an efficient image watermarking scheme based on dwt...
TRANSCRIPT
An Efficient Image Watermarking Scheme
based on DWT-DFRNT Dual Transform Method
for Digital Content Protection
M. Kim1 , D. Li2, , S. Hong
1
1 Department of Computer and Information Sciences, Towson University,
Towson, Maryland, USA 2 Department of Computer Science, University of Yanbian,
Yanji, China
Abstract. With the rise of digital contents, their copyright protection has
emerged as a serious issue and digital watermarking can be introduced as an
effective method for solving the issue. In this paper, we propose a new digital
image watermarking algorithm based on the dual transform method, discrete
wavelet transform (DWT) and discrete fractional random transform (DFRNT).
We generate a watermark through a two-dimensional (2D) barcode and the
block code encoder, and the generated watermark image is embedded into
DWT-DFRNT using quantization technique in order to ensure robustness and
security of the algorithm. Experimental results present that our proposed
method has improved the extraction performance of watermark and ensured the
imperceptibility, security, and robustness of the watermark against image signal
processing attacks such as compression and noise adding by increasing the
hidden computational complexity of information and combining DWT and
DFRNT transforms.
Keywords: Watermarking, Discrete Wavelet Transform (DWT), Discrete
Fractional Random Transform (DFRNT), 2D-barcode, Quantization
1 Introduction
The number of digital media contents has continued to increase rapidly due to
advances in network and multimedia technologies. As a result, copyright infringement
of digital content is also increasing and it is now recognized as a serious problem.
Digital watermarking [1, 2] may be used as an effective method for identifying the
copyright ownership of digital content from unauthorized use and distribution. In
digital watermarking, a watermark which contains the copyright information is
embedded into the digital media content such as images, audio, and video, and may be
extracted when the copyright ownership of the digital content needs to be verified [2,
3]. In the previous researches, many digital watermarking methods have been
Corresponding author
ITCS 2013, ASTL Vol. 25, pp. 345 - 348, 2013 © SERSC 2013
345
proposed. There are discrete cosine transform (DCT) [4, 5], DWT [6, 7], singular
value decomposition (SVD) [8], and DFRNT [9] single domain based watermarking
techniques. To complement single transform based watermarking, dual transform
domain based watermarking techniques such as DWT-DCT [10] and DWT-SVD [11]
have also been proposed.
In this paper, we propose an efficient digital image watermarking algorithm based
on the dual transform method which combines DWT and DFRNT domains using 2D
barcode in order to the imperceptibility, security, and robustness of the watermark
against image signal processing and various attack.
2 Proposed Algorithm
2.1 Techniques in Proposed System
In watermark generation process, we use 2D barcodes which have various
information capacities and the self-error correction function and the block code
pattern for encoding a 2D barcode. The process increases the hidden computational
complexity of information. The transform method used in the proposed system
consists of a combination of DWT and DFRNT [12] dual domain in order to ensure
the imperceptibility, security, and robustness of the watermark due to the frequency
decomposition ability of DWT that extracts robust coefficients and the unpredictable
random characteristic of DFRNT.
2.2 Watermarking Process
Fig. 1. Watermark embedding process
The watermark embedding process is summarized as follows: First, we generate the
watermark image through the 2D barcode and block code encoding. Next, the original
host image signal is decomposed into three sub-bands, H, V, and D, through the two-
level 2D-DWT, and DFRNT is performed on the sub-band coefficients. And then, we
embed the watermark image into the sub-band coefficient value using quantization
technique. Finally, we performed the inverse DFRNT (IDFRNT) and inverse DWT
(IDWT) and obtain the watermarked image. Fig. 1 shows steps in the watermark
embedding process. The watermark extraction process is performed in such a way that
the embedding process is inversely applied to the watermarked image signal.
Proceedings, The 2nd International Conference on Information Technology and Computer Science
346
3 Experimental results
In order to evaluate the performance of our proposed algorithm, we have performed
the experiments with respect to the imperceptibility and robustness of the watermark
in various conditions including compression and noise adding attacks.
We used a 512512 standard gray level image of size as the sample host image and
a 2121 cell QR code, and 22 block code encoder. While 2D-DWT is performed, the
image signal is scanned in 88 block unit. In order to evaluate the imperceptibility
and robustness, we compute the peak signal to noise ratio (PSNR) and Bit Error Rate
(BER). BER value is calculated as follows:
(1)
(a) Original image (b) Watermarked image
Fig. 2. Image signals before and after embedding watermark.
The comparison of the original image with the watermarked image without being
attacked is shown in Fig. 2. The value of the quantization coefficient Q value was set
to 25 and the acquired PSNR is 40.01 dB.
Table 1. Experimental results by various attacks.
Attack type Compression Gaussian noise
Watermark
BER=7.48% BER=17.6%
QR Code
BER=0.09% BER=1.84%
We performed the experiments by carrying out compression and noise adding attacks,
and the extracted watermarked image and restored QR code after being attacked are
shown in Table 1. In the case of compression attack, JPEG Quality Factor (QF) was
set to 65. In the case of Gaussian noise attack, mean was set to 0 and variance was set
An Efficient Image Watermarking Scheme based on DWT-DFRNT Dual Transform Method for Digital Content Protection
347
to 0.001. The experimental results verify that the proposed algorithm ensures the
imperceptibility and robustness of the watermark.
4 Conclusions
In this paper, we have proposed an efficient image watermarking scheme based on
DWT-DFRNT dual transform method for Digital Content Protection. We used a 2D
barcode which contains watermark information and the block code encoder to
generate watermarks. We performed various experiments to evaluate the performance
of the proposed algorithm. The experimental results presented that the proposed
algorithm has good imperceptibility and ensure robustness of watermark against
image compression and Gaussian noise attacks.
Acknowledgments. This research project was supported by the National Natural
Science Foundation of China (Grant No. 61262090).
References
1. Cox, I. J., Miller, M. L., Linnartz, J. M. G., Kalker, T.: A Review of Watermarking
Principles and Practices. Digital Signal Processing for Multimedia Systems. Marcel Dekker,
Inc. 461-485 (1999)
2. Cox, I. J., Miller, M. L.: The first 50 years of electronic watermarking. Journal on Applied
Signal Processing. 2002(2), 126-132 (2002)
3. Petitcolas, F. A. P., Anderson, R., Kuhn, M. G.: Information hiding – A survey. Proceedings
of the IEEE. 87(7), 1062-1078 (1999)
4. Chu, W. C.: DCT-Based Image Watermarking Using Subsampling. IEEE Transactions on
Multimedia. 5(1), 34-38 (2003)
5. Dogan, S., Tuncer, T., Avci, E., Gulten, A.: A New Watermarking System Based on
Discrete Cosine Transform (DCT) in Color Biometric Images. Journal of Medical Systems.
36(4), 2379-2385 (2012)
6. Lu, W., Sun, W., Lu, H.: Novel robust image watermarking based on subsampling and DWT.
Multimedia Tools and Applications. 60(1), 31-46 (2012)
7. Wu, S. Q., Huang, J. W., Shi, Y. Q.: Efficiently self-synchronized audio watermarking for
assured audio data transmission. IEEE Transactions on Broadcasting. 51(1), 69-76 (2005)
8. Chang, C., Tsai, P., Lin, C.: SVD-based digital image watermarking scheme. Pattern
Recognition Letters. 26(10), 1577-1586 (2005)
9. Guo, J., Liu, Z., Liu, S.: Watermarking based on discrete fractional random transform.
Optics Communications. 272(2), 344-348 (2007)
10. Su, Q., Yuan, H.: Combined dwt-dct digital audio watermarking. Energy Procedia. 13, 432-
438 (2011)
11. Li, Q., Yuan, C., Zong, Y.Z.: Adaptive DWT-SVD domain image watermarking using
human visual model. In: Proceedings of 9th International Conference on Advanced
Communication Technology, Vol. 3, pp. 1947--1951. Gangwon-Do, Korea (2007)
12. Liu, Z., Zhao, H., Liu, S.: A discrete fractional random transform. Optics Communications.
255(4), 357-365 (2005)
Proceedings, The 2nd International Conference on Information Technology and Computer Science
348