Quality Evaluation Of Static Point Clouds Encoded Using MPEG Codecs

Perry, Stuart; Cong, Huy Phi; Cruz, Luís A. da Silva; Prazeres, Joao; Pereira, Manuela; Pinheiro, António; Dumic, Emil; Alexiou, Evangelos; Ebrahimi, Touradj

doi:10.1109/icip40778.2020.9191308

Cited by 73 publications

(63 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ICIP2020 Database [48]: This database has six contents, including both human bodies and inanimate objects. It includes 112 test PCs, where each PC ref is processed into three different distortion types at six levels.…”

Section: Resultsmentioning

confidence: 99%

Layered Projection-Based Quality Assessment of 3D Point Clouds

Chen

Long

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Point clouds are subject to various distortions during point cloud processing missions, any of which may lead to quality degradation. Consequently, predicting point cloud quality has attracted a lot of attention. In this paper, a layered projection-based point cloud quality metric (LP-PCQM) is proposed. We layer the distorted point cloud and its original version firstly and then extract the geometry and color features of layers. The geometry feature is obtained using the projection-based method and the color features are extracted upon RGB by using the point-based method. Finally, the LP-PCQM is a weighted linear combination of an optimal subset of these pooled geometry and color features of layers. To verify the performance of LP-PCQM, we compare it with other eight metrics including both point-based metrics and projection-based metrics on the WPC, SJTU-PCQA, and ICIP2020 database respectively. Experimental results show that the proposed metric exhibits better and more robust performance.

show abstract

Section: Resultsmentioning

confidence: 99%

Layered Projection-Based Quality Assessment of 3D Point Clouds

Chen

Long

et al. 2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The proposed method is evaluated using two recent 3D point cloud datasets: sjtu [26] and ICIP20 [27].…”

Section: Datasets and Experimental Setupmentioning

confidence: 99%

Deep Learning-Based Quality Assessment Of 3d Point Clouds Without Reference

Chetouani

Quach

Valenzise

et al. 2021

2021 IEEE International Conference on Multimedia &Amp; Expo Workshops (ICMEW)

View full text Add to dashboard Cite

Point cloud (PC) quality assessment is of fundamental importance to enable the efficient processing, coding and transmission of 3D data for virtual/augmented reality, autonomous driving, cultural heritage, etc. The quality metrics proposed so far aim at quantifying the distortion in the PC geometry and/or attributes with respect to a reference pristine point cloud, using simple features extracted by the points. In this work, we target instead a blind (no-reference) scenario in which the original point cloud is not available. In addition, we learn features from data using deep neural networks. Given the limited availability of subjectively annotated datasets of corrupted point clouds, and the consequent difficulty to learn in an end-to-end fashion PC quality features, in this work we use instead a two-step procedure. First, we extract from local patches three relevant low-level features which have been commonly used in other PC quality metrics, i.e., geometric distance, local curvature and luminance values. Afterwards, we employ a deep neural network to learn, from these lowlevel features, a mapping to the PC ground truth mean opinion score. Our results on two state-of-the-art PC quality datasets show the potential of the proposed approach. The code is

show abstract

“…Furthermore, we use the point-to-plane (p2plane) metric with the mean squared error (MSE) as distance measure to quantify the quality of the compressed point cloud. This quality measurement method has been found to be the most accurate one when compared to the point-to-point metric and Hausdorff distance measure [8]. We calculate the peak-signal-to-noise ratio (PSNR) of the MSE as described in [11] for a better comparability between the different point clouds.…”

Section: A Evaluation Of Different Update Intervalsmentioning

confidence: 99%

On the Potential of Rate Adaptive Point Cloud Streaming on the Point Level

Laniewski

Aschenbruck

2021

2021 IEEE 46th Conference on Local Computer Networks (LCN)

View full text Add to dashboard Cite

In many robotic applications as well as in the area of immersive multimedia, large point clouds need to be transmitted over a network preferably in real-time. Typically, the point cloud is first compressed to a significantly smaller size, before it is transmitted as a simple file transfer. A drawback of this method is that it requires a point cloud to be fully available for the compression process to begin, inducing additional delay. In this paper, we explore the potential of streaming incoming points directly from an ongoing laser scan. Therefore, we propose a compression algorithm that operates on an ongoing point stream. Furthermore, we propose and compare different rate adaptation methods that dynamically adapt the compression rate to the currently available network data rate. Our evaluations show promising results in terms of rate fluctuations around the available network data rate and the achieved quality of our solution.

show abstract

Quality Evaluation Of Static Point Clouds Encoded Using MPEG Codecs

Cited by 73 publications

References 13 publications

Layered Projection-Based Quality Assessment of 3D Point Clouds

Layered Projection-Based Quality Assessment of 3D Point Clouds

Deep Learning-Based Quality Assessment Of 3d Point Clouds Without Reference

On the Potential of Rate Adaptive Point Cloud Streaming on the Point Level

Contact Info

Product

Resources

About