Point clouds are one of the most promising technologies for 3D content representation. However, the subjective and objective quality assessment of this type of imaging still remains an open problem. In this experiment, we conduct subjective quality assessment of geometry-only point clouds subject to octree-based compression at different quality levels. The test contents were displayed using Screened Poisson surface reconstruction, without including any textural information, and they were rated by subjects in a passive way using 2D animated video sequences. The subjective evaluations were conducted in five independent laboratories in different countries. In , we provide the results of our analysis, including inter-laboratory comparisons, benchmarking of state-of-the-art objective quality metrics, and correlation results with subjective scores obtained from another test involving visualization of raw point clouds.
In the provided URL link, you may find:
- The raw subjective scores that were obtained from every laboratory.
- The complete set of correlation results for every pair combination of laboratories.
- Benchmarking results between subjective scores from every lab and state-of-the-art objective metrics.
- Correlation coefficients for the two types of content visualization examined (i.e., raw point clouds against reconstructed mesh objects using Screened Poisson).
 E. Alexiou, M. Bernardo, L. S. Cruz, L. G. Dmitrovic, R. Duarte, E. Dumic, T. Ebrahimi, D. Matkovic, M. Pereira, A. Pinheiro and A. Skodras, “Point cloud subjective evaluation methodology based on 2D rendering,” 10th International Conference on Quality of Multimedia Experience (QoMEX), Sardinia, Italy, 2018