Journal of
Imaging Article
Computationally Efficient Robust Color Image Watermarking Using Fast Walsh Hadamard Transform Suja Kalarikkal Pullayikodi *
ID
, Naser Tarhuni, Afaq Ahmed and Fahad Bait Shiginah
Department of Electrical and Computer Engineering, Sultan Qaboos University, 33 Muscat, Oman;
[email protected] (N.T.);
[email protected] (A.A.);
[email protected] (F.B.S.) * Correspondence:
[email protected] Received: 8 September 2017; Accepted: 10 October 2017; Published: 13 October 2017
Abstract: Watermark is the copy deterrence mechanism used in the multimedia signal that is to be protected from hacking and piracy such a way that it can later be extracted from the watermarked signal by the decoder. Watermarking can be used in various applications such as authentication, video indexing, copyright protection and access control. In this paper a new CDMA (Code Division Multiple Access) based robust watermarking algorithm using customized 8 × 8 Walsh Hadamard Transform, is proposed for the color images and detailed performance and robustness analysis have been performed. The paper studies in detail the effect of spreading code length, number of spreading codes and type of spreading codes on the performance of the watermarking system. Compared to the existing techniques the proposed scheme is computationally more efficient and consumes much less time for execution. Furthermore, the proposed scheme is robust and survives most of the common signal processing and geometric attacks. Keywords: spreading; hadamard; robust; attacks; computational
1. Introduction Now-a-days, internet and multimedia technologies have become part and parcel of the human life. The information shared on the world wide web (www) is proliferating day by day. Due to the recent advances in digital technologies, these contents can easily be copied, downloaded and manipulated [1]. Hacking and piracy turns out to be the byproduct of the digital revolutions. To prevent the anonymous manipulation of the digital contents and to protect the copyright, multimedia researchers introduced the copy deterrence mechanism called watermarks which are usually logos, stamps, Gaussian random patterns or pseudo noise sequences that are embedded in the media to be protected. Watermarking does not ensure that the watermarked image is completely secure because there can be intentional manipulations of the image (collusion, confusion and ownership deadlock), which actually try to destroy the embedded watermark or non-intentional manipulations like geometrical and signal processing operations (cropping, scaling, filtering, A/D conversion, etc.) [2]. The embedding algorithm should be designed in such a way that the embedded watermark survives the above mentioned attacks and can be extracted from the attacked image by the decoder. Such algorithms are known as robust and secure algorithms. The introduction of the watermarks should not degrade the perceptual quality of the original image. But as the amount of watermark content in the image to be protected (cover image) increases, the perceptual transparency shows significant reduction [3]. At the same time, more the amount of watermarks, the more robust will be the watermarked image against distortions and attacks. So the most important and desirable properties of the watermarks are perceptual quality and robustness even though there exists trade off among those properties. Along with that the good payload size, reduced false positive rate and information hiding capacity are also expected [4,5]. Watermarks can be added to the digital image by modifying the pixel values as per the values of the watermarks which is known as spatial domain watermarking [6]. Alternately, by altering J. Imaging 2017, 3, 46; doi:10.3390/jimaging3040046
www.mdpi.com/journal/jimaging
J. Imaging 2017, 3, 46
2 of 13
the transform domain values using common transforms like Discrete Cosine Transform (DCT) [3,4], Discrete Wavelet Transform (DWT) [7–9], Fast Walsh Hadamard Transform (FWHT) [10,11] or Fast Fourier Transform (FFT). Even though spatial domain techniques are simple to implement they are less robust against attacks [6]. On the other hand, frequency domain techniques show good robustness and perceptual quality but implementation is complex compared to spatial domain techniques [8]. Some researchers combine two transforms to achieve the advantages of both the transforms [12–14]. Watermarking has various applications like authentication, copyright protection, military and satellite applications, broadcast monitoring (ensuring the proper broadcast at the proper timing), transaction tracking/finger printing (finding out the source of leakage of multimedia signal), medical applications and many more [2,3]. Based on the application, the properties should be set and accordingly the embedding and decoding algorithm should be modified. Among different ways of the watermarking models, the communication based model is the most important and widely used. From a communication point of view, watermarking system can be viewed as a process of communicating a message (watermark), through the fixed bandwidth channel (cover image), from the watermarking embedder to the watermarking decoder [1,3,4,8]. Specifically, some researchers apply the concept of CDMA based communication system in which watermark will be spread using the orthogonal spreading code which is generated using the security key and is added to the selected frequency coefficients [7]. Decoding is usually done using correlation of the received image with the spreading codes. Spread spectrum (SS) watermarking is considered to be the most efficient, secure and robust watermarking scheme which yields high Signal to Noise Ratio (SNR), enhanced robustness and low distortion [1,3]. As in DS-CDMA (Direct Sequence—CDMA), different watermarks are spread using mutually orthogonal codes or different parts of a single watermarks are spread using separate codes [2,4]. The additive noise and signal distortions in communication systems are similar compared with the unintentional manipulations on the image and malicious attacks in watermarking. As the watermark is spread throughout the image bandwidth such that the energy of the embedded watermark in any particular frequency bin is very small, it is difficult for the eavesdropper to detect and destroy the watermark [2]. Most of the existing literatures use a single code to spread and few researchers have worked on multiple mutually orthogonal spreading codes [11,14]. As each part of the watermark is spread throughout the cover image, the length of each spreading code can be increased providing more security. If N independent and orthogonal codes are used, the attacker has to break N codes [3]. The selection of transform for the decomposition of image and the properties of spreading codes determine the efficiency of the data spreading. This paper is organized as follows. Section 2 explains the proposed algorithm of embedding and decoding followed by the simulation results in Section 3 and finally, Section 4 concludes the paper. Mainly we have used, three different cover images and three different watermarks for analysis as shown in succeeding sections. 2. Walsh Hadamard Transform and Related Works Walsh Hadamard Transforms (WHT) are generalized version of Fourier transforms. The transformation matrix of the WHT consists of only value of 1 and −1 which directly implies that the calculation of Walsh Hadamard Transform of an image consists of " only real# additions and 1 1 subtractions [10]. The elementary matrix of Walsh Hadamard Transform is . Higher order 1 −1 matrices (HN ) are calculated as the kronecker product (×) of lower order matrices as given below [15]. HN = HN −1 × H1 " HN =
H N −1 H N −1
H N −1 − H N −1
(1) # (2)
J. Imaging 2017, 3, 46
3 of 13
"
# 1 1 where H1 = . 1 −1 The rows of the Hadamard matrix are orthogonal and the transpose (HN T ), conjugate (HN ∗ ) and inverse of the matrix (HN −1 ) are the same. H N = H N ∗ = H N −1 = H N T
(3)
Complexity of calculating the Walsh Hadamard Transform is O(Nlog2 N) real additions and subtractions. Equations of forward Walsh Hadamard Transform of a one dimensional matrix u(m) is v u N −1 u U (k) = 1/t N ∑ u(m) (−1)b(k,m) 0 ≤ k ≤ N − 1
(4)
m =0
k = k0 + 2k1 + · · · · · · +2n−1 k n−1 , k0 , k1 , · · ·, k n−1 are binary values m = m0 + 2m1 + · · · · · · +2n−1 mn−1 , m0 , m1 , · · ·, mn−1 are binary values n −1
b(k, m) =
∑ k i mi
i =0
The two dimensional transform is found out by calculating the one dimensional WHT of the rows first followed by WHT of the columns. The transformation matrix can be Walsh ordered or Hadamard ordered. Walsh ordered matrix has elements arranged in increasing order of frequency (sequency). Sequency is the more generalized 24 term for frequency which is the number of sign changes among the elements of a row in the matrix. But Hadamard matrix has elements arranged in hadamard order (0, 4, 6, 2, 3, 7, 5, 1). The inverse and forward transformations are identical except for a scalar multiplication factor. WHTs are computationally very efficient and hardware implementations are cost effective as both forward and inverse transformations can be done through the same block which reduces the space, complexity and cost considerably [11,14]. Most of the energy is contained in less than 50% of the coefficients yielding good energy compaction properties and high resiliency towards low quality compression. So WHT turns out to be the best choice for robust, low cost implementations [2,11]. Only few numbers of researchers worked on FWHT based algorithms. F. Rajkumar et al. combined the image decomposition method, Singular Value Decomposition (SVD) and FWHT and used the entropy (measure of randomness of the pixel values) of the blocks to decide the block to embed the watermark [14]. In [16], Anthony et al. developed a robust watermarking algorithm, but computationally, it takes more time as both watermark and original image are to be frequency transformed. The watermark is embedded in the coefficients selected by the security key. Moreover the DC coefficients of each block should be stored in a separate file. Some of the researchers used complex Hadamard transform. Aris et.al. have applied FWHT on the watermark and has added to the DC coefficients of the host image and they could achieve good visual perception [17]. In [18], the authors tried to add the watermark in the Hadamard coefficients. But no attack analysis is performed instead they studied the effect of scaling factor on the embedding process. In [19], Ishikawa et.al. proposed an illumination based watermarking technique which used a combination of FWHT and DCT. With this technique they were able to attain 100% accuracy and robustness against compression. In [9] complex Hadamard transform is used and the watermark is embedded in the phase of the transform coefficients. Santi. P. Marity et al. used spread spectrum based watermarking using FWHT. They have incorporated characteristics of Human Visual System (HVS) also to make the system more robust and perceptually better [20]. Combining HVS and FWHT offered good robustness against fading as well. The VLSI implementation of the similar algorithm has been done using FPGA which can be used for the blind assessment of Quality of Service (QoS) of the wireless fading channel. In [21] authors
J. Imaging 2017, 3, 46
4 of 13
have combined the effect of DCT and FWHT. FWHT of the watermark to be embedded is calculated before embedding into the DC coefficients. DCT of each of the 8 × 8 blocks is calculated to extract the DC coefficients of each block and they showed that the proposed technique is better than DCT based technique. E.E. Abdullah et al. combined Single Value Decomposition and FWHT and formulated a high rate, robust and easily implementable algorithm [22]. E. Mwangi used the rows of the Hadamard matrix to spread the watermark using CDMA technique and spread watermark is embedded into the selected DWT coefficients [23]. 3. Proposed Algorithm The proposed algorithm makes use of CDMA based spread spectrum concept, explained earlier, for embedding and decoding the watermark. The spreading codes used are the hadamard codes which are orthogonal codes. 3.1. Computational Efficient Calculation In this paper the computation of WHT is done using a customised algorithm developed for calculating 8 × 8 WHT cells. The algorithm has been implemented in MATLAB (2013) using the idea of fast computation of Discrete Fourier Transform (DFT), without using any loops and using only matrix additions and subtractions. With this customized technique, computation of WHT of an 8 × 8 image takes only 0.15 milliseconds which is considerably less than the time for execution using existing MATLAB implementation which consumed 1 milliseconds for the same applied image. The cyclomatic complexity of the proposed algorithm is found to be equal to one which means that there is a single path from input to output. On the other hand cyclomatic complexity of MATLAB implemented WHT and DCT is 12. A new, computationally efficient CDMA based watermarking algorithm using above mentioned WHT is formulated with a specific selection of frequency coefficients which has better robustness and performance compared to the existing similar algorithms using other transforms like DCT [3]. Variation of robustness with change in spreading code length, number of spreading codes and type of spreading code is studied in detail. 3.2. Embedding Algorithm Even though watermarking of the color images are not done by most of the researchers, some of the existing algorithms watermark in all the three (R,G,B) color planes, and some of the researchers embed only in the blue plane as changes in the blue component will not be perceived by the human eye much. Here, the color image to be watermarked is converted to YUV format and the watermark is embedded only in the Y plane based on the property of human visual system (HVS) which states that the human eye is more vulnerable to changes in the color component compared to that in the intensity component [9]. The Y matrix that is to be watermarked is divided into k number of 8 × 8 blocks (i1 , i2 , i3 , i4 , . . . , ik ). Each of these blocks is transformed using Fast Walsh Hadamard Transform to get the transformed blocks (h1 , h2 , h3 , h4 , . . . , hk ). The binary watermark (B) is converted into one dimensional string (b1 , b2 , b3 . . . , bn ) each of the values either 1 or −1. For spreading, r numbers of, polar orthogonal hadamard codes are used. As shown in Figure 1 mutually similar mid band coefficients along the 3rd, 4th, 5th and 6th rows are selected for embedding the watermark and rest of the coefficients are left unchanged. Mid frequency coefficients are selected as the changes on them do not cause much changes to the perceptual quality compared to the lower frequency coefficients. In addition, these mid band frequencies are less affected by compression. The selected frequency coefficients along xth row are made into a single vector rx such that length of rx is same as that of spreading codes used. Each bit of the watermark in each modifying row is spread using different spreading codes, then, spread watermarks are added to the selected coefficient vector. The embedding equation is, W x = r x + α(s1 b1 + s2 b2 + · · · + sr br )
(5)
J. Imaging 2017, 3, 46
5 of 13
where α is the modulation index or gain of embedding (selected based on the allowed distortion and the required robustness), rx is the xth row used for watermarking and W x is the FWHT of the xth row of the watermarked image. Watermarked image can be obtained after finding the inverse FWHT of all the watermarked blocks and reconstructing into their original locations. After watermarking, the Y plane isJ. Imaging replaced back 2017, 3, 46and the image is converted back to RGB format. 5 of 12 J. Imaging 2017, 3, 46
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1 1 1 1 1 1 1 1
-
-
-
-
-
-
-
-
. . .
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
. . .
-
-
-
-
-
-
-
-
. . .
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1
1 1 1 1 1 . 1 1 1 1 1 . 1 1 1 1 1 .
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1
1 1 1 1 - - - - - - . . 1 1 1 1 - - - - - - . . 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 . . -
-
-
-
-
-
-
-
-
-
-
-
1 1 1 1 1 1 1 1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
. 1 . 1 . 1
. 1 1 1 1 1 1 1 1 1 1 1 1 . - - 1 1 1 1 1 1 . - - 1 1 1 1 1 1
. .
-
-
-
-
-
-
-
1 1 1 1 1 1 -1 - 1 -
-
1- 1- 1- 1- 1 1 1- 1-
-
-.
.- . -
-
-
-
-
-
-
- -- - - -
-
-
--
-
-
-
-
-.
-.
- - - -
-
.- . - .
-
-
-
-
-
-
1 - 1 - 1 1 1- 1- 1- 1-
-
-
1- -1 1 1. .1 . 1 1 - 1 -
-
-
-
-
1 1 1 1 1 1 1 1
-
-
. .
-
1 1 1 1 1 1 1 1
-
1 1 1 1 . . - 1 1 1 1 . - . 1 1 1 1 . . 1 1 1 1 - - - . . . . 1 1 1 1 . - - - . . . 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
--
--
--
.-
-
-
-
-
-
-
-
-
-
-
-
-
. . .
-
-
-
-
-
-
-
-
. . .
1 1 1 - - . . 1 1 1 - - . . 1 1 1 1 1 1 1 1 1 1 . .
-
1 1 1 1 1 1 1 1 -
-
1 - 1 1 .1 . 1 . 1 1 1-
-
-
-
-
-
-
-
-
-
-
-
1 1 1 1 1 1 1 1 1 1
-
- 1- 1 - 1 - 1 - 1 - 1 1 1 -
-
-
-
-
1- 1 1 - 1 - 1 - 1 - 1 - 1 -
-
.- . - .
-
-
1- 1 1 - 1 - 1 - 1 - 1 - 1 -
-
-
. . .
.- . - . -
-
-
-
. . . . . .
1 1 . 1 . 1. 1 1 1 1
. . .
1 1 . 1 . 1. 1 1 1 1
. . .
1 1 1 1 1 1 1 11 1
1 11 11 11 11 1 1 1 1 11 1 . 1 . 1. 1 1 1 1 1 1 1. .1 .1 1 11 1 1 11 1 1 1 1 11 1 1 1 1 1 1 1 . 1 . 1. 1 1 1 1
. . .
1 1 1 1 1 1 -1 -1 -
-
1- 1 - 1 - 1 - 1 1 1 - 1 -
-
. - . -.
-
-1 -1
1 1. 1. .1 1 1-
-
- 1- 1 - 1 - 1 - 1 - 1 1 1 -
1 1 1 1 1 1 -1 -1 -
-
1- 1 - 1 - 1 - 1 1 1 - 1 -
-
. - . -.
-
-1 -1
1 1. 1. .1 1 1-
-
-
1- 1- 1 1 - 1 - 1 -1 - 1
-
. . .
-
-
-
1 - 1 1 . 1 . 1. 1 1 - 1 -
. . .
- .- . - . -
-
-
-
. . .
. . .
-
-
-
-
-
-
-- - - - - - - -- -
-
-
1 1 .1 . 1 . 1 1 1 1
-
1- 1 1 - 1 - 1 - 1 - 1 -1
-
-
-
-
--
-
.- . - .
-
-
- -
-
-
.- . - . -
-
-
-
-
-
-
-
-
. . . . . .
-
1- 1- 1 1 - 1 - 1 -1 - 1
1 - 1 1 . 1 . 1. 1 1 - 1 -
-
. . .
-
1- 1 - 1 - 1 - 1 - 1 1 1 -
-
- . - . -.
-
-
-
-
. . .
. - . -.
-
-
-
-
. . .
-
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1
. . .
1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1
. . .
-
-
. . .
-
-
. . .
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
5 of 12
. . .
1. Location midcoefficients band coefficients for embedding. Figure Figure 1. Location of mid of band used forused embedding. Figure 1. Location of mid band coefficients used for embedding. -
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
. . .
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
. . .
3.3. Decoding Algorithm 3.3. Decoding Algorithm 3.3. Decoding TheAlgorithm received watermarked image is converted to YUV format and Y plane is split into 8 × 8 The received watermarked image is converted to YUV format and Y plane is split into 8 × 8 blocks as in the embeddingimage side. is The split block is transformed to domain using watermarked converted to YUV format Y frequency planedomain is split into 8 Walsh × 8 Walsh blocks The as inreceived the embedding side. The split block is transformed to and frequency using Hadamard Transform (FWHT). Hadamard spreading generated as blocks as in the embedding side. The The split same block orthogonal, is transformed to frequency domaincodes usingare Walsh Hadamard Transform (FWHT). The same orthogonal, Hadamard spreading codes are generated as at the embedding side andThe the same correlation of theHadamard modified spreading rows withcodes each of spreading Hadamard Transform (FWHT). orthogonal, arethe generated as codes is at the embedding side and the correlation of the modified rows with each of the spreading codes at the embeddingThe sidesign and the correlation of thevalue modified rows withthe each of the spreading codes bits. is calculated. of the correlation determines embedded watermark If the is calculated. The sign of the correlation value determines the embedded watermark bits. If the calculated. The sign of the correlation value determines the embedded watermark bits. If the correlation value is negative the embedded watermark bit is assumed to be 0 and if it is positive the correlation value is negative the embedded watermark bit is assumed to be 0 and if it is positive the correlation valuewatermark is negative bit theisembedded watermark bit is assumed be 0 and if it is positive the below in embedded considered to be 1. Embedder andto decoder outputs are shown embedded watermark bit Embedderand anddecoder decoderoutputs outputs are shown below embedded bitisisconsidered considered to to be be 1. 1. Embedder are shown below in in Figurewatermark 2. Figure 2. Figure 2.
Figure 2. Embedding and Decoding procedure. 2. Embedding and Decoding procedure. Figure Figure 2. Embedding and Decoding procedure.
4. Performance Evaluation and Simulation ResultsResults 4. Performance Evaluation and Simulation 4.1. Performance Measures 4.1. Performance Measures The performance of the algorithm is measured and compared based on the following metrics. The performance of the algorithm is measured and compared based on the following metrics. Mean Square Error (MSE), is the measure of the deviation of watermarked image from the original Mean Square Error (MSE), is the measure of the deviation of watermarked image from the original cover image, which is given by,
cover image, which is given by, ∑ ∑ ( ( , )− ( , ))² MSE = (6) ∑ ∑ ( ( , )− ( , ))² MSE = (6) where C is the original cover image intensity, WI is watermarked image Intensity, m and n are the where C is the original intensity, WI isRatio watermarked image Intensity, m and n are the width and height of the covercover image.image Peak Signal to Noise (PSNR) is the measure of similarity width and height of the cover image. Peak Signal to Noise Ratio (PSNR) is the measure of similarity of the original cover image and the watermarked image. It is the direct measure of perceptual quality.
J. Imaging 2017, 3, 46
6 of 13
4. Performance Evaluation and Simulation Results 4.1. Performance Measures The performance of the algorithm is measured and compared based on the following metrics. Mean Square Error (MSE), is the measure of the deviation of watermarked image from the original cover image, which is given by, MSE =
1 mn
n
m
∑ ∑ (C(i, j) − W I (i, j))2
(6)
i =1 j =1
where C is the original cover image intensity, WI is watermarked image Intensity, m and n are the width and height of the cover image. Peak Signal to Noise Ratio (PSNR) is the measure of similarity of the original cover image and the watermarked image. It is the direct measure of perceptual quality. PSNR = 20 log
(max(max(C ))) √ MSE
(7)
To quantify the quality of the extracted watermark the Bit Error Rate (BER) is used as indication of the similarity of the original binary watermark and extracted watermark. The BER counts the bit by bit difference between the original watermark and the extracted watermark, thus indicating the efficiency and robustness of the watermarking algorithm against image distortions and attacks. 4.2. Computational Complexity and Execution Time A 512 × 512 sized Lena image is watermarked with a 64 × 64 sized watermark and the time taken for embedding and decoding using the proposed algorithm is measured on HP Laptop EliteBook 8470p with Windows 7 Operating System, 8 GB RAM, clock speed of 2.8 GHz with Intel i5 core CPU. It is found that the total time consumed by the algorithm is 0.665 s. Compared to DCT based algorithms [3] and FWHT based algorithms, the proposed algorithm takes much less time. The time taken by different algorithms for embedding the same sized watermark is compared in the Table 1 given below. Table 1. Execution Time for Different Algorithms. Existing FWHT Algorithms and Similar DCT Algorithm
Time for Embedding and Decoding (Seconds)
Samee (2013) S. Maity et al. (2012) M.K. Kundu et al. (2011) Anthony et al. (2003) Proposed Algorithm
1.9 3 7.5 3 0.665
4.3. Perceptual Quality Compared to DCT based techniques, the proposed algorithm shows less perceptual quality. When we watermark the 512 × 512 as shown in Figure 3, with a 32 × 32 bit of watermark, proposed technique yields a PSNR value of 41.8 dB and similar DCT technique [3] shows a PSNR value of 50.8dB. Even though the PSNR of the proposed algorithm is numerically less, however, the resultant perceptual quality for PSNR of 41.8 dB is high.
4.3. Perceptual Quality Compared to DCT based techniques, the proposed algorithm shows less perceptual quality. When we watermark the 512 × 512 as shown in Figure 3, with a 32 × 32 bit of watermark, proposed technique yields a PSNR value of 41.8 dB and similar DCT technique [3] shows a PSNR value of J. Imaging Even 2017, 3,though 46 7 of 13 50.8dB. the PSNR of the proposed algorithm is numerically less, however, the resultant perceptual quality for PSNR of 41.8 dB is high.
(a)
(b)
(c)
Figure 3. (a) Cover Image (b) Watermark (c) Decoder output. Figure 3. (a) Cover Image (b) Watermark (c) Decoder output.
4.4. Robustness Analysis 4.4. Robustness Analysis FWHT is resilient to low quality compression compared to other common transforms like DCT FWHT is resilient to low quality compression compared to other common transforms like DCT and DWT. Lena image, of size 512 × 512 shown in Figure 4 below is watermarked with a watermark and DWT. Lena image, of size 512 × 512 shown in Figure 4 below is watermarked with a watermark of size 32 × 32. Figure 5 below shows that as the compression quality decreases, the extracted of size 32 × 32. Figure 5 below shows that as the compression quality decreases, the extracted watermark is getting distorted and the BER at the extractor increases. Compared to the other watermark is getting distorted and the BER at the extractor increases. Compared to the other algorithms, algorithms, the proposed algorithm is able to withstand low quality compression up to a quality the proposed algorithm is able to withstand low quality compression up to a quality factor of 15. factor of 15. Comparison of robustness of the proposed algorithm and DCT has been done in the case Comparison of robustness of the proposed algorithm and DCT has been done in the case of various of various attacks as shown in Table 2. The results reinforce the fact that proposed technique is more attacks as shown in Table 2. The results reinforce the fact that proposed technique is more robust to all robust to all the tested attacks compared to the similar DCT algorithm. the tested attacks compared to the similar DCT algorithm.
Imaging2017, 2017,3,3,46 46 J.J.Imaging
of12 12 77of
(a) (a)
(b) (b)
Figure4. 4.(a) (a)Lena LenaImage Image(b) (b)32 32××32 32watermark. watermark. Figure Figure 4. (a) Lena Image (b) 32 × 32 watermark. CompressionQuality QualityVs VsBER BER Compression
50 50
BER BER (%) (%)
40 40 30 30 20 20
BER-FWHT -FWHT BER BER- -DCT DCT BER
10 10 00 -10 00 -10
50 50
100 100
CompressionQuality Quality Compression
Figure 5. Compression Quality Vs BER for proposed technique (FWHT) and for similar DCT technique. Figure5.5.Compression CompressionQuality QualityVs VsBER BERfor forproposed proposedtechnique technique(FWHT) (FWHT)and andfor forsimilar similarDCT DCTtechnique. technique. Figure
Table2. 2.Extracted Extractedwatermark watermarkAfter AfterDifferent DifferentAttacks. Attacks. Table
Attacks/Distortions(Level) (Level) Attacks/Distortions Original Original Scaling (6 (6 times) times) Scaling GaussianNoise(μ Noise(μ==0, 0,σ² σ²==0.005) 0.005) Gaussian GaussianNoise(μ Noise(μ==0.01, 0.01,σ² σ²==0.005) 0.005) Gaussian GaussianNoise(μ Noise(μ==0, 0,σ² σ²==0.05) 0.05) Gaussian
DCTTechnique Technique(Samee) (Samee) FWHT FWHTTechnique Technique DCT
BER (%) BER BER (%) (%) BER BER (%) BER (%) BE BER (B( BER BE
BER (%) BER (B(( BER BER BE BER BER (%) BER (%) BER (%) BE
40 BER --FWHT -FWHT 20 30 20 DCT BER DCT BER BER ---FWHT 20 20 BER -FWHT - DCT BER BER DCT 30 30 BER 20 BER --FWHT DCT BER -FWHT - DCT 30 10 10 30 20 20 BER -FWHT - DCT BER 10 30 10 BER BER -FWHT - DCT 20 20 30 BER -FWHT - DCT BER 10 10 DCT 30 BER -FWHT 10 BER DCT 10 20 -- DCT BER --FWHT DCT BER 10 10 20 BER -FWHT DCT 20 BER DCT BER BER --FWHT 10 20 BER -FWHT 0 0 20 10 10 BER -FWHT BER -FWHT --FWHT DCT 10 10 20 20 000 10 000 BER BER -FWHT - DCTBER BER - DCT 20 BER - DCT BER - DCT 0 0 00 00-10 50 100 50 100 10 10 BER - DCT BER BER -- DCT DCT 50 50 100 100 0 10 -10 10 00 00 BER - DCT 50 50 100 -10 -10 00 0 10 10 0-10 50 Quality 100 100 50 100 -10 10 0 50 100 0 50 100 -10 Compression -10 0 Compression Quality 0 50 100 -10 Compression Compression Quality Quality -10 00 00 Compression 50 50 Quality 100 100 Compression -10 50 Quality 100 00 00-10 50 100 -10 Compression Quality J. Imaging 2017, 3, 46 8 of 13 Compression Quality -10 -10 Compression Quality Compression Quality 50 0 0 00 00-10 Compression Quality 50 100 100 0 50 100 Compression Quality Compression Quality 50 100 Compression Quality -10 Compression Quality -10 0 50 100 -10 0-10 50 100 (FWHT) 0Vs 50 Quality 100 Compression Figure 5. Quality BER for proposed technique (FWHT) and similar DCT 0Vs 50 100 5. BER for proposed technique for DCT -10 Quality FigureFigure Figure 5. Compression Compression 5. Compression Compression Quality Quality Vs BER Vs for BER proposed for proposed technique technique (FWHT) (FWHT) and for forand and similar for similar similar DCT technique. technique. DCT technique. technique. -10 Compression Quality Compression Quality -10 Compression Quality Figure 5. 5. Compression Quality Quality Vs Vs for BER proposed for proposed technique technique (FWHT) (FWHT) and similar for similar DCT DCT technique. Compression Quality FigureFigure 5. Compression Compression Quality Vs BER BER for proposed technique (FWHT) and for forand similar DCT technique. technique.
Figure 5. Quality Vs BER for proposed technique (FWHT) and for DCT Compression Quality Compression Quality Figure 5. Compression Quality Vs BER for proposed technique (FWHT) and for similar DCT technique. Compression Quality Figure 5. Compression Compression Quality Vs BER for proposed technique (FWHT) and for similar similar DCT technique. technique. Compression Quality Table 2.Vs Extracted watermark After Different Attacks. Figure 5. Compression Quality Vs BER for proposed technique (FWHT) and for similar DCT technique. 5. BER for technique (FWHT) and for DCT FigureFigure 5. Compression Compression QualityQuality Vs BER BER for proposed technique (FWHT) and for for similar DCT technique. technique. Figure 5. Compression Compression Quality Vs BER for proposed proposed technique (FWHT) and for similar similar DCT technique. technique. Figure 5. Quality Vs for proposed technique (FWHT) and similar DCT Table 2. Extracted watermark After Different Attacks. Table 2. Extracted watermark After Different Attacks. 5. Compression Vs BER for proposed technique (FWHT) for similar DCT technique. TableQuality Table 2. Extracted 2. Extracted watermark watermark After Different After Different Attacks. Attacks. FigureFigure 5. Compression Compression Quality VsExtracted BER for proposed technique (FWHT) and forand similar DCT technique. technique. Table Table 2. 2. Extracted watermark watermark After Different After Different Attacks. Attacks. Figure 5. Compression Vs BER for proposed technique (FWHT) and for similar DCT technique. Figure 5. Quality Vs BER for proposed technique (FWHT) and for similar DCT TableQuality 2. Extracted watermark After Different Attacks. Table 2. Extracted watermark After Different Attacks. Figure 5. Compression Quality Vs BER for proposed technique (FWHT) and for similar DCT technique. Table 2. watermark After Different Attacks. Table 2. Extracted watermark After Different Attacks. FigureFigure 5.Attacks/Distortions Compression Quality VsExtracted BER for proposed technique (FWHT) and forand similar DCT technique. Figure 5. Compression Compression Quality Vs BER for proposed proposed technique (FWHT) and for similar similar DCT technique. 5. Quality Vs BER for technique (FWHT) for DCT technique. (Level) DCT Technique (Samee) FWHT Technique Table 2. Extracted watermark After Different Attacks. Figure 5. Compression Quality Vs BER for proposed technique (FWHT) and for similar DCT technique. Table 2. Extracted watermark After Different Attacks. Table 2. Extracted watermark After Different Attacks. Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Table 2. Extracted watermark After Different Attacks. Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Table 2. Extracted watermark After Different Attacks. Attacks/Distortions Attacks/Distortions (Level) (Level) DCT DCT Technique Technique (Samee) (Samee) FWHT FWHT Technique Technique Attacks/Distortions Attacks/Distortions (Level) (Level) DCT Technique DCT Technique (Samee) (Samee) FWHT FWHT Technique Technique 2. Extracted watermark After Different Attacks. Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Table Table 2.(Level) Extracted watermark AfterTechnique Different Attacks. Table 2. Extracted watermark After Different Attacks. Attacks/Distortions DCT Technique (Samee) FWHT Technique Attacks/Distortions (Level) DCT (Samee) FWHT Technique Table 2. Extracted watermark After Different Attacks. Table 2. Extracted watermark After Different Attacks. Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Table 2. Extracted watermark After Different Attacks. Table 2. Extracted watermark After Different Attacks. Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Original Table 2. Extracted watermark After Different Attacks. Original Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Original Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Table 2. Extracted watermark After Different Attacks. Original Original Original Original
Attacks/Distortions (Level) Technique (Samee) FWHT Technique Original Original Attacks/Distortions (Level) DCT DCT Technique (Samee) FWHT Technique Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Original Original Attacks/Distortions (Level) Technique (Samee) FWHT Technique Attacks/Distortions (Level) DCT DCT Technique (Samee) FWHT Technique Original Original Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Original Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Original Original Attacks/Distortions (Level) DCT Technique (Samee) FWHT Technique Original Scaling (6 times) Scaling (6 times) Scaling (6 times) Scaling Scaling (6Original times) (6 times) times) Original Scaling Scaling (6 times) (6 Original Original Scaling (6Original times) Scaling (6 times) times) Original Scaling (6 times) Scaling (6 Original Scaling (6 times) Scaling (6 Scaling (6 times) times) Scaling (6 times) times) Scaling (6 Scaling (62σ²times) Gaussian Noise(μ == 0, Gaussian Noise(μ ====0.005) 0, σ² Scaling (6=times) times) Scaling Gaussian Gaussian Noise(μ Noise(μ 0,σ(6 σ²=times) 0,0.005) 0.005) σ² === 0.005) 0.005) Scaling (6 Gaussian Noise (µ Scaling (6σ² times) Gaussian Gaussian Noise(μ Noise(μ ==0,0, ====0, 0.005) σ² 0.005) Scaling (6 times) Gaussian Noise(μ 0, σ² 0.005) Gaussian Noise(μ σ² Scaling (6 times) Gaussian Noise(μ = 0, σ² = Scaling (6 times) Gaussian Noise(μ ==0, 0,0.005) σ² == 0.005) 0.005) Gaussian Noise(μ = 0, σ² 0.005) Gaussian Noise(μ σ² Gaussian Noise(μ 0, σ² σ²====0, 0.005) Gaussian Noise(μ 0,0.005) σ² == 0.005) 0.005) Gaussian Noise(μ == 0, Gaussian Noise(μ = 0, σ² = 0.005) Gaussian Noise(μ 0, σ² σ² 0.005) Gaussian Noise(μ == ==0.01, 0.005) Gaussian Noise(μ ==0.01, σ² Gaussian Noise(μ ==2σ² σ² 0.005) Gaussian Gaussian Noise(μ Noise(μ 0.01, σ² 0.01, 0.005) σ²== ===0.005) 0.005) Gaussian Noise(μ 0, ===0, 0.005) Gaussian Noise(μ 0,==0.005) σ² 0.005) Gaussian Gaussian Noise(μ Noise(μ ==σ² 0.01, 0.005) σ² Gaussian Noise(μ =0.01, 0, σ² σ² Gaussian Noise (µ == σ = Gaussian Noise(μ =0.01, 0.01, =0.005) 0.005) Gaussian Noise(μ =σ² 0,==0.005) σ² = 0.005) Gaussian Noise(μ 0.01, σ² = 0.005) 0.005) Gaussian Noise(μ = 0, = Gaussian Noise(μ = 0.01, σ² 0.005) Gaussian Noise(μ = 0.01, σ² = 0.005) Gaussian Noise(μ = 0.01, σ² = 0.005) Gaussian Noise(μ 0.01, σ² Gaussian Noise(μ 0.01,==σ² σ² 0.005) Gaussian Noise(μ 0.01, σ² == 0.005) 0.005) Gaussian Noise(μ == 0.01, == 0.005) Gaussian Noise(μ = σ² 0.01, σ² = 0.005) Gaussian Noise(μ 0.01, 0.005) Gaussian Noise(μ 0.01, σ² 0.005) Gaussian Noise(μ == 0, Gaussian Noise(μ ====0, σ² Gaussian Noise(μ === 0.01, σ² ===0.05) 0.005) Gaussian Gaussian Noise(μ Noise(μ 0, ==σ² σ² 0, 0.05) σ² 0.05) Gaussian Noise(μ 0.01, σ²=====0.05) 0.005) Gaussian Gaussian Noise(μ Noise(μ = 0, σ² = = 0, 0.05) σ² 0.05) Gaussian Noise(μ 0.01, σ² 0.005) Gaussian Noise(μ = 0.01, σ² = 0.005) 2 Gaussian Noise(μ = 0, σ² = 0.05) Gaussian Noise(μ = 0, σ² = 0.05) Gaussian Noise(μ ==0.01, σ² =0,0.05) 0.005) Gaussian Noise(μ = 0, σ² = Gaussian Noise (µ 0, σ = 0.05) Gaussian Noise(μ = σ² = 0.05) Gaussian Noise(μ = 0, σ² = 0.05) Gaussian Noise(μ σ² Gaussian Noise(μ 0, σ² σ²====0, 0.05) Gaussian Noise(μ 0,0.05) σ² == 0.05) 0.05) Gaussian Noise(μ == 0, Gaussian Noise(μ ==(Density—0.01) 0, σ² == 0.05) Gaussian Noise(μ = 0, σ² = 0.05) Gaussian Noise(μ 0, σ² 0.05) Salt & Pepper Noise (Density—0.01) Salt & Pepper Noise Gaussian Noise(μ = 0, σ² = 0.05) Salt &Salt Salt Pepper & Noise(μ Pepper Noise Noise Gaussian Noise(μ =(Density—0.01) 0,0.05) σ² = 0.05) Salt & Pepper & Pepper Noise Noise Gaussian =(Density—0.01) 0, σ² σ²=(Density—0.01) Gaussian Noise(μ σ² = 0.05) Salt &Salt Pepper Noise (Density—0.01) & Pepper Noise (Density—0.01) Gaussian =(Density—0.01) 0, ==0,0.05) Salt Pepper Noise (Density—0.01) Salt & Noise(μ Pepper Noise (Density—0.01) Salt& &Salt Pepper Noise (Density—0.01) Salt & Pepper Noise (Density—0.01) & Pepper Noise (Density—0.01) Salt & Pepper Noise (Density—0.01) & Pepper (Density—0.01) Salt &Salt Pepper NoiseNoise (Density—0.01) Salt & Pepper Noise (Density—0.01) Salt & Pepper Noise (Density—0.1) Salt & Pepper Noise (Density—0.01) Salt & Pepper Noise (Density—0.01) Salt & Pepper Noise (Density—0.1) Salt & Salt Pepper & Pepper Noise Noise (Density—0.1) (Density—0.1) Salt & Pepper Noise (Density—0.01) Salt & Pepper (Density—0.01) Salt & Salt Pepper & Pepper Noise Noise (Density—0.1) (Density—0.1) Salt & Pepper Noise (Density—0.01) Salt & Noise (Density—0.01) Salt& & Pepper NoiseNoise (Density—0.1) Salt & Pepper Noise (Density—0.1) Salt Pepper Noise (Density—0.01) Salt & Pepper Noise (Density—0.1) Salt &Pepper Pepper Noise (Density—0.1) Salt & Pepper Noise (Density—0.1) & Noise (Density—0.1) Salt & Pepper Noise(Density—0.1) (Density—0.1) Salt& &Salt Pepper Noise Salt & Pepper Pepper Noise (Density—0.1) Salt Pepper Noise (Density—0.1) Salt & Pepper Noise (Density—0.1) Salt & &Salt Pepper Noise (Density—0.1) & Pepper Sharpen (3 ×× 3) Sharpen ×× 3) Salt Pepper Noise (Density—0.1) Sharpen Sharpen (3Noise 3)(3 (3(Density—0.1) 3) & Pepper Noise (Density—0.1) Sharpen Sharpen (3 ×× 3) ×× 3) Salt & &Salt Pepper Noise (Density—0.1) Salt & Pepper Sharpen (3Noise 3)(3 Sharpen (3(Density—0.1) 3) Salt Pepper Noise (Density—0.1) Sharpen (3 × 3) Sharpen (3 × 3) Sharpen (3 × 3) Sharpen Sharpen (3 × 3)(3 Sharpen (3 ×× 3) 3) Sharpen (3(3××3)3)(3 × 3) Sharpen Sharpen Sharpen (3 3)(3 ×× 3) Blur (2 ×× 2) Blur ×× 2) Sharpen (3 ×××(2 3) BlurSharpen (2 Blur 2) (2 2) Sharpen (3 3) Blur (2 Blur ×× 2) (2 ×× 2) Sharpen (3 3)(3 Sharpen × Blur (2 (2 2) Sharpen (3 × 3) Blur (2 2) Blur × 2) Blur (2 × 2)3) Blur (2 × 2) Blur Blur (2 × 2)(2 Blur (2 ×× 2) 2) Blur (2 ××× 2) Median filtering (3 3) Blur (2 × 2) Median filtering Blur (2 × 2) Blur (2 2) Median Median filtering filtering (3 × 3)(3 (3 ××× 3) 3) Blur (2Blur ×filtering 2)(2 2) Median Median filtering (3 3) Blur (2 ×filtering 2)(2 Median filtering (3×××××2) 3)(3 Median (3 ×× 3) 3) Blur Blur (2 × 2) Median filtering (3 3) Median filtering (3 3) Median filtering (3 Median filtering Median filtering (3 ×× 3) 3)(3 Median filtering (3 ×× 3) 3) Median filtering (3 × 3)5) Gaussian filtering (5 5) Gaussian filtering (5 Median filtering (3 3) Median filtering (3 Gaussian Gaussian filtering filtering (5××3)×××3) 5) (5×××××3) 5) Median filtering Median filtering (3(3 × Median filtering (3 3)5) Gaussian Gaussian filtering filtering (5 5) (5 Median filtering (3 × 3) Median filtering (3 Gaussian filtering (5 × 5) Gaussian filtering (5 Median filtering (3 × 3) Gaussian filtering (5 × 5) Gaussian filtering (5×××3)5) 5) Gaussian filtering (5 × 5) Gaussian filtering Gaussian filtering (5 × 5)(5 Gaussian filtering (5 ×× 5) 5) Gaussian filtering (5 × 5) Gaussian filtering (5 ×× 5) 5)(5 Crop (50%) Gaussian filtering Crop (50%) Gaussian filtering (5 Crop Crop (50%) (50%) Gaussian filtering (5 ×× 5) 5) Crop Crop (50%) Gaussian filtering (5 5)(5 Gaussian filtering (5(50%) × ×× 5)5) Crop (50%) Gaussian filtering × 5) Crop (50%) Gaussian filtering (5 Crop (50%) Crop (50%) Crop (50%) Crop Crop (50%)(50%) Crop (50%) (50%) Crop Crop (50%) Crop (50%) Crop (50%) Crop (50%) Crop(50%) (50%)(50%) Crop Crop (50%) Crop
4.5. Effect of Spreading Code Length on Robustness The 512 × 512 Lena image is watermarked with a 16 × 16 watermark. For watermarking, three different lengths spreading codes, 512 bits, 256 bits and 128 bits are used. When we use 512 bits spreading code the whole row is used for embedding and for 256 bits and 128 bits only half and quarters of the selected coefficients respectively in a row are used for embedding. The BER performance is shown in Figure 6. Robustness of the algorithm in each of the above cases is noted and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression (JPEG) compared to others and in the case of detailed attack analysis also 512 bit spreading codes show best robustness against all the attacks and 128 bit code the least.
BER (%) BER (%) BER (%)
BER (%)
BER BER (%) BER (%)(%) BER BER (%) (%)
BER (%) BER (%) BER (%) BER (%) BER (%) BER (%) BER (%)
BER (%)BER (%) BER (%)
BER (%)
BER (%)BER BER BER (%) (%) (%) BER BER (%) (%) BER (%) BER (%)
spreading code whole is used for embedding and 256 bits and 128 bits only half and different lengths spreading 512 bits, bits are used. When we use 512 bitsbits and 128 spreading code the whole row iscodes, used for embedding and forfor 256 bits and 128 bits only half and spreading code the whole row isrow used for spreading embedding code and the whole for row bits is and used 128 for bits embedding half and for 256 b The ××512 512 Lena image The 512 is watermarked 512 Lena image with is a×abits watermarked 16 ×aand 16 watermark. with a watermarking, For 16 ×watermarking, 16only watermark. three For watermarking, thre The 512 ×512 512 Lena image is watermarked with a256 16 watermark. For three The ×the 512 Lena image isis watermarked with ×128 16 watermark. For watermarking, three The 512 512 Lena image watermarked 16 × 256 16 watermark. For watermarking, three spreading code the whole row is×codes, used for embedding and for 256 bits and 128 bits only half and spreading code the whole row iscodes, used for embedding and for 256 bits and 128 bits only half and spreading code the whole row used foriswith embedding and for 256 bits and 128 bits only half and different lengths spreading 512 bits, 256 bits and 128 bits are used. When we use 512 bits different lengths spreading 512 bits, 256 bits and 128 bits are used. When weonly use 512 bits different lengths spreading codes, 512 bits, 256 bits and 128 bits are used. When weonly use 512 bits spreading code the whole row is used for embedding and for 256 bits and 128 bits half and spreading code the whole row is used for embedding and for 256 bits and 128 bits only half and The 512 × 512 Lena image The 512 is × watermarked 512 Lena image with a watermarked 16 × 16 watermark. with a For 16 × watermarking, 16 watermark. three For watermarking, three spreading code the whole row is used for embedding and for 256 bits and 128 bits half and quarters of the selected coefficients respectively in a row are used for embedding. The BER spreading code the whole row is used for embedding and for 256 bits and 128 bits only half and quarters of the selected respectively a512 row are used for embedding. The BER quarters of the selected coefficients respectively quarters of inthe a in selected row are coefficients used for embedding. respectively The in aused. BER row arebits used for 512 embe different lengths spreading different codes, lengths 512 spreading bits, 256 codes, bits and 128 bits, bits 256 are bits and When 128 bits we are use 512 bits When we use bi different lengths spreading codes, 512 bits, 256 bits and 128 bits are used. we use 512 different lengths spreading codes, 512 bits, 256 bits and 128 bits are used. When weWhen use 512 bits different lengths spreading codes, 512 bits, 256 bits and 128 bits are used. When we use 512 bits 4.5. Effect of Spreading 4.5. Code Effect Length ofcoefficients Spreading on Robustness 4.5. Code Effect Length of Spreading on Robustness Code Length onused. Robustness quarters ofcode the selected coefficients respectively in row are used for embedding. The BER quarters ofcode the selected coefficients respectively inembedding row are used for embedding. The BER quarters of the selected coefficients respectively in aa256 row are used for embedding. The BER spreading code the whole row is used for and for 256 bits and 128 bits only halfwe and spreading the whole row is used for embedding for 256 bits and 128 bits only half and spreading the whole row is used for embedding and for 256 bits and 128 bits only half and quarters of the selected coefficients respectively in aa512 row are used for embedding. The BER quarters of the selected coefficients respectively in row are used for embedding. The BER different lengths spreading different codes, lengths 512 spreading bits, 256 codes, bits and 128 bits, bits are bits used. and When 128 bits we are use used. 512 bits useonly 512 bitsabo quarters of the selected coefficients respectively in aaand row are used for embedding. The BER performance is shown in Figure 6. Robustness of the algorithm in each of the above cases is noted quarters the selected coefficients respectively in afor row are used for embedding. The BER performance isof shown in Figure 6. Robustness ofembedding the algorithm in each of the above cases isWhen noted performance is shown in Figure 6. Robustness performance of the is algorithm shown in in Figure each of 6. the Robustness above cases of the is noted algorithm in each of the spreading code the spreading whole row code is used the for whole embedding row is used and for embedding 256 bits and and 128 for bits 256 only bits half and and 128 bits half an spreading code the whole row is used for and for 256 bits and 128 bits only half and spreading code the whole row is used for embedding and for 256 bits and 128 bits only half and spreading code the whole row is used for embedding and for 256 bits and 128 bits only half and performance isofspreading shown in Figure Figure 6.respectively Robustness ofalgorithm the algorithm in used each of the above cases is128 noted performance is shown inimage Figure 6. Robustness of the algorithm each of the above cases isand noted performance is shown in Figure 6. Robustness of algorithm in each of the above cases is noted quarters the selected coefficients respectively in ain row are used for embedding. The BER quarters of512 the selected coefficients respectively in athe row are for embedding. The BER quarters ofcode the selected coefficients in aand row are used for embedding. The BER performance shown in Figure 6. Robustness of the in each of the cases is noted performance is shown in Figure 6. Robustness of the algorithm in each of the above cases is noted spreading the whole row code is used the for whole embedding row is used for for embedding 256 bits and and 128 for bits 256 only bits half and bits only halfthree andwat performance is shown 6. Robustness of the algorithm in each of the above cases is The 512 ×is Lena The 512 is ×watermarked 512 Lena image The with 512 is a×watermarked 16 512 ×spreading Lena 16 watermark. image with is ain watermarked For 16 ×above watermarking, 16 watermark. with acompression three For 16 ×noted watermarking, 16 watermark. For and tabulated Table 3. It can be seen that the 512 bit code is more resistant to performance is shown in Figure 6. Robustness of the algorithm each of the above cases is noted and tabulated Table 3. Itin can be seen that the 512 bit spreading code is more resistant toThe compression and tabulated Table 3. It can be seen that and the tabulated 512 bit spreading Table 3. code It can is be more seen resistant that the to 512 compression bit spreading code is more resistan quarters of the selected quarters coefficients of the selected respectively coefficients in a row respectively are used in for a embedding. row are used The for BER embedding. The BE quarters of the selected coefficients respectively in a row are used for embedding. BER quarters of the selected coefficients respectively in a row are used for embedding. The BER quarters of the selected coefficients respectively in a row are used for embedding. The BER and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression performance is shown in Figure 6. Robustness of the algorithm in each of the above cases is noted performance is shown in Figure 6. Robustness of the algorithm in each of the above cases is noted performance is shown in Figure 6. Robustness of the algorithm in each of the above cases is noted J. Imaging 2017, 3, 46 J. Imaging 2017, 3, 46 J. Imaging 2017, 3, 46 8 of 12 8 of 12 and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression quarters of the selected quarters coefficients of the selected respectively coefficients in a row respectively are used in for a embedding. row are used The for BER embedding. The BER and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression different lengths spreading different codes, lengths 512 spreading bits, different 256 codes, bits lengths and 512 spreading 128 bits, bits 256 are codes, bits used. and 512 When 128 bits, bits we 256 are use bits used. 512 and bits When 128 bits we are use used. 512 bits When (JPEG) compared to others and in the case of detailed attack analysis also 512 bit spreading codes and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression (JPEG) compared to others and in the ofthe detailed attack analysis also 512 bit spreading codes (JPEG) compared to shown others and in the (JPEG) of detailed compared attack others analysis and inin 512 case bit of detailed codes attack also 512 bit performance is in Figure 6. iscase Robustness shown in Figure of the 6.to algorithm Robustness in each of the of the above incases each is the above cases is note performance is shown 6.case Robustness of the algorithm each ofspreading the above cases isanalysis noted performance is shown inperformance Figure 6. Figure Robustness of algorithm in each of the above cases is noted performance is shown in Figure 6. Robustness ofof the algorithm inalso each ofalgorithm the above cases isofnoted noted (JPEG) compared to others and in the case of detailed attack analysis also 512 bit spreading codes (JPEG) compared to others and in the case of detailed attack analysis also 512 bit spreading codes (JPEG) compared to others and in the case detailed attack analysis also 512 bit spreading codes J. Imaging 2017, 3, 46 9 of 13 and tabulated Table 3. Itin can be seen that the 512 bit spreading code is more resistant to compression and tabulated Table 3.others It can be seen that the 512 bit spreading code is more resistant to compression and tabulated Table 3. It can be seen that the 512 bit spreading code isbits more resistant to compression (JPEG) compared to others and the case of detailed attack analysis also 512 bit spreading codes (JPEG) compared to others and in the case of detailed attack analysis also 512 bithalf spreading codes performance is shown performance in Figure 6. isall Robustness shown in Figure of the 6. algorithm Robustness in each of the of algorithm the above in cases each isbit of noted the above cases isand noted (JPEG) compared to and in the case of detailed attack analysis also 512 bit spreading codes spreading code the spreading whole row code is used the for spreading whole embedding row code is128 used and the for whole for embedding 256 row is and used and 128 for for bits embedding 256 only bits and and 128 for bits 256 only bits half and 128 b show best robustness against all the attacks and 128 bit code the least. (JPEG) compared to others and in the case of detailed attack analysis also 512 bit spreading codes show best robustness against the attacks and bit code the least. showtabulated best robustness against all the attacks show and best 128 robustness bit code the against least. all the attacks and 128 code the least. and tabulated Table and 3. It tabulated can be seen Table that 3. the It 512 can bit be seen spreading that the code 512 is bit more spreading resistant code to compression is more resistant to compressio and tabulated Table 3. It can be seen that the 512 bit spreading code is more resistant to compression and Table 3. It can be seen that the 512 bit spreading code is more resistant to compression and tabulated Table 3.against It can be seen that the 512 bit spreading code isleast. more resistant to compression show best robustness all the attacks and 128 bit code the least. show best robustness against all the attacks and 128 bit code the least. show best robustness against all the attacks and 128 bit code the 4.5. Effect ofrobustness Spreading 4.5. Code Length ofothers Spreading on Robustness 4.5. Code Effect Length ofdetailed Spreading oncode Robustness Code Length on Robustness (JPEG) compared to and in the case of detailed attack analysis also 512 bit spreading codes (JPEG) compared toEffect others and in the of attack analysis also 512 bit spreading codes (JPEG) compared torobustness and in the case of detailed attack analysis also 512 bit spreading codes show best against all the attacks and 128 bit the least. show best against all the attacks and 128 bit code the least. and tabulated Table and 3.others It tabulated can be seen Table that 3. the Itcase 512 can bit be seen spreading that the code 512 isof bit more spreading code to compression is more resistant to compression show best robustness against all the attacks and 128 bit code the least. quarters of the selected quarters coefficients of the selected respectively quarters coefficients of in the selected row respectively are coefficients used infor aresistant embedding. row respectively are used The in for aBER embedding. row used TheforBER embe show best against the attacks and 128 bit code the least. (JPEG) compared to others compared and in the case others of and in the attack case analysis detailed also attack 512 bit analysis spreading also codes 512are bit spreading code (JPEG) compared to others and into the case ofaattack detailed attack analysis also 512 bit spreading codes (JPEG) compared torobustness others and in the case of detailed analysis also 512 codes (JPEG) compared to(JPEG) others and inall the case of detailed detailed attack analysis alsobit 512spreading bit spreading codes show best robustness against all the attacks and 128 bit code the least. show best robustness against all the attacks and 128 bit code the least. show best robustness against all the attacks and 128 bit code the least. (JPEG) compared to (JPEG) others compared and in the to case others of detailed and in the attack case analysis of detailed also attack 512 bit analysis spreading also codes 512 bit spreading codes performance is shown performance in Figure 6. is Robustness shown performance in Figure of the 6. is algorithm Robustness shown in in Figure each of the of 6. algorithm the Robustness above in cases each of the is of noted algorithm the above in cases each is of noted the abo The × robustness 512 Lena image The 512 is × watermarked 512 Lena image The with 512 is a × watermarked 16 512 × Lena 16 watermark. image with is a watermarked For 16 × watermarking, 16 watermark. with a three For 16 × watermarking, 16 watermark. three For wat Compression Quality Vs BER for Different best show against best all robustness the attacks against and 128 all bit the code attacks the and least. 128 bit code the least. show best robustness against all the attacks and 128 bit code the least. showshow best512 robustness against all the attacks and 128 bit code the least. show best robustness against all the attacks and 128 bit code the least. Compression Quality Vs BER for Different Compression Quality Vs BER for Different Compression Quality Vs BER for Different Compression Quality Vs BER for Different Compression Quality Vs BER for Different Compression Quality Vs BER for Different Spreading Code Length show best robustness show against best all robustness the attacks against and 128 all bit the code attacks the and least. 128 bit code the least. and tabulated Table and 3. It tabulated can be seen Table that and 3. the It tabulated 512 can bit be seen spreading Table that 3. the code It 512 can is bit be more seen spreading resistant that the code to 512 compression is bit more spreading resistant code to compression is more resistan Compression Quality Vscodes, BER for Different Compression Quality Vs BER for256 Different Spreading Code Length Spreading Code Length Spreading Code Length Compression Quality Vsand BER for Different different lengths spreading differentcodes, lengths 512 spreading bits, different 256 bits lengths 512 spreading 128 bits, bits are codes, bits used. and 512 When 128 bits, bits we 256 are use bits used. 512 and bits When 128 bits we are use used. 512 bits When Compression Quality Vs BER for Different Spreading Code Length Spreading Code Length Spreading Code Length Spreading Code Length Spreading Code Length Spreading Code Length Compression Quality Vsfor BER for Different Compression Vs BER Different (JPEG) compared tospreading (JPEG) othersrow compared andcode inused the to case (JPEG) others ofQuality detailed compared and the attack to case others analysis of detailed and the attack 512 case bit analysis spreading of detailed codes attack 512 analysis spreading codes 512 Spreading Code Length Compression Quality Vs BER for Different spreading code the whole is the for spreading whole embedding row code isinfor used and the for whole for embedding 256 row bitsalso isin and used and 128 for bits embedding 256 only bitsalso half and and 128bit for bits 256 only bitsalso half and and 128bitb 50 50 50 50 Compression Quality Vs BER Compression for Different Quality Vs BER forfor Different Compression Quality Vs BER Different Compression Quality Vs BER for Different Compression Quality Vs BER for Different Spreading Code Length Spreading Code Length Spreading Code Length 50 50 50 show best robustness show against best all robustness the attacks show against and best 128 all robustness bit the code attacks the against and least. 128 all bit the code attacks the and least. 128 bit code the least. 50 50 Spreading Code Length Spreading Code Length Compression Quality Vs BER Compression fora Different Quality Vs BER fora Different quarters of the selected quarters coefficients of50 the50selected respectively quarters coefficients of inthe selected row respectively are coefficients used infor embedding. row respectively are used The inforaBER embedding. row are used TheforBER embe Spreading Code Length Spreading Code Length Spreading Code Length 512 bits 4040 50 Spreading Code Length Spreading40 Code512 Length bits 512 bits 512 bits 50 50 40 performance is shown performance in 40 Figure 6. is Robustness shown performance in Figure of the 6. is algorithm Robustness shown in in Figure each of the of 6. algorithm the Robustness above in cases each of the is of noted algorithm the above in cases each is of noted the abo 40 40 512 bits 512 bits 512 bits 50 40 50 40 40 50 512 bits 512 bits 512 bits 50 50 40 256 bits 256 bits 256 bits 512 bits 256 bits 30 50 50 and tabulated Table and 3. It tabulated can be seen Table that and 3. the It tabulated 512 can bit be seen spreading Table that 3. the code It 512 can is bit be more seen spreading resistant that the code to 512 compression is bit more spreading resistant code to compression is more resistan 30 30 30Vs BER Compression Quality Vs BER Compression for Different Quality Compression Different Vs BER for Different 256 bits 512 256 bitsfor 40 256Quality 40 512 bits 512 bits 30 256 bits 256 bits bits 30 40 256 128 bits bits 30 256 bits 40 30 30 to Spreading Code40 Length Spreading Code512 Length Code Length 30 40 bits 512 bits 40 128 512 bits 512 40in 30 128 128 bits 512 bits bits 512 bits Spreading (JPEG) compared to(JPEG) others compared and the case (JPEG) others of detailed compared and in the attack to case others analysis of detailed and also in the attack 512 case bit analysis spreading of detailed alsocodes attack 512 bitanalysis spreading alsocodes 512 bit 128 bits 128 bits 128 bits 256 256 bits bits 256 40 40 2020 30 20 20 512 128 128 bits bits 512 bits 128 bits 128 30 30 bits 20 256 bits 256 bits 256 bits 20 20 bits and 256bit bits 256 bits 256 show best robustness show against bestall robustness the attacks show against and best 128 all robustness bit the code attacks the against and least. 128 all the code attacks the least. 128 bit code the least. 50 50 50 30 30 20 30 20 20 30 30 20 128 bits bits 128 bits 128 bits 256 bits 30 30 1010 20 10 10 256 128 bits 128 bits bits 128 128 128 bits bits 128 bits 20 20 10 10 10 40 40 20 40 512 10 20 10 128 bits 128 bits 10 512 512 bits 20 20 20 10 0 10 20 20Vs BER 0 100Compression Quality Compression for Different Quality 0Vs BER Compression for Different Quality Vs BER for Different 10 256 bits 256 bits 256 bits 30 30 30 10 00 10 0 Spreading 00 010 Code Length Spreading Code0 Length Spreading 10 10 00 50 100 50 50 100 100 50Code Length 100 100 0 10 0 50 100 128 bits 128 bits 128 bits 0 50 100 0 50 100 0 0 0 0 Compression 50 Quality 50 100 500 Quality Compression Quality100 100 00 Compression 50 Compression Quality 20 20 Quality 20 100 00 0 50 50 50 0 Compression Compression Quality Compression Quality 0 Compression 50 Quality 50 Quality 500Compression 100 100 100 Quality Compression 0 0 00 Compression 50 50 100 0 0 50 100 100 0 50 050 Quality 100 10 10 10 100 40 40 Compression Quality Compression Quality Compression Quality 512 bits 512 bits 512 bits Figure 6. 40 Compression Quality Vs BER in the case of different spreading code length. 0 50 0 100 50 100 Compression Quality Quality Figure Compression Quality BER in of different spreading code length. Compression Quality Compression Quality Compression Quality Figure 6. 6. Compression Quality VsVs BER thethe case ofCompression different spreading code length. Figure 6. Compression Quality Vs BER in thein Figure case of 6.case different Compression spreading Quality code Vs length. BER in the case of different spreading code l 256 spreading bits 256 bits 256 bits Figure 6. Compression Quality Vs BER in thein case of different code length. Figure 6. Compression Compression Quality VsQuality BER in thein case of different spreading code code length. Compression Quality Compression Quality Figure 6. Compression Vs BER the case of different spreading code length. 06. 0in 0spreading 30 30 30 Figure 6. Quality Vs BER the case of different code length. Figure 6. Compression Compression Quality Vs BER in the case of different spreading code length. Figure 6. Compression Quality Vs BER thein case of different spreading length. Figure Quality Vs BER the case of different spreading code length. Table 3. Extracted watermark after different attacks in the case of128different spread code128length. bits bits 128 bits 50 0 in 100 50 100 50 100 of different Figure 6. 0 Compression Quality Vs BER in the case of different spreading code length. Figure 6. watermark Compression Quality Vs BER case of different spreading code length. Figure 6.3.Compression Quality Vs BER in the case of different spreading code length. Extracted watermark after different attacks in the case different spread code length. Table 3. after attacks in the case of0of different spread code length. TableFigure 3.Table Extracted after different Table attacks 3.the inExtracted the case of watermark different spread after different code length. attacks in the case spread co 206.watermark 20in 20 Figure 6. Compression Figure Quality 6.different Vs Compression BER the Quality case of different Vs BER in spreading the case of code different length. spreading 6. Extracted Compression Quality VsQuality BER thein case of different spreading code length. Figure Compression Vs BER the case of different spreading code length. Figure 6.watermark Compression Quality Vs BER in the case of different spreading code length. 3. Extracted watermark after different attacks in the case of different spread code length. TableTable 3. Extracted after different attacks in in the case of different spread code length. Table 3. Extracted watermark after different attacks in the case of different spread code length.code length.
Compression Quality Compression Quality Compression Quality TableFigure 3. Extracted after6. different attacks in the case of different code length. Table 3.watermark Extracted watermark after different attacks in the case ofspread different spread code length. length. Table 3. watermark after different in the case of different spread code length. 6. Extracted Compression Figure Quality Vs Compression BER in theattacks Quality case of different Vs BER in spreading the case of code different length. spreading code length. Table 3. Extracted watermark after different attacks in the case of different spread code (Level) Length—512 Length—256 Length—128 Attacks (Level) Length—512 Length—256 Length—128 (Level) Length—512 Length—256 Length—128 10 10 10 Attacks (Level) Length—512 Attacks Length—256 (Level) Length—128 Length—512 Length—256 Len Table 3.Attacks Extracted watermark after different attacks in the case of different spread code length. 3.Attacks Extracted watermark after different attacks in the case of different spread code length. TableTable 3. Extracted watermark after different attacks in the case of different spread code length. Attacks (Level) Length—512 Length—256 Length—128 Attacks (Level) Length—512 Length—256 Length—128 Attacks (Level) Length—512 Length—256 Length—128 3. Extracted watermark Table 3.after Extracted different watermark attacks in after the different case of different attacks in spread the case code of length. different spread code length. Table 3.Attacks Extracted watermark after different attacks the case ofspread different spread code length. TableTable 3. Extracted watermark after different attacks in the case of different code length. Table 3.Attacks Extracted watermark after different attacks in thein case of different spread code length. (Level) Length—512 Length—256 Length—128 Attacks (Level) Length—512 Length—256 Length—128 (Level) Length—512 Length—256 Length—128 Attacks (Level) Length—512 Length—256 Length—128 TableFigure 3. Extracted watermark Table 3.after Extracted watermark attacks in after the different case of different attacks in spread the case of length. different spread code length. Original 0Original 0in 0in 6.Attacks Compression Figure Quality 6.different Vs Compression BER theFigure Quality case ofLength—512 6.different Vs Compression BER spreading the Quality case of code different Vscode length. BER in spreading the case of code different length. spreading code l Original Attacks (Level) Length—256 Length—128 Attacks (Level) Length—512 Length—256 Length—128 Original Original (Level) Length—512 Length—256 Length—128
Original Original Original Attacks (Level) Length—512 Attacks (Level)Attacks Length—512 Length—256 Length—128 Attacks (Level) Length—512 Length—128 Attacks (Level) Length—512 Length—256 Length—128 Original Original Original 0 Original 50 (Level) 0 (Level) 100 Length—256 50 Length—256 0 Length—512 100 Length—128 50 Length—256 100 Length—128 Attacks Original (Level) Attacks Length—512 Length—256 Length—512 Length—128 Length—256 Length—128 Original Original Scaling (6)Compression Table 3. Extracted watermark Table 3.(6) after Extracted different watermark Table attacks after Extracted the different case of watermark different attacks spread after the different case code of length. different attacks spread the case code of length. different spread co Quality3.in Compression Quality Compression Quality in Scaling Scaling (6)Original Scaling (6) in Scaling (6) Original Original Original Original Scaling (6) (6) Scaling (6)Scaling Scaling (6) Scaling (6) Original Scaling (6) Original Scaling (6) Attacks (Level) Attacks (Level) Length—512 Attacks Length—256 (Level) Length—512 Length—128 Length—256 Length—512 Length—128 Length—256 Len Scaling (6) Scaling Scaling (6) 2σ=2 0.05) Gaussian noise (μ == 0, = 0.05) 2 0, Gaussian noise (μ =(6) σ Gaussian noise (μ = Scaling 0, σ(6) Gaussian noise (μ = 0,Quality σ2 =of 0.05) 0.05) 2Quality 2Vs Scaling (6) (6) of6.different Scaling (6) 20.05) 22 =Scaling (6) Scaling Figure 6. Compression Figure 6. Compression BER in the Figure Quality case Vs Compression BER in spreading the case code different Vs length. BER in spreading the case of code different length. spreading code l Gaussian noise (μ = 0, σ = 0.05) Gaussian noise (µ = 0, σ = Gaussian noise (μ = 0, σ = 0.05) Gaussian noise (μ = 0, σ 0.05) 2 2 Gaussian noise (μ (6) =noise 0, =(μ Gaussian noise (μ0.05) 0, σ2 == 0.05) 0.05) Gaussian noise (μσ= 0, σ== 0, =Scaling 0.05) Gaussian σ Scaling (6) Original Original Original 2 2 2 2 Gaussian noise σ00.05) = 0.05) Gaussian noise (μ(μ σ2=σ =0,=0.05) Gaussian noise (μ 0, =(μ 0.05) Gaussian noise ==0, 0, Gaussian noise (μ =σ==2=220, σ0.05) Gaussian noise (μ ===noise 0, σσ Gaussian (μ = 0, σ2 = 00.05) 22==00.05) 2 noise Gaussian noise (μ 0, Gaussian (μ3.=in 0,the σdifferent 0.05) = 0.05) Gaussian noise (μ 0, =(μ00.05) Gaussian σ222watermark 0.05) Gaussian noise (μ 0, =00.05) 0.05) Gaussian noise (μ =2after 0, σσσ =0, Table 3. Extracted watermark Table 3. Extracted Table attacks after Extracted case of watermark different attacks in spread after thedifferent case code of length. different attacks in spread the case code of length. different spread co Gaussian noise (μ =noise 0, σ =(μ 00.05) Gaussian ===22different 0, σ ===noise 00.05) 2σ Gaussian noise (μ = 0, = 00.05) Gaussian noise (μ 0, σ 00.05) Gaussian noise (µ = 0, σ = 0.05) 2 2 2 Gaussian noise (μ 0, σ = 00.05) GaussianGaussian noise (μ (6) =noise 0, Gaussian σ (μ (μ (6) = 0, σ = 0.05)Scaling (6) = 0.05) = 0,noise σ = 00.05) Scaling Scaling 2 2 2 Gaussian σ2 .005) 00.05) Gaussian noise (μ(μ 0, σ=2.005) =0,2σ00.05) Gaussian noise (μ==noise 0, σ=20.01, =(μ 00.05) Gaussian noise =σ ==.005) 20.01, Gaussian noise (μ σ0, =00.05) Gaussian noise (μ 0.01, Gaussian noise (μ(Level) =Length—512 0.01, σ2 = .005) Attacks (Level) Length—512 Length—256 Length—128 Length—256 Length—512 Length—128 Length—256 Len 2Attacks Gaussian noise (μ 0, Gaussian σ (μ = 0, σ2Attacks = 00.05) 2= 200.05) Gaussian noise =.005) =noise Gaussian noise (μ(Level) 0, =(μ Gaussian noise (μ=σ 0, Gaussian noise (μ ===20.01, 0.01, σ==2200.05) =σ2.005) .005) Gaussian noise (μ ==noise 0.01, σ =σ .005) Gaussian (μ 0.01, σ == .005) 2 00.05) 2 Gaussian noise (μ = 0.01, σ = Gaussian noise (μ = 0.01, σ .005) 2 2 2 Gaussian noise (μ = σ = Gaussian noise (μ = 0, Gaussian σ noise (μ = 0, σ = 00.05) = 00.05) Gaussian noise (μ = 0.01, σ = .005) Gaussian noise (µ = 0.01, σ = 0.005) 2 2 2 Gaussian noise (μ = 0, Gaussian σ =2 0.05) noise2 (μ = 0, Gaussian σ = 0.05) noise (μ = 0, σ = 0.05) 2 = .005) Gaussian noise (μ 0.01, σ0.001) =Salt .005) Gaussian noise (μ = (Density 0.01, σ0.001) Salt and Pepper noise =2 0.001) Gaussian noise (μ =noise 0.01, σ(Density == .005) Original Original Original Salt and Pepper Salt and Pepper noise (Density Pepper noise (Density = 0.001) 22 = = 2 == Gaussian noise (μ ==noise 0.01, Gaussian σσ0.001) noise (μand = 0.01, σ2 = .005) .005) Gaussian noise (μ 0.01, σ = .005) Salt and Pepper noise (Density = 0.001) Gaussian noise (μ = 0.01, σ .005) Gaussian noise (μ 0.01, = .005) Salt and Pepper noise (Density = Salt and Pepper (Density = 0.001) Salt and Pepper noise (Density = 0.001) Salt and Pepper noise (Density = 0.001) Salt and Pepper noise (Density = 0.001) 2 2 Salt and Pepper noise (Density = 0.001) Gaussian noise (μ = 0.01, Gaussian σ noise (μ = 0.01, σ = .005) = .005) 2 2 noise (μ(Density = 0, Gaussian σnoise noise (μ==0.001) 0, Gaussian σ = 00.05) noise (μ = 0, σ2 = 00.05) == 00.05) SaltGaussian and Pepper noise 0.001) Salt and Pepper Salt and Pepper noise (Density ==0.001) Salt and Pepper noise (Density =(Density 0.001) Salt and Pepper noise (Density = 0.1) Scaling (6) Scaling (6) Scaling Salt and Pepper noise (Density 0.1) Salt and Pepper noise (Density = 0.1) Salt and Pepper noise(6) (Density = 0.1) Salt and Pepper noise Salt (Density and Pepper = 0.001) noise (Density = 0.001) Salt and Pepper noise = ==0.001) Salt noise (Density =(Density 0.001) SaltPepper and Pepper noise (Density 0.1) Salt and noise (Density ===0.001) Saltand and Pepper noise (Density 0.1) Salt and Pepper noise (Density 0.1) Salt and Pepper noise (Density ==(Density 0.1) SaltPepper and Pepper noise (Density 0.1) Salt and Pepper noise (Density 0.1) Salt and Pepper noise = 0.1) Salt and Pepper noise Salt (Density and Pepper = 0.001) noise (Density = 0.001) 2 2 Gaussian noise (μ =(Density 0.01, Gaussian σ(Density noise ==0.01, Gaussian σ = .005) noise (μ = 0.01, 2σ2 = .005) .005) Salt and Pepper noise 0.1) SaltPepper and Pepper noise 0.1) Salt and noise = 0.1) Salt and Pepper noise (Density =(Density 0.1)= (μ 2× 2 = 0.05) Sharpen (5 ×=0.05) 5) noise (μ = 0, Gaussian σ noise (μ = 0, Gaussian σ noise (μ(5=×0,5)σ = 0.05) = Sharpen (5 5) Sharpen (5 × 5) Sharpen Salt and Pepper noise Salt (Density and Pepper = 0.1) noise (Density = 0.1) SaltGaussian and Pepper noise (Density = 0.1) Salt and Pepper Salt and Pepper noise (Density 0.1)= 0.1) Sharpen (5noise 5) Sharpen (5 × 5) 5) Sharpen (5 ×× 5) Sharpen (5 ×(Density Sharpen (5 =×(Density 5) =noise Sharpen (5 ×× 5) Sharpen (5 5) Salt and Pepper noise Salt and Pepper 0.1) (Density = 0.1) Salt and Pepper noise Salt (Density and Pepper Salt (Density and Pepper = 0.001) noise (Density = 0.001) Sharpen (5= ×0.001) 5) noise Sharpen (5 5) Sharpen (5 × 5) (3 ×××= 3) Gaussian noise (μ = 3) 0, Gaussian σ noise (μ = 0, Gaussian σ2 ×= 5) noise 0, σ2 = 00.05) 00.05) 00.05) Sharpen (5 × 5) Blur ×23) Blur (3Blur ×Sharpen Blur (μ (3 ×= 3) Sharpen (5 5) Sharpen (5 (5 × 5) Sharpen (5 ×(3 5) Sharpen (5 × 5) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Sharpen (5 ×Blur 5) (5 ×(Density 5) (3Pepper × Sharpen 3) Salt and Pepper noise Salt (Density and = 0.1)noise Salt and Pepper = 0.1)noise (Density = 0.1) Blur (3 3) Blur (3 × 3) 2 =×.005) Blur(μ (3=×0.01, 3) Gaussian noise Gaussian σ noise (μ = 0.01, Gaussian σ2 = .005) noise (μ = 0.01, σ2 = .005) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) 4.6. Number of Watermark Bits Robustness 4.6. Effect of of Number of (3 Watermark Bits onon Robustness 4.6. Effect ofEffect Number ofBlur Watermark Bits on Robustness 4.6. Effect of3)Number of Watermark Bits on Robustness ×of3) Blur (3 ×Robustness 4.6. Effect Effect of Number Number of Watermark Watermark Bits on Robustness Sharpen (5 ×Watermark 5) Sharpen (5 × 5) Sharpen (5 × 5) 4.6. Effect Effect of Number Number ofNumber Watermark Bits on on Robustness 4.6. Effect of Bits on 4.6. of of Watermark Bits Robustness 4.6. Effect of Number of Watermark Bits on Robustness 4.6. of of Bits on Robustness 4.6. Effect of Number of Watermark Bits on Robustness SaltAs andthe Pepper noise Salt (Density and Pepper = 0.001) noise Salt (Density and Pepper = 0.001) noise (Density = 0.001) number of embedded watermark bits increases the robustness the algorithm against As the number of embedded watermark bits increases the robustness ofof the algorithm against As the number of embedded watermark As bits the increases number the of embedded robustness watermark of the bits increases against the robustness of the a 4.6. Effect ofofof Number ofofWatermark Bits on Robustness 4.6. Effect of Number ofofWatermark Bits on Robustness 4.6. Effect of Number Watermark Bits on Robustness As the number embedded watermark bits increases the robustness ofalgorithm theof algorithm against As the number embedded watermark bits increases the robustness of the algorithm against As the number embedded watermark bits increases the robustness ofdegradations the algorithm algorithm against 4.6. Effect of Number of 4.6. Watermark Effect of Number Bits on Robustness of Watermark Bits on Robustness 4.6. Effect of Number of Watermark Bits on Robustness As the number of embedded watermark bits increases the robustness of the algorithm against 4.6. Effect of Number of Watermark Bits on Robustness As the number of embedded watermark bits increases the robustness the against 4.6. Effect of Number of Watermark Bits on Robustness As the number of embedded watermark bits increases the robustness of the algorithm against attacks also increases as shown in Figure 7. On the other hand, there occurs in the there occurs deg Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) As the number of embedded watermark bits increases the robustness of the algorithm against attacks also as shown inBits Figure On theon other there occurs inhand, the attacks increases asEffect shown in Figure attacks 7.noise On7. also the increases other hand, ashand, shown there(Density occurs in Figure degradations 7.degradations On the in other the 4.6.also Effect of increases Number of Watermark on Robustness 4.6. Effect of Number of 4.6. Watermark of Number Bits on Robustness of Watermark Bits Robustness Salt and Pepper noise Salt (Density and Pepper = 0.1) Salt (Density and Pepper = 0.1) noise = 0.1) attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the As the number of embedded watermark bits increases the robustness of the algorithm against As the number of embedded watermark bits increases the robustness of the algorithm against Asperceptual the number of embedded watermark bits increases the robustness of the algorithm against attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the the attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the quality of the watermarked image. Different watermarks used for simulations are shown attacks also increases as shown inimage. Figure 7. increases On the other hand, there occurs degradations in perceptual ofwatermarked the watermarked Different watermarks used for simulations arewatermarks shown perceptual quality ofof the image. perceptual Different quality watermarks of the watermarked used for simulations image. Different arealgorithm shown used for again simul As the number of embedded As the number watermark of embedded bits watermark the robustness bits increases of the the algorithm robustness against ofagainst the algorithm Asquality the number ofthe embedded watermark bits increases the robustness of the As the number embedded watermark bits increases the robustness of the algorithm against As the number of embedded watermark bits increases the robustness of the algorithm against perceptual quality of the watermarked image. Different watermarks used for simulations are shown perceptual quality of the watermarked image. Different watermarks used for simulations are shown perceptual quality of watermarked image. Different watermarks used for simulations are shown As the number of embedded watermark bits increases the robustness of the algorithm against attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the perceptual quality of the watermarked image. Different watermarks used for simulations are shown perceptual quality of the watermarked image. Different watermarks used for simulations are shown As the number of embedded As the number watermark of embedded bits increases watermark the robustness bits increases of the the algorithm robustness against of the algorithm against perceptual quality of the watermarked image. Different watermarks used for simulations are shown in Figure 8. perceptual quality of the watermarked image. Different watermarks used for simulations are shown in Figure 8. in Figure 8. in Figure 8. attacks also increases attacks as shown also increases in Figure as 7. shown On the in other Figure hand, 7. On there the occurs other degradations hand, there occurs in the degradations in th attacks also increases as shown in Figure 7. On the other hand, there occurs degradations in the attacks also increases as shown inEffect Figure 7. Ononother the other there occurs degradations in the Sharpen (5ofshown × Number 5) Sharpen (5ofOn × Number 5) Sharpen (5hand, ×there 5) attacks also increases as in Figure 7. the hand, occurs degradations in the 4.6. Effect of Number of 4.6. Watermark Effect Bits on Robustness of 4.6. Watermark Bits Robustness of Watermark Bits on Robustness in Figure 8.quality in Figure Figure 8.quality in Figure 8. attacks also increases as shown in Figure 7. On the other hand, there occurs in the perceptual quality ofwatermarked the watermarked image. Different watermarks used fordegradations simulations are shown perceptual of the image. Different watermarks used for simulations arethe shown perceptual of the watermarked image. Different watermarks used for simulations areoccurs shown in 8. in Figure 8. attacks also increases attacks as shown also increases in Figure as 7. shown On the in other Figure hand, 7. On there the occurs other degradations hand, there in degradations in the in Figure 8. in Figure 8. perceptual quality perceptual the quality of image. the watermarked Different watermarks image. Different used for simulations are shown for are show perceptual ofwatermarked the watermarked image. Different watermarks used for simulations aresimulations shown perceptual quality ofquality theof watermarked image. Different watermarks used for simulations are used shown perceptual quality of the watermarked image. Different watermarks used forwatermarks simulations areshown shown perceptual ofAs thethe watermarked image. Different watermarks used for simulations are inrobustness in Figure 8.perceptual in Figure 8. quality in Figure 8.quality perceptual of the watermarked quality of image. the watermarked Different watermarks image. Different used for watermarks simulations used are shown forof simulations are against shown As the number of embedded number watermark of embedded As bits the increases number watermark the of embedded robustness bits increases watermark of thethe algorithm robustness bits increases against thethe algorithm of the a in Figure 8. in Figure 8. Blur (3 × 3) Blur (3 × 3) Blur (3 × 3) in Figure 8. in Figure 8. in Figure 8. Figure 8. in Figure 8. in Figure 8. attacks also increases attacks as shown also increases in Figure attacks as 7. shown Onalso theincreases in other Figure hand, as 7. shown On there theoccurs in other Figure degradations hand, 7. On there theoccurs in other the degradations hand, there occurs in the deg perceptual quality ofperceptual the watermarked quality of image. perceptual the watermarked Different quality watermarks of image. the watermarked Different used forwatermarks simulations image. Different used are shown forwatermarks simulationsused are shown for simul 4.6. Effect 8. of Number of 4.6. Effect 8. of Number Bits on Robustness of 4.6. Effect 8. of Number Bits on Robustness of Watermark Bits on Robustness in Figure in Watermark Figure in Watermark Figure
As the number of embedded As the number watermark of embedded As bits theincreases number watermark the of embedded robustness bits increases watermark of thethe algorithm robustness bits increases against of thethe algorithm robustness against of the a attacks also increases attacks as shown also increases in Figure attacks as 7. shown Onalso theincreases in other Figure hand, as 7. shown On there theoccurs in other Figure degradations hand, 7. On there theoccurs in other the degradations hand, there occurs in the deg perceptual quality ofperceptual the watermarked quality of image. perceptual the watermarked Different quality watermarks of image. the watermarked Different used forwatermarks simulations image. Different used are shown forwatermarks simulationsused are shown for simul in Figure 8. in Figure 8. in Figure 8.
J. Imaging 2017, 3, 46 Imaging 2017, 2017, 3, 3, 46 46 J.J.Imaging Imaging J.J. 2017, Imaging 2017, 3, 3, 46 46
99 of 12 of 13 12 10 of 99 of 12 of 12 Compression Compression quality quality Vs Vs BER BER for for different different watermark watermark sizes sizes Compression Compression quality quality Vs Vs BER BER for for different different watermark watermark sizes sizes
BER (%) BER (%) BER (%)
40 40 40 35 35 35 30 30 30 25 25 25 20 20 20 15 15 15 10 10 10 5 55 0 00
16x16 16x16 16x16 16x16 32x32 32x32 32x32 32x32 41x41 41x41 41x41 41x41 45x45 45x45 45x45 45x45
0 00
50 50 50
100 100 100
Compression Compression Quality Quality Compression Compression Quality Quality
Figure 7. Compression Quality Vs BER in Figure the case of different watermark sizes. Figure 7. 7. Compression Compression Quality Quality Vs Vs BER BER in in the the case case of of different different watermark watermark sizes. sizes. Figure Figure 7. 7. Compression Compression Quality Quality Vs Vs BER BER in in the the case case of of different different watermark watermark sizes. sizes.
(a) (a) (a)
(b) (b) (b)
(c) (c) (c)
(d) (d) (d)
Figure Figure 8. 8. Watermarks Watermarks used used (a) (a) 16 16 ×× 16 16 bits bits (b) (b) 32 32 ×× 32 32 bits bits (c) (c) 41 41 ×× 41bits 41bits (d) (d) 45 45 ×× 45 45 bits. bits. Figure 41bits(d) (d)45 45××× bits. Figure Watermarks used (a) 16 16 bits (b) 32 ×× 32 45 bits. Figure8.8. 8.Watermarks Watermarksused used(a) (a)16 16×××16 16bits bits(b) (b)32 32× 32 bits bits (c) (c) 41 41 ×××41bits 41bits (d) 45 4545 bits.
4.7. Effect of Number of Spreading Codes on Robustness 4.7. Effect EffectofofNumber Numberof ofSpreading Spreading Codes Codes on on Robustness Robustness 4.7. Effect As the number of spreading codes increases, the number of bits of watermark that can be As the the number number of of spreading spreading codes codes increases, increases, the the number number of of bits bits of of watermark watermark that that can can be be As the number of watermark embedded also increases. When we use 4 spreading codes, in each of the selected mid band row 4 codes, inin each of of the selected midmid band rowrow 4 bits embedded increases. When we use spreading codes, each selected band 44 embedded also also increases. increases.When Whenwe weuse use444spreading spreading codes, in each of the the selected mid band row bits can be embedded and as per the proposed algorithm total of 1024 bits watermark can be can be embedded and asand per the proposed algorithmalgorithm total of 1024 bitsof be embedded at bits can be as the total 1024 watermark can bits can be embedded embedded and as per per the proposed proposed algorithm total ofwatermark 1024 bits bits can watermark can be be embedded at the maximum in a 512 × 512 image. Same way with one spreading code 256 bits and the maximum in amaximum 512 × 512 image. Same way withSame one spreading bits andcode with 256 16 embedded at in ×× 512 image. way one spreading bits embedded at the the maximum in aa 512 512 512 image. Same way with withcode one256 spreading code 256spreading bits and and with 16 spreading codes 4096 bits watermark can be embedded. Different watermarks used in this codes16 4096 bits watermark canbits be embedded. watermarks used in watermarks this section are shown in with spreading codes watermark can Different used in with 16 spreading codes 4096 4096 bits watermarkDifferent can be be embedded. embedded. Different watermarks used in this this section are shown in Figure 9. So, in effect as the number of spreading codes increases, the number Figure 9. So, in effect as the number spreading increases, the number of watermark bits that section are shown in 9. effect as number of codes the section are shown in Figure Figure 9. So, So, in inof effect as the the codes number of spreading spreading codes increases, increases, the number number of watermark bits that can be embedded and thus the robustness increases as shown in Figure 10. can be embedded and thus the robustness increases asrobustness shown in Figure 10. as of watermark bits embedded and thus increases of watermark bits that that can can be be embedded and thus the the robustness increases as shown shown in in Figure Figure 10. 10.
(a) (a) (a)
(b) (b) (b)
(c) (c) (c)
Figure Figure 9. 9. Watermarks Watermarks used used (a) (a) 16 16 ×× 16 16 bits bits (b) (b) 32 32 ×× 32 32 bits bits (c) (c) 64 64 ×× × 64bits. 64bits. Figure 32 bits (c) 64 64bits. Figure 9. Watermarks used (a) 16 ×× 16 bits (c) 64 ×× 64bits. Figure9. 9.Watermarks Watermarksused used(a) (a)16 16× 16 bits bits (b) (b) 32 32 ×× × 32 32 bits (c) 64 64bits. Compression Compression Quality Quality Vs Vs BER BER for for different different number number of of Compression Vs for Compression Quality Quality Vs BER BERcodes for different different number number of of spreading spreading codes spreading codes spreading codes
40 40 40 30 30 30
BER (%) BER (%) BER (%)
11 code code 11code code 44 codes codes 44codes codes 16 16 codes 16 codes 16 codes codes
20 20 20 10 10 10 0 00
0 00
50
50 50 Quality Compresion Compresion Quality Compresion Compresion Quality Quality
100 100 100
Figure 10. Compression Quality Vs BER in the case of different number of spreading codes. Figure Figure 10. 10. Compression Compression Quality Quality Vs Vs BER BER in in the the case case of of different different number number of of spreading spreading codes. codes. Figure Figure 10. 10. Compression Compression Quality Quality Vs Vs BER BER in in the the case case of of different different number number of of spreading spreading codes. codes.
J. Imaging 2017, 3, 46 J. Imaging 2017, 3, 46 J. Imaging 2017, 3, 46
11 10 of of 13 12 10 of 12
4.8. Effect of Type of Spreading Codes on Robustness 4.8. 4.8. Effect Effect of of Type Type of of Spreading Spreading Codes on Robustness Researchers used many types of codes for spreading the watermark bits. Gold code, PN code, Researchers used many for spreading the bits. code, code, Researchers used many types of of codes codes forspreading spreadingcodes. the watermark watermark bits. Gold code,PN PNtypes code,of Hadamard codes are the types commonly used The effect of Gold the different Hadamard codes are the commonly used spreading codes. The effect of the different types Hadamard codes are the commonly used spreading codes. The effect of the different types of spreading spreading codes on the performance of the proposed FWHT based algorithm is studied in detail.ofWe spreading codes oncode, the performance of theand proposed FWHT based algorithm studied in detail. We11 codes the performance of the proposed FWHT based algorithm isPN studied in detail. We have used haveon used Gold Hadamard code MATLAB generated codesis for spreading. Figure have used Gold code, Hadamard code andperformance MATLAB PN generated PNspreading. codes for(JPEG) spreading. Figure 11are Gold code, Hadamard code and MATLAB generated codes compression for Figure 11 PN shows that shows that Hadamard codes have better against and codes shows that Hadamard codes have better performance against compression (JPEG) and PN codes are Hadamard codes have better performance against compression (JPEG) and PN codes are the worst. the worst. Detailed analysis of the robustness of the algorithm against different attacks reveals that the worst.analysis Detailed of the robustness of the similar algorithm againstattacks different attacks reveals that in Detailed ofanalysis thecode robustness of the algorithm against different reveals Hadamard Hadamard and Gold sequences show almost robustness. As shown in that Figure 12, only Hadamard and Gold code sequences show almost similar robustness. As shown in Figure 12, only in and Gold code sequences show almost similar robustness. As shown in Figure 12, only in the case of the case of blurring attack PN codes outperforms Hadamard sequences. the case of blurring attackoutperforms PN codes outperforms blurring attack PN codes HadamardHadamard sequences. sequences. Compression Quality Vs BER
BER (%) BER (%)
45 Compression Quality Vs BER 4540 4035 3530 3025 2520 2015 1510 10 5 5 0 0 0 50 Compression 0 50 Quality Compression Quality
hadamard hadamard pn sequence pn sequence Gold code Gold code
100 100
Figure Compression Quality BER case different types spreading codes. Figure 11.11. Compression Quality VsVs BER inin thethe case ofof different types of of spreading codes. Figure 11. Compression Quality Vs BER in the case of different types of spreading codes. Effect of type of spreading codes in the case of different attacks Effect of type of spreading codes in the case of different attacks 30 30 25 25 20 20 15 15 10 10 5 5 0 0
Hadamard Gold PN Sequence Hadamard Gold PN Sequence
Figure 12. BER after different attacks in the case of different types of spreading codes. Figure 12. BER after different attacks in the case of different types of spreading codes. Figure 12. BER after different attacks in the case of different types of spreading codes.
5. 5. Conclusions Conclusions
5. Conclusions This a detailed analysis ofof digital image watermarking, various Thispaper paperdid did a detailed analysis digital image watermarking,itsits variousproperties propertiesand and applications. Proposed a CDMA based spread spectrum technique of watermarking using customized This paper did a detailed analysis of digital image watermarking, its various properties and applications. Proposed a CDMA based spread spectrum technique of watermarking using 8× 8 Fast Walsh Transform forTransform color images compared the performance similar applications. Proposed a CDMA based spread spectrum technique of watermarking using customized 8 × Hadamard 8 Fast Walsh Hadamard forand color images and compared thewith performance existing DCT based algorithm. Simulation results showed that the proposed algorithm exhibits customized 8 × 8 Fast Walsh Hadamard Transform for color images and compared the performance with similar existing DCT based algorithm. Simulation results showed that the proposed algorithm better robustness against at a specified and is highly resilient to low quality to compression with similar existing DCTattacks based algorithm. results theresilient proposed algorithm exhibits better robustness against attacks Simulation atgain a specified gainshowed and is that highly low quality compared to DCT algorithm. Analyzed in detail the effect of number, length and type of spreading exhibits better robustness against attacks at a specified gain and is highly resilient to low compression compared to DCT algorithm. Analyzed in detail the effect of number, length quality and type codes on the compared robustness ofthe therobustness algorithm. Another major advantage of the technique istype low compression DCT algorithm. Analyzed in detail the effect ofproposed number, length of spreading codes onto of the algorithm. Another major advantage of theand proposed computational complexity and thus the low cost. The only drawback of FWHT is its low information of technique spreadingiscodes on the robustness of the algorithm. Another major advantage of the proposed low computational complexity and thus the low cost. The only drawback of FWHT is its technique is low computational complexity low cost.quality. The only of FWHT is its is low information hiding capacity and thusand thethus low the perceptual So,drawback the proposed algorithm low information hiding capacity and thus the low perceptual quality. So, the proposed algorithm is
J. Imaging 2017, 3, 46
12 of 13
hiding capacity and thus the low perceptual quality. So, the proposed algorithm is a good candidate for applications that require computationally efficient, low cost and robust watermarking. Author Contributions: Suja conceived, designed and performed the experiments; Suja and T.Naser analyzed the data. A.Afaq and Fahad involved in the discussions and analysis of the results. Suja wrote the paper and T.Naser reviewed the paper. Conflicts of Interest: The authors declare no conflict of interest.
References 1.
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
14. 15. 16.
17.
18. 19.
Vassaux, B.; Bas, P.; Chassery, J.M. A New CDMA Technique for Digital Image Watermarking. In Proceedings of the 2001 International Conference on Image Processing, Thessaloniki, Greece, 7–10 October 2001; Volume 3, pp. 983–986. Al-Haj, A.M. Advanced Techniques in Multimedia Watermarking: Image, Video and Audio Applications; Information Science Reference: Hershey, PA, USA, 2010. Cox, J.; Kilian, J.; Leighton, F.T.; Shamoon, T. Secure Spread Spectrum Watermarking for Multimedia. IEEE Trans. Image Process. 1997, 6, 1673–1687. [CrossRef] [PubMed] Samee, M.K.; Gotze, J. CDMA Based Reversible and Blind Watermarking Scheme for Images. Int. J. Future Comput. Commun. 2013, 2, 138. [CrossRef] Shrivastava, V.; Saini, L.K. A Survey of Digital Watermarking Techniques and its Applications. Int. J. Comput. Sci. Trends Technol. IJCST 2014, 2, 303–313. Bamatraf, A.; Ibrahim, R.; Najib, M. A New Digital Watermarking Algorithm Using Combination of Least Significant Bit (LSB) and Inverse Bit. J. Comput. 2011, 3, 1–8. Masoumi, M.; Rezaei, M.; Hamza, A.B. A blind spatio-temporal data hiding for video ownership verification in frequency domain. AEU-Int. J. Electron. Commun. 2015, 69, 1868–1879. [CrossRef] Patil, P.; Bormane, D.S. DWT Based Invisible Watermarking Technique for Digital Images. Int. J. Eng. Adv. Technol. IJEAT 2013, 2, 603–605. Vishwanatham, V.M.; Reddy, G.S.; Jagadeesh, P. A Hybrid Digital Watermarking Algorithm for Colour Images Based on DWT and DCT; Annals Computer Science Series: Mirton, Romania, 2012. Khan, M.I.; Sarkar, M.I.H. An Efficient Image Watermarking Scheme Using BFS Technique Based on Hadamard Transform. Smart Comput. Rev. 2013, 3, 298–308. Lee, L.T.; Pai, Y.T. DS-CDMA based Watermarking: A low power, high security algorithm for embedded systems. Int. J. Innov. Comput. Inf. Control 2010, 7, 5893–5908. Anju, R. Modified Algorithm for Digital Image Watermarking Using Combined DCT and DWT. Int. J. Inf. Comput. Technol. 2013, 3, 691–700. Garimella, A.; Satyanarayana, M.V.V.; Murugesh, P.S.; Niranjan, U.C. ASIC for Digital Color Image Watermarking. In Proceedings of the IEEE 11th Digital Signal Processing Workshop and IEEE Signal Processing Education Workshop, Taos Ski Valley, NM, USA, 1–4 August 2004; pp. 292–296. Chaturvedi, N. Comparison of Digital Image watermarking Methods DWT & DWT-DCT on the Basis of PSNR. Int. J. Innov. Res. Sci. Eng. Technol. 2012, 1, 147–153. Franklin, R.V.; GRS, M.; Santhi, V. Entropy based Robust Watermarking Scheme using Hadamard Transformation Technique. Int. J. Comput. Appl. 2011, 12, 14–21. Shen, J.; Tan, S.H.; Anthony, T.S.H. A Robust Digital Image-in-Image Watermarking Algorithm Using the Fast Hadamard Transform. In Proceedings of the 2003 International Symposium on Circuits and Systems, Bangkok, Thailand, 25–28 May 2003. Marjuni, A.; Logeswaran, R.; Ahmad Fauzi, M. F. An image watermarking scheme based on fast Walsh Hadamard transformation and discrete Cosine transformation. In Proceedings of the International Conference on Networking and Information Technology, Manila, Philippines, 11–12 June 2010; pp. 289–293. Shanthakumari, R.; Parvathavarthini, S. An Adapative Watermarking Process in Hadamard Transform. Int. J. Adv. Inf. Technol. 2014, 4, 1. Ishikawa, Y.; Uehira, K.; Yanaka, K. Practical evaluation of illumination watermarking technique using orthogonal transforms. J. Display Technol. 2010, 6, 351–358. [CrossRef]
J. Imaging 2017, 3, 46
20.
21. 22.
23.
13 of 13
Maity, S.P.; Kundu, M.K. A Blind CDMA Image Watermarking Scheme in Wavelet Domain. In Proceedings of the IEEE International Conference on Image Processing-ICIP 2004, Singapore, 24–27 October 2004; Volume 4, pp. 2633–2636. Wah, W.H. Image Watermarking and Data Hiding Techniques. Ph.D. Thesis, Hong Kong University of Science and Technology, Hong Kong, China, August 2003. Abdallah, E.E.; Hamza, A.B.; Bhattacharya, P. Robust block based image watermarking scheme using fast Hadamard transformation and single value decomposition. In Proceedings of the 18th International Conference on Pattern Recognition, Hong Kong, China, 20–24 August 2006; Volume 3, pp. 673–676. Mwangi, E. A wavelet based image watermarking scheme with a CDMA/Hadamard embedding technique. In Proceedings of the 9th International Symposium on Signal Processing and Its Applications, Sharjah, UAE, 12–15 February 2007; pp. 1–4. © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
