International Journal of Engineering Trends and Technology (IJETT) – Volume 13 Number 7 – Jul 2014
Video Compression Techniques: A Review
Ajay#1,Mahesh Prasad Parsai*2
#
Electronics and Telecommunication,
Jabalpur Engineering College, Jabalpur
Madhya Pradesh (India), 482002
Abstract— This article presents, the review of various
compression algorithms used for the video compression. The
video compression has become a basic requirement in the
modern digital scenario. The video compression is started with
the Motion JPEG and Motion JPEG2000. In which the each
frame is consider as a picture for the JPEG and JPEG 2000
compression respectively. But these techniques do not use the
video compression at all. Video signal has high temporal
redundancies due to the high correlation between successive
frames. Actually, this redundancy has not been exploited enough
by these video compression techniques. So the MPEG is exploited
as a solution to this problem. But the MPEG is slow due to 3D
transformations of the video blocks. So the Accordion based
video compression is found to be a better idea.
Keywords— Video Compression Techniques.
I. INTRODUCTION
Moving digital video images have been integrated with
programs. The difficulty in implementing moving digital
video is the tremendous bandwidth required for the encoding
of video data. For example, a quarter screen image (320 x 240
pixels) playing on an RGB video screen at full speed of 30 fps
requires storage and transmission of 6.9 million bytes per
second (MB/s). This data rate is simply prohibitive, and so
means of compressing digital video suitable for real-time
playback are a necessary step for the widespread introduction
of digital motion video applications[1-3].
Many digital video compression algorithms have
been developed and implemented. The compression ratios of
these algorithms vary according to the subjective acceptable
level of error, the definition of the word compression, and
who is making the claim[4-5].
An ideal video compression technique should have the
following characteristics:
 Will produce levels of compression rivaling MPEG
without objectionable artifacts.
 Can be played back in real time with inexpensive
hardware support.
 Can degrade easily under network overload or on a
slow platform.
 Can be compressed in real time with inexpensive
hardware support.
Video compression techniques made feasible a
number of applications[6-9]. Four distinct applications of the
compressed video can be summarized as: (a) consumer
ISSN: 2231-5381
broadcast television, (b) consumer playback, (c) desktop video,
and (d) videoconferencing.
Consumer broadcast television: which includes digital video
delivery to homes, typically requires a small number of highquality compressors and a large number of low-cost
decompressors. Expected compression ratio is about 50:1.
Consumer playback applications: such as CD-ROM libraries
and interactive games, also require a small number of
compressors and a large number of low-cost decompressors.
The required compression ratio is about 100:1.
Desktop video: which includes systems for authoring and
editing video presentations, is a symmetrical application
requiring the same number of encoders and decoders. The
expected compression ratio is in the range from 5:1 to 50:1.
Videoconferencing applications: also require the same
number of encoders and decoders, and the expected
compression ratio is about 100:1.
II. JPEG AND MOTION JPEG
The JPEG standard, ISO/IEC 10918, is the single most
widespread picture compression format of today. It offers the
flexibility to either select high picture quality with fairly high
compression ratio or to get a very high compression ratio at
the expense of a reasonable lower picture quality. Systems,
such as cameras and viewers, can be made inexpensive due to
the low complexity of the technique[2,7-9].
JPEG image compression contains a series of
advanced techniques. The main one that does the real image
compression is the Discrete Cosine Transform (DCT)
followed by a quantization that removes the redundant
information (the “invisible” parts).
A digital video sequence can be represented as a
series of JPEG pictures. The advantages are the same as with
single still JPEG pictures – flexibility both in terms of quality
and compression ratio. The main disadvantage of Motion
JPEG (a.k.a. MJPEG) is that since it uses only a series of still
pictures it makes no use of video compression techniques. The
result is a lower compression ratio for video sequences
compared to “real” video compression techniques like MPEG.
The benefit is its robustness with no dependency between the
frames, which means that, for example, even if one frame is
dropped during transfer, the rest of the video will be unaffected.
III. JPEG 2000 AND MOTION JPEG 2000
JPEG 2000 was created as the follow-up to the successful
JPEG compression, with better compression ratios. The basis
http://www.ijettjournal.org
Page 318
International Journal of Engineering Trends and Technology (IJETT) – Volume 13 Number 7 – Jul 2014
was to incorporate new advances in picture compression
research into an international standard. Instead of the DCT
transformation, JPEG 2000, ISO/IEC 15444, uses the Wavelet
transformation. The advantage of JPEG 2000 is that the
blockiness of JPEG is removed, but replaced with a more
overall fuzzy picture[2,7-9].
compression of full-motion video consisting of small frames
and requiring slow refreshments. The data rate required is 940 Kbps, and the target applications include interactive
multimedia and video telephony. This standard requires the
development of new model-based image coding techniques
for human interaction and low-bit-rate speech coding
techniques.
As with JPEG and Motion JPEG, JPEG 2000 can also be
used to represent a video sequence. The advantages are equal
to JPEG 2000, i.e., a slightly better compression ratio
compared to JPEG but at the price of complexity.
The disadvantage reassembles that of Motion JPEG. Since
it is a still picture compression technique it does not take any
advantage of the video sequence compression. This results in
a lower compression ration compared to real video
compression techniques. The viewing experience if a video
stream in Motion JPEG 2000 is generally considered not as
good as a Motion JPEG stream, and Motion JPEG 2000 has
never been any success as a video compression technique.
IV. H.261/H.263
The H.261 and H.263 are not International Standards but only
Recommendations of the ITU. They are both based on the
same technique as the MPEG standards and can be seen as
simplified versions of MPEG video compression.
They were originally designed for video-conferencing over
telephone lines, i.e. low bandwidth. However, it is a bit
contradictory that they lack some of the more advanced
MPEG techniques to really provide efficient bandwidth use.
The conclusion is therefore that H.261 and H.263 are not
suitable for usage in general digital video coding.
Fig 1- 3D DCT video compression
The MPEG algorithm is intended for both asymmetric and
symmetric applications. Asymmetric applications are
characterized by frequent use of the decompression process,
while the compression process is performed once. Examples
include movies-on-demand, electronic publishing, and
education and training. Symmetric applications require equal
use of the compression and decompression processes.
Examples include multimedia mail and videoconferencing.
When the MPEG standard was conceived, the following
features were identified as being import: random access, fast
forward/reverse searches, reverse playback, audio-visual
synchronization, robustness to errors, editability, format
flexibility, and cost-trade-off. These features were described
in detail by LeGall The MPEG standard consists of three parts:
synchronization and multiplexing of video and audio; video;
and audio.
V. THE MPEG VIDEO COMPRESSION STANDARD
The MPEG compression algorithm is intended for
compression of full-motion video[3-6]. The compression
method uses interframe compression and can achieve
compression ratios of 200:1 through storing only the
differences between successive frames. The MPEG approach
is optimized for motion-intensive video applications, and its
specification also includes an algorithm for the compression
of audio data at ratios ranging from 5:1 to 10:1. The MPEG
first-phase standard (MPEG-1) is targeted for compression of
320x240 full motion video at rates of 1 to 1.5 Mb/s in
applications, such as interactive multimedia and broad cast
television. MPEG-2 standard is intended for higher resolutions,
similar to the digital video studio standard, CCIR 601, EDTV,
and further leading to HDTV. It specifies compressed bit
Fig 2. Parameters of MPEG Algorithms
streams for high-quality digital video at the rate of 2-80 Mb/s.
The MPEG-2 standard supports interlaced video formats and a
paper formatting requirements is to use this document as a
number of features for HDTV.
template and simply type your text into it.
The MPEG-2 standard also addresses scalable video coding
VI. MPEG FRAME STRUCTURE
for a variety of applications, which need different image
resolutions, such as video communications over ISDN In the MPEG standard, frames in a sequence are coded using
networks using ATM. The MPEG-4 standard is intended for three different algorithms, as illustrated in Figure 3.
ISSN: 2231-5381
http://www.ijettjournal.org
Page 319
International Journal of Engineering Trends and Technology (IJETT) – Volume 13 Number 7 – Jul 2014
Fig. 3 Types of frames in the MPEG standard
I frames (intra images) are self-contained and coded using a
DCT-based technique similar to JPEG. I frames are used as
random access points to MPEG streams, and they give the
lowest compression ratios within MPEG. P frames (predicted
images) are coded using forward predictive coding, where the
actual frame is coded with reference to a previous frame (I or
P). This process is similar to H.261 predictive coding, except
the previous frame is not always the closest previous, as in
H.261 coding.
VII. ACCORDIAN BASED VIDEO COMPRESSION
The basic idea is to represent video data with high correlated
form. Thus, we have to exploit both temporal and spatial
redundancies en video signal. The input of our encoder is so
called video cube, which is made up of a number of frames.
This cube will be decomposed into temporal frames which
will be gathered into one frame (2 dimensions). The final step
consists of coding the obtained frame. In the following, we
detail the method design steps[1,6,7-9].
Many experiences had proved that the variation of the 3D
video signal is much less in the temporal dimension than the
spatial one. Thus, pixels, in 3D video signal, are more
correlated in temporal domain than in spatial one this could be
traduced by the following expression: for one reference pixel
I(x,y,t) where: I: pixel intensity value, x; y: space coordinate
of the pixel, t: time (video instance). This assumption will be
the basis of the proposed method where we will try to put
pixels - which have a very high temporal correlation - in
spatial adjacency.
VIII.
ACCORDION BASED REPRESENTATION
To exploit this succeeding assumption, we start by carrying
out a temporal decomposition of the 3D video signal, the
figure 4 shows temporal and spatial decomposition of one
8X8X8 video cube:
Fig. 4. Spatial and temporal decomposition principle
Frames obtained by following the temporal decomposition
will be called .temporal frames.. These latter are formed by
gathering the video cube pixels which have the same column
rank. According to the mentioned assumption, these frames
have a stronger correlation compared to spatial frames. To
increase correlation in Accordion Representation we reverse
the direction of event frames. Figure 4 illustrates the principle
of this representation.
The .Accordion representation. is obtained following a
process having as input the GOP frames(I 1..N) and has as
output the resulting frame IACC. The inverse process has as
input the IACC frame and as output the coded frames (I 1..N).
Fig. 5. Accordion representation
1.
2.
3.
4.
5.
6.
ISSN: 2231-5381
IX. ACC-JPEG VIDEO COMPRESSION ALGORITHM
Decomposition of the video in groups of frames (GOP) is
performed as a beginning of the compression procedure.
Accordion Representation of the GOP is obtained as per
given in the previous section.
Decomposition of the resulting "IACC" frame into 8x8
blocks.
For each 8x8 block: Discrete cosine Transformation
(DCT) is applied.
Then Quantification of the obtained coefficients is
perform for each block.
Course in Zigzag of the quantized coefficients is done for
each block.
http://www.ijettjournal.org
Page 320
International Journal of Engineering Trends and Technology (IJETT) – Volume 13 Number 7 – Jul 2014
7.
Entropic Coding of the coefficients (RLE, Huffman) after
the step 6 to get the coded output.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
Fig. 6 ACC - JPEG diagram coding
X. CONCLUSION
The Video compression has gone through an evolution in the
last century. In this paper, we have studied the Different
techniques of the video compression, the Motion JPEG or
MJPEG is the basic technique of video compression, which
uses the
2D-DCT and consider each frame as a picture.
The MJPEG 2000 is similar to the MJPEG but it uses the
DWT in place of DCT. The MPEG is the actual technique
which uses the third dimension of the video for the
compression purpose. Video signal has high temporal
redundancies due to the high correlation between successive
frames. Actually, this redundancy has not been exploited
enough by these video compression techniques. The problem
is tried to resolve using MPEG. But as the MPEG uses the 3D
transformation and it’s a time taking operation so we prefer
the block by block transformation still the speed of operation
is an issue for this. To overcome this problem, we may prefer
to use the accordion based video compression in which the 2D
transformation is used to compress the accordion frames. Due
to which it act fast. Accordion is formed with the multiple
frames so it also has the features and effects of the temporal
motion.
[9]
Tarek Ouni, Walid Ayedi, Mohamed Abid, “New low complexity DCT
based video compression Method”, IEEE, pp. 202-207, 2009.
Borko Furht and Raymond Westwater,“Video Presentation And
Compression”, pp 171-204.
A. N. N. T. R. K.R., “Discrete cosine transforms”. in IEEE transactions
on computing, pp. 90.93, 1974.
N. Boinovi and J. Konrad, “Motion analysis in 3d dct domain and its
application to video coding”. vol. 20, pp. 510.528, 2005.
G. M.P. Servais, “Video compression using the three dimensional
discrete cosine transform”. in Proc.COMSIG, pp. 27.32, 1997.
M. B. T. Q. N. A. Molino, F. Vacca, “Low complexity video codec for
mobile video conferencing”. in Eur. Signal Processing Conf.
(EUSIPCO), (Vienna, Austria), pp. 665.668, Sept 2004.
“An explanation of video compression techniques”, www.axis.com
B. Furht, “A Survey of Multimedia Techniques and Standards, Part I:
JPEG Standard,” Real-Time Imaging Journal, Vol. 1, No. 1, pp. 49-67,
1995.
B. Furht, “A Survey of Multimedia Compression Techniques and
Standards,” Part II: Video Compression,” Real-Time Imaging Journal,
Vol. 1, pp. 319-337, 1995.
If we compare the performance of the different
techniques, we found that the MJPEG provides a compression
ratio up to 1:10. For the MJPEG 2000 it is 1:20, whereas for
the MPEG it’s about 1:30. For the ACC – JPEG it’s about
1;25 but with a simplicity in the algorithm and the relatively
fast speed of operation.
The described techniques are expectantly helpful for
the further study and the research in the field of video
compression. We can find some new method by combining
the feature of different techniques to get the more better
performance in terms of compression ratio, PSNR and bit rate.
ISSN: 2231-5381
http://www.ijettjournal.org
Page 321