Comparatives study of Various Techniques against Camcorder Piracy in Theater

Full text

Comparatives study of Various Techniques against
Camcorder Piracy in Theater
Nilesh Kumar Dubey
Devang Patel Institute of Advance Technology and Research,
CHARUSAT
nileshdubey.ce@charusat.ac.in
Hardik Modi
Chandubhai S. Patel Institute of Technology, CHARUSAT
hardikmodi.ec@charusat.ac.in
Abstract—for the long decay the cinema industries are
suffering from camcorder piracy due to that every year cinema
industries are losing billions of dollars. Most important cause
of piracy is camcorder piracy, in which pirates record the
movie during playback in theater. DCI (Digital cinema
Initiative) given many rule and regulation to protect from
piracy but still it is increasing, reason is there is no concrete
technical solution exist. In this paper various techniques that
can be opted in future for fighting against the piracy is taken.
There are two types of solution exist for the overcoming the
piracy problem one is through deterring the pirate and other is
not to let it be captured through camera. Watermarking is one
way to deter pirates and watermark can be embedded in
frames or audio of videos. Modulations techniques can be used
to generate flicker that cannot be detected by HVS but create
noise in camcorder recorded videos.
Keywords—HVS, Watermarking, Psychohvisual,
Modulation, Cinema, Theatre Piracy, camcorder.
I. INTRODUCTION
A lot of intense work and interest in making a film. It is
genuine simple to record it with a video recording device and
offer in shoddy rate in the commercial center. Every year
cinema industries of world are losing approximately $30
billion. Piracy causes a big threat to any country economies
and many legal acts implemented to protect against cinema
piracy. Motion picture industry approximated that about 80%
of the movie piracy take place from theatre illegal recording.
Theater piracy is continually big threat for cinema industries
in box office collection. In theater piracy, pirates records
movie through high quality camera and circulate on internet
or using DVDs before its official release. Then again, with
the consistent improvement of multimedia innovation, there
have been increasingly top notch recordings in individuals'
regular day to day existences through fast systems and top
quality hardware. In spite of the fact that the advancement of
technology can advance individuals' lives, it can likewise
make tremendous misfortunes item copyright proprietors in
light of the fact that computerized media items can be copied
and appropriated effectively with superb account gadgets.
Hence, in the system period of the computerized media
industry, it is important to investigate powerful answers for
anticipate theft and ensure protected innovation privileges of
advanced media items. Many techniques have been proposed
to tackle piracy problem but due to flaws in each techniques
still do not have concrete solution. A concrete solution will
come with incremental work to protect against theatre piracy.
Camcorder cinema piracy can be protected by two types of
techniques one is active technique and another is passive. In
active techniques cinema file is processed for protecting
against piracy before it plays back in theatre so that piracy
can be overcome. In passive techniques the cinema file is
unprocessed for protecting against piracy; piracy is defeated
by processing the pirated cinema file only. Many papers are
opted active technique and based on watermarking and
temporal psycochovisual modulation. In watermarking
method a secrete message is embedded in frame or audio of
cinema file before cinema file is play back in theatre and
now once piracy from theatre is done then the pirated video
can be used to extract the embedded watermark. Watermark
message consist information of show time, location and
theatre id of cinema where play back. So that by extracted
watermark theatre identity as well pirate identity can be
traced by estimating geometrical distortion occurs in pirated
cinema that may occur at the time of recording through
camcorder. This paper has been written to present all the
techniques till date opted for the protection of cinema against
the theatre piracy.
II. TECHNIQUES AGAINST CINEMA PIRACY
Many active and passive techniques are proposed to deal
with theatre piracy of cinema but till now not the single
technique is implemented to protect the cinema industry
from piracy. In this section all the existing techniques are
discussed with its problem in implementation.
I). Min-Jeong Lee, Kyung-Su Kim and Heung-Kyu Lee,
proposed watermarking [1] based solution to deal the theatre
piracy. In the proposed system authors embedded payload
consisting show time and location as watermark in cinema
file before it playback in theatre. Later on once the piracy is
done by the pirate then the pirated copy can be used to
extract the embedded information i.e. identity of theatre and
geometrical distortion that may occurred in pirated copy
during the piracy can be traced to know the possible position
of pirate. This techniques is divided in two part first part is
about watermark embedding and extraction, second part is
about estimation of geometrical distortion in pirated copy to
trace the pirate.
Watermark embedding process involves following steps:
1. Generate a basic watermark pattern which
consist information about show time and show
location. Generate basic watermark of size
//MmNnu
, where M and N symbolize
width and height of to be watermarked file, and
m and n symbolize number of repetition in row
and column respectively
2018 4th International Conference on Computing Communication and Automation (ICCCA)
978-1-5386-6947-1/18/$31.00 ©2018 IEEE 1

2. Repeat basic watermark pattern (
wb
)
mnu
time to attain periodicity and periodic
watermark (
w
) of size
MNu
will be
generated.
3. Now watermark
w
is embedded in each frame
(I) of cinema file while following below
equation:
(, ) (, ) (, ) (, )I xy Ixy xy wxy
w
KG
 u
where
( , ) ( mod , mod )
MN
wxy w x y
bmn
. (1)
Here
K
is universal weighting component and
(, )xy
G
is
a local component of each pixel (x,y) according to human
visual system.
Water mark extraction process involve following steps:
1. Preprocess the recorded copy according to eq. 2
because blind watermarking approach is used.
22
(, )
''
(, ) (, ) (, ) (, )
2(, )
xy v
Ixy xy I xy xy
w
xy
V
PP
V

 
(2)
Here

.
P
and

2.
V
are local mean and variance of
pirated file and
2
v
is defined as average noise variance.
Now the extracted watermark is defined as
''
'
w
wI I 
. (3)
2. Now analyze
'
w
using local auto correlation
function to estimate the occurred geometrical
distortion in pirated file to trace the pirate and
recover the hidden message in pirated file.
II). Yuta Nakashima, Ryuki Tachibana, and Noboru
Babaguchi [2] suggested audio watermarking based solution
to deprive piracy and pirate. Nakashima et. al embed
watermark in audio channel of movie, whole process of
embedding and detection of pirate and piracy take place in
following way:
1. Psychoacoustic based watermarking approach is used.
Watermark is embedded in such manner that embedded
signal cannot be audible to the movie watcher and watermark
is embedded is embedded in multiple soundtrack channel.
2. Position of pirate can be estimated from embedded
watermark as shown in fig.1. In this approach each channel
of sound track is taken as host signal (HS) and after adding
watermark signal in HS a watermarked host signal (WHS)
presented. Now when movies are playing back, the audio of
movie spread in air from various speakers located at multiple
part of theatre. Now, when pirate record the movie then
recorded audio will be amalgamation of all watermarked host
signals.
Fig 1: Process diagram of deterring the pirate using watermark
3. Recorded audio will be used for estimate the position
of pirate in theatre and it is possible because of delayed
occurred in recorded audio which comes from multiple
speaker located at different places in theatre.
III). Ahmad, Baudry and Chupeau et al. proposed flicker
[3] forensic based approach to deter the pirate, in this
approach they utilized the flicker generated during recording
of video which is displayed on LCD screen. Flicker effect
comes in recorded video because of interplay difference
between LCD and recording devices. The whole process of
deterring piracy is listed below:
1. Modeled the luminance flicker that can be generated
while recording the video from LCD display as in eq. 4
[,] cos 2 2
1
ff
bl bl
fkr B k r
ff
cs
SS

§·
¨¸
©¹
(4)
Where
[,]fkr
is discrete signal and observed as the
recorded luminance, k and r are sampling index for frame
and row respectively,
sin ( ) sin ( )
1
BccfT
ss
bl
T
, fbl is back
light frequency in LCD, fc is frame rate, fs is sampling rate.
2. Estimated the vertical radial frequency of flicker
because flicker signal energy in comparison to recorded
video is too low.
3. Match the relation that pirated video is captured
through seized recorded video by utilizing the estimated
vertical radial frequency in step 2.
IV). Li Li et. al presented video watermarking [4] scheme
to protect cinema from piracy that is based on H.264/AVC
HDTV watermarking approach this proposed method is
robust against the various geometric attack and algorithm is
mentioned below:
1. Extract frame from video which is to be watermarked
and apply RGB to YUV conversion and then apply DCT
transformation on Y factor only.
2. Now convert the transformed image into u8 8
different block and compute just noticeable difference value
for each u8 8 block.
2

3. Combine the all processed region of whole frame.
4. If the added watermark size is 1 bit then insert the
Watermark pattern to only initial two frame of the sequential
five frame collection.
5. Embedding process applied multiple time on multiple
frames and then assembled these all watermark embedded
frame to get watermarked video back.
III. POSSIBLE ATTACK ON PROTECTED VIDEO
Most of the technique for protecting cinema against the
piracy is based on watermarking so it is important to study
the characteristics of watermarking and possible attack may
be carried out over watermarked file to destroy or manipulate
the watermark. There are many attack till now explored that
may affect the content of watermark in watermarked file.
Once watermarked file is tempered in such a way that
removes or change the meaning of watermark then there is
no significance of embedding watermark so developed
algorithm must robust against the possible attack that may
affect the watermark file. Here in this section all the possible
attacks are discussed that must be taken under consideration
at the time of watermarking.
Frame Averaging and Statistical Analysis: In frame
averaging and statistical attack, [6-7] attackers take multiple
watermarked frames and predict the watermark and remove
it from watermarked file.
Frame Dropping: As in any video there are multiple
frames of same information and if the attacker able to
estimates the frame [8-9] with watermark can remove that
frame before presenting and no one can notice the missing
frame.
Frame Cropping: In this attack, attacker may crop some
portion of the frame and then present the cropped frame.
Extraction of watermark from the cropped file becomes very
hard [8-9].
Geometrical distortion: This may arise in recorded
watermarked file because of angle, scaling, distance and
rotation during recording. Filters to be applied: Any low pass
filter may be applied to degrade the quality of watermark
information on pirated copy for example median filter can be
applied to remove the noise from the frame but it may
remove the watermark too from the frame [10].
Noise can be added: Any signal generally contain
takeoffs from the perfect signal that would be created by our
model of the signal generation process. Such takeoffs are
alluded to as noise. There are mainly two kind of noise is
existing that can be applied over a watermarked frame to
deteriorate the watermark, first is Gaussian Noise and second
is Salt and Pepper noise. Gaussian noise is a statistical kind
of noise in which probability distribution function is
equivalent to the normal distribution that is also known as
Gaussian noise. Another noise that is very frequently used by
attackers is salt and pepper noise that is also called as
impulse noise that can be implemented on image by sharp
and sudden change in image signal.
Collusion: At times, it is feasible for an assailant to
acquire numerous watermarked duplicates [11-14]. The
assailant can frequently misuse this circumstance to evacuate
watermarks, without knowing the watermarking calculation.
An attack that needs a few watermarked duplicates is known
as agreement assaults. There are two fundamental sorts of
collusion attack:
Type-1: In this sort of collusion attack, attacker gets a
few duplicates of a similar work, with various watermarks.
Here, the aggressor attempts to discover the video outlines
which are comparable in nature. Thus, outlines having a
place with a similar scene have a high level of connection.
The assailant at that point isolates different scenes of the
video. At that point factual normal of the neighboring edges
is done to combine the diverse checks and figures another
plain casing. Sort 1 intrigue assault must be effective if
progressive edges are sufficiently diverse.
Type-2: In this kind of attack, the assailant gets a few
distinct duplicates that contain a similar watermark and
studies them to find out about the calculation. At that point a
few duplicates are found the middle value of by the assailant.
In the event that all duplicates have a similar reference
design added to them, at that point this averaging task would
return something that is shut to the example. At that point,
the normal example can be subtracted from the duplicates to
produce a plain video [15].
IV. CONCLUSION
In this paper theatre piracy problem is discussed with the
work that is carried out to protect cinema industry from the
piracy. There are various approaches that are used to give a
concrete solution against the theatre piracy but not the single
approach is applied over full length movie file. All approach
used small size video with very less number of frames to test
the given algorithm. On the other hand actual movie size is
very large with the millions of frame. Most of the work is
based on watermarking scheme either movie frame or audio
channels are used to embed the watermark and later on once
the pirated copy come in market then it is used to extract the
watermark form the pirated copy and then further
investigation applied to identify the place and device of
piracy. Any signal generally contains takeoffs from the
perfect signal that would be created by our model of the
signal generation process. Such takeoffs are alluded to as
noise.
REFERENCES
[1] Lee, M., Kim, K., & Lee, H. (2010). Digital Cinema Watermarking
for Estimating the Position of the Pirate. IEEE Transactions on
Multimedia, 12(7), 605-621. doi:10.1109/tmm.2010.2061221.
[2] Li, L., Dong, Z., Lu, J., Dai, J., Huang, Q., Chang, C., & Wu, T.
(2015). AN H.264/AVC HDTV watermarking algorithm robust to
camcorder recording. Journal of Visual Communication and Image
Representation, 26, 1-8. doi:10.1016/j.jvcir.2014.08.009.
[3] Hajj-Ahmad, A., Baudry, S., Chupeau, B., Doerr, G., & Wu, M.
(2017). Flicker Forensics for Camcorder Piracy. IEEE Transactions
on Information Forensics and Security, 12(1), 89-100.
doi:10.1109/tifs.2016.2603603.
[4] Nakashima, Y., Tachibana, R., & Babaguchi, N. (2009).
Watermarked Movie Soundtrack Finds the Position of the Camcorder
in a Theater. IEEE Transactions on Multimedia, 11(3), 443-454.
doi:10.1109/tmm.2009.2012938.
[5] Gosavi, C. S., & Mali, S. N. (2017). Secure, Robust Video
Watermarking to Prevent Camcorder Piracy. Indian Journal of
Science and Technology, 10(18), 1-10.
doi:10.17485/ijst/2017/v10i18/111377 .
[6] Zhao, Y., & Lagendijk, R. (n.d.). Video watermarking scheme
resistant to geometric attacks. Proceedings. International Conference
on Image Processing. doi:10.1109/icip.2002.1039908.
[7] Dubey, N. K., & Kumar, S. (2014). A Review of Watermarking
Application in Digital Cinema for Piracy Deterrence. 2014 Fourth
3

International Conference on Communication Systems and Network
Technologies, 236-240. doi:10.1109/csnt.2014.131.
[8] Dubey, N. K., & Kumar, S. (2015). An Effective Approach of
Distortion-Resistant Video Watermarking for Piracy Deterrence.
International Journal of Security and Its Applications, 9(1), 283-294.
doi:10.14257/ijsia.2015.9.1.27.
[9] Liu, L., Li, R., & Gao, Q. (2005). A robust video watermarking
scheme based on DCT. 2005 International Conference on Machine
Learning and Cybernetics. doi:10.1109/icmlc.2005.1527856.
[10] Lagzian, S., Soryani, M., & Fathy, M. (2011). Robust watermarking
scheme based on RDWT-SVD: Embedding data in all subbands. 2011
International Symposium on Artificial Intelligence and Signal
Processing (AISP). doi:10.1109/aisp.2011.5960985.
[11] Xu, X., Zhang, R., & Niu, X. (2010). Image synchronization using
watermark as RST invariant descriptors. 2010 IEEE International
Conference on Information Theory and Information Security.
doi:10.1109/icitis.2010.5689703.
[12] Zhang, J., Liu, S., & Yin, Z. (2009). Blind digital watermarking
method against geometric attack. Journal of Computer Applications,
28(9), 2249-2251. doi:10.3724/sp.j.1087.2008.02249.
[13] Wang, Y., & Moulin, P. (2007). Capacity and optimal collusion attack
channels for Gaussian fingerprinting games. Security, Steganography,
and Watermarking of Multimedia Contents IX.
doi:10.1117/12.706575.
[14] Lin, Y., & Abdulla, W. H. (2014). Proposed Audio Watermarking
Scheme. Audio Watermark, 95-121. doi:10.1007/978-3-319-07974-
5_4.
[15] K., & P. (2015). 3rd Level DWT & SVD Based Video Watermarking.
International Journal of Science and Research (IJSR), 4(11), 1483-
1487. doi:10.21275/v4i11.nov151496.
4