Free viewpoint video quality assessment based on morphological multiscale metrics

Abstract

In this paper, two image quality metrics based on morphological multiscale decompositions have been applied for the evaluation of free viewpoint video sequences. These sequences are synthesized using decompressed depth maps in the Depth-Image-Based Rendering synthesis process. Since the synthesis introduces edge distortion in the synthesized image/video, morphological filters are used for their ability to maintain important geometric information (i.e., edges) across different resolution levels. The edges displacement in different resolution scales is evaluated by means of the Mean Square Error. The adopted metrics show higher correlation with human judgment than state-of-the-art image quality measures used in this context

From pyramids to state‐of‐the‐art: a study and comprehensive comparison of visible–infrared image fusion techniques

“…Constructed in two ways as morphological low‐pass pyramid and morphological band‐pass pyramid, it uses morphological filters to decompose the image. Most of the pyramid schemes are based on the usage of linear filters but research has been done in [7, 11] to investigate the role and scope of non‐linear filtering in pyramid construction. It has also been suggested that linear filter‐based pyramid construction techniques manipulate and change the intensities of the objects hence, consequently the location of contours and edges also changes.…”

“…Erosion removes comparatively brighter regions, which are smaller than the considered structuring element. The dilation operation afterwards ensures that the original size of the brighter regions greater than the structuring element is retained [11]. A general morphological pyramid‐based decomposition process is shown in Fig.…”

Investigating Epipolar Plane Image Representations for Objective Quality Evaluation of Light Field Images

“…Some of them consider common distortions [5] such as gaussian blur, additive noise, etc. Some publicly available datasets consider LF specific distortions [6] [7][8] such as compression, view synthesis related distortions. There is still a need for datasets who offers acquisition and display related distortions with the corresponding subjective scores.…”

“…In Free Viewpoint Video dataset(FVV) [6], seven different compression algorithms have been used to code depth maps. Coded depth maps are used to synthesize intermediate views to generate horizontal parallax videos with 50 frames.…”

“…In table 2, Pearson correlation coefficients with subjective scores are presented for MW-PSNR [12], GMSD [16] and NICE [11] metrics on FVV dataset [6]. All metrics have been tested on the same dataset with both LF representations.…”