PCQM: A Full-Reference Quality Metric for Colored 3D Point Clouds

Abstract

3D point clouds constitute an emerging multimedia content, now used in a wide range of applications. The main drawback of this representation is the size of the data since typical point clouds may contain millions of points, usually associated with both geometry and color information. Consequently, a significant amount of work has been devoted to the efficient compression of this representation. Lossy compression leads to a degradation of the data and thus impacts the visual quality of the displayed content. In that context, predicting perceived visual quality computationally is essential for the optimization and evaluation of compression algorithms. In this paper, we introduce PCQM, a full-reference objective metric for visual quality assessment of 3D point clouds. The metric is an optimallyweighted linear combination of geometry-based and color-based features. We evaluate its performance on an open subjective dataset of colored point clouds compressed by several algorithms; the proposed quality assessment approach outperforms all previous metrics in terms of correlation with mean opinion scores.

Visual Quality of 3D Meshes With Diffuse Colors in Virtual Reality: Subjective and Objective Evaluation

“…For quality assessment of 3D colored point clouds, a data-driven metric (PCQM) has been recently introduced by Meynet et al [31]. Our metric considers the same initial collection of color and geometric features as [31]. Nevertheless, moving from point cloud domain to mesh domain implies major adaptations in the computation of these features.…”

“…To address the color-related aspects of our metric, we consider the features introduced in the 2D image-difference framework of Lissner et al [30]. Their color features have recently been used successfully for the quality assessment of colored 3D point clouds [31]. We refer to our metric as CMDM (for Color Mesh Distortion Measure).…”

“…However, choosing the best combination model is a crucial problem. For prediction of the color-image differences [30], the authors used a factorial combination model, while Meynet et al considered a linear model for their point cloud quality metric [31]. In our case, we chose to consider a linear model: (1) to make the optimization easier and (2) because we tried nonlinear models such as Minkowski pooling, which did not provide better performance.…”

A Reduced Reference Metric for Visual Quality Evaluation of Point Cloud Contents

“…A generalized Hausdorff distance has been proposed to improve the performance of geometry-based metrics [6], as well as a scale-invariant, point-to-distribution geometry metric based on Mahalanobis distance [7]. Recently, curvature statistics have also been proposed in order to estimate the distortion of a point cloud with respect to its reference [8], and they have been extended to include color information [9]. Viola et al incorporate color distortion in geometrybased metrics, using luminance histogram information [10], whereas Diniz et al use local binary pattern descriptors to estimate texture distortion [11].…”

Layered Projection-Based Quality Assessment of 3D Point Clouds

“…Most studies in the past had emphasized on the geometric distortion measurement of PC object, such as the point-to-point (po2point), point-toplane (po2plane), point-to-mesh (po2mesh) [18] and planeto-plane (pl2plane) [19]. Recently, curvature statistics which were initially introduced and applied on polygonal meshes [31] have also been proposed to estimate the distortion of a PC concerning its reference [20] and they have been extended to include color information [21]. Different from the above methods, Yang et al [22], Liu et al [23], Viola et al [24], and Alexiou et al [25] developed new research directions for PCQA from breakthroughs, such as deep learning, histogram, or SSIM [32] and so on.…”