IEEE Consumer Electronics Magazine - July 2019 - 21

plane is chosen as the projection plane using orthographic
projection, neighboring points in the xy plane can be grouped
as a patch. In such a case, a patch may be represented by two
different images: a geometry image and a texture image. A
patch in a geometry image contains the z coordinate information where xy coordinate information can be derived from the
patch formation. In the same way, a patch in a texture image
contains the color information. Therefore, the geometric
image may be represented by one color component (gray
scale) and the texture image by three components.
A number of patches may be generated and packed into one
geometry image and one texture image for each frame. Geometry images and texture images form two separate video frames
and are input to video compression (i.e., the reference software
of high efficiency video coding TM 16.16). Once patches for
geometry and texture are generated and encoded, some auxiliary information must be encoded as well, mainly consisting of
how to put the patches back to the original point-cloud structure. For this purpose, additional data called auxiliary patch
information (API) and an occupancy map (OM) are encoded in
parallel. In each generated image for geometry or texture, there
are pixels belonging to patches and those that are not.
To distinguish the pixels in the given image, an OM with
binary data associated to the given block size indicates if the
current block belongs to a patch. If the block size becomes
one, it is possible to encode the point-cloud as lossless (at
least to preserve the number of input points after encoding).
An OM corresponds to each set of geometry and texture
images in a frame. Then, API is necessary to group patch
blocks in the occupancy map. An OM and API are separately
compressed using simple arithmetic coding with run-length
coding. This completes the encoding process.

REMAINING TECHNICAL ISSUES UNTIL 2020
It is a bit early to predict the exact shape of PCC in 2020, the
date when the international standard (IS) will be published.
However, it is possible to discuss what is ahead on the way to
the IS. Currently, for each of the three PCC categories, there is a
TM: two geometry-based compression approaches and one
3D-to-2D mapping-based video compression approach. The
work to harmonize TMC1 and TMC3 has already begun, which
may eventually lead to only two TMs in the final standard. It is
also notable that many functional blocks, such as arithmetic
coding, octree decomposition, and color conversion, are the
common tools in the three TMs. Therefore, it is also expected
that the common functional blocks will be unified across TMs.
The CfP for PCC addressed a typical bit-rate range for PCC
applications to be 5-40 megabits/s. The current TMs are
expected to meet the goal in compression efficiency with high
probability. At every MPEG meeting (there are four meetings
each year), core experiments (CEs) are designed and conducted
to upgrade TMs, which include new technologies based on the
CEs. The CEs are a working mechanism to ensure that all of the
requirements to address in MPEG PCC are supported by TMs.
While PCC is one important technology being discussed by
MPEG, MPEG is formulating another important one called

MPEG Immersive (MPEG-I) [10]. In fact, PCC is a part of the
MPEG-I standard. MPEG-I is meant to develop standards to represent, compress, and deliver next-generation immersive media.
Point-cloud representation is considered one of the immersive
media formats, and it is covered by part 5 of the MPEG-I standard. Additionally, MPEG-I will provide technologies to deal with
360° video compression, immersive audio, and immersive media
delivery to enable fully immersive media experiences to users.

CONCLUSION
In this article, we briefly reported the initiation of MPEG PCC
standardization by introducing the requirements, the applications, and the main technical concepts of three TMs. This is a
new kind of standard in which video and graphics compression
tools mingle and possibly produce good compression efficiency
with many new features (e.g., view dependency, lossless compression, and various scalabilities). Several remaining technical
issues need to be addressed. Our future research will focus on
the development of those technologies. Our research will particularly concentrate on how to utilize the efficient video coding
technology to compress 3D point-cloud data.

ABOUT THE AUTHORS
Li Cui (lcui2000@gmail.com) is working toward her Ph.D.
degree at Hanyang University, South Korea.
Rufael Mekuria (rufael@unified-streaming.com) is a staff
research engineer at Unified Streaming, The Netherlands.
Marius Preda (marius.preda@it-sudparis.eu) is an associate professor at Institut Mines-Télécom, France, and the chair
of Moving Picture Experts Group's 3D Graphics group.
Euee S. Jang (esjang@hanyang.ac.kr) is a professor in
the Department of Computer Science and Engineering at
Hanyang University, South Korea.

REFERENCES
[1] O. Devillers and P.-M. Gandoin, "Geometric compression for interactive transmission," in Proc. IEEE Visualization 2000, pp. 319-326.
[2] Information Technology-Coding of Audio-Visual Objects-Part 2:
Visual, ISO/IEC 14496-2, 1998.
[3] M. Bourges-Sévenier and E. S. Jang, "An introduction to the MPEG4 Animation Framework Extension," IEEE Trans. Circuits Syst. Video
Technol., vol. 14, no. 7, pp. 928-936, 2004. doi: 10.1109/TCSVT.2004.
830662.
[4] C. Loop, Q. Cai, S. O. Escolano, and P. A. Chou, "Microsoft voxelized upper bodies: A voxelized point cloud dataset," ISO/IEC JTC1/
SC29/WG11, Rep. M72012, Geneva, Switzerland, May 2016.
[5] "Print cloud compression requirements," ISO/IEC JTC1/SC29/
WG11, Gwangju, South Korea, Doc. W17353, Jan. 2018.
[6] "PCC test model category 1 v1," ISO/IEC JTC1/SC29/WG11,
Gwangju, South Korea, Doc. W17340, Jan. 2018.
[7] "PCC test model category 2 v1," ISO/IEC JTC1/SC29/WG11,
Gwangju, South Korea, Doc. W17348, Jan. 2018.
[8] "PCC test model category 3 v1," ISO/IEC JTC1/SC29/WG11,
Gwangju, South Korea, Doc. W17349, Jan. 2018.
[9] R. L. de Queiroz and P. A. Chou, "Compression of 3D point clouds
using a region-adaptive hierarchical transform," IEEE Trans. Image Process., vol. 25, no. 8, pp. 3947-3956, 2016. doi: 10.1109/TIP.2016.2575005.
[10] "MPEG-I technical report on immersive media," ISO/IEC JTC1/
SC29/WG11, Torino, Italy, Doc. W17069, July 2017.

JULY 2019

IEEE CONSUMER ELECTRONICS MAGAZINE

IEEE Consumer Electronics Magazine - July 2019

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - July 2019