales.…”