We use cookies to give you the best experience possible. By continuing we’ll assume you’re on board with our cookie policy

The undertaking was started with the acquisition procedure and a survey conducted by old research workers. Literature reviews done at the earlier phase to explicate overview of theoretical background of undertaking and reexamine old undertaking of the research country. This chapter gives an overview sing the theory that is applied into this undertaking.

2.2 Digital Watermarking Basic Principles

The Watermark Extraction Or Recovery System... TOPICS SPECIFICALLY FOR YOU

Fridrich. J ( 1998 ) , in his paper [ 1 ] provinces all watermarking methods portion the same edifice blocks: an implanting system and the water line extraction or recovery system. In general, any watermarking strategy consists of three parts:

The water line

The encoder ( Insertion or implanting algorithm )

The decipherer ( Verification or Extraction algorithm )

Each proprietor has a alone water line to integrate the water line into the object. The object besides known as original image, cover object or host image. Merely the authorised users should derive entree to the watermarked informations.

The phases of watermarking procedure which comprises of the embedding, the distribution, the extraction and the determination, are described in the undermentioned subdivisions.

2.2.1 Embedding Phase

In this embedding phase, the original image ( H ) to be watermarked is preprocessed before implanting a water line ( W ) . Early watermarking strategies worked in the spacial sphere, where the water line is added by modifying pixel values of the host image. Examples of such techniques are Substitution Watermarking and Additive Watermarking [ 2 ] . For the instance of implanting in the transform sphere, this may affect change overing the image to the desired sphere such as the Discrete Cosine Transform ( DCT ) , the Discrete Fourier transform ( DFT ) and the Wavelet Transform ( WT ) domains. Then, to acquire the watermarked image, reverse transform is performed.

The water line to be embedded can be a binary image, a spot watercourse or a pseudo-random figure. The key is used to bring forth a more unafraid water line. The water line key is private and merely the authorised individual is known. It ensures that merely authorized individual are able to observe the water line informations. Figure 2.1 illustrates the encoding procedure.

Figure 2.1 Watermark Embedding

Mathematically, this can be written as

E ( H, W, K ) = H* ( 2.1 )

Where E is an encoder map and K is the secret key.

The end product is the watermarked information. It is perceptually indistinguishable to H and is obtained by executing an reverse transform on the altered transform coefficients.

It is base on ballss through the transmittal channel. The digital watermarked merchandise will be transmitted through some ways such as cyberspace, or transmittal within pen-drive. The channel for the watermarked informations could be lossy, noisy and dependable channel. In the procedure of transmittal and distribution of the watermarked image, this will lend mistakes to the watermarked image. All these uses on the watermarked image have to be seen as an onslaught on the embedded information. Thus the standard watermarked informations may be different from the original watermarked informations. Detailss of the onslaughts will be described in subdivision 2.6.

2.2.2 Extracting Phase

Figure 2.2 illustrates the decryption procedure.

Figure 2.2 Watermark Extracting

Let ‘s denote a decipherer map as D. D takes a watermarked image H* whose ownership is to be determined and recovers a water line W* from the image utilizing the secret key ( K ) . This watermarking technique is said to be unafraid since the key used at implanting, is needed for extraction. It is difficult to take or change the message from the informations without cognizing the key. Mathematically, this is written as

D ( H* , H, K ) = W* ( 2.2 )

Where D is a decipherer map and K is the secret key.

The extracted water line ( W* ) will be used in the determination devising phase. The watermarking system examines the extracted informations by measuring the similarity between the original water line image ( W ) and the extracted water line image ( W* ) during this decoding phase.

The correlativity is calculated between the cured water line image ( W* ) and the original water line image ( W ) for each of the pels. Correlation is used as the similarity step of two images with an appropriate threshold.

Mabtoul. S, Ibn-Elhaj. E, and Aboutajdine, 2006 in their paper [ 3 ] defines correlativity as

( 2.3 )

Where W is original watermark image and W* is the cured water line image

Meanwhile, Krishnan Nallaperumal, R.K Selvakumar, S. Rajapandian, K.ArulMozhi and C.Nelson Kennedy Babu, 2006 said in their diary, a higher correlativity indicates the being of the water line in that set [ 9 ] . It is 1, if the original water line resembles the extracted water line and 0 otherwise [ 3 ] , which can be represented as

( 2.4 )

Where degree Celsius is the correlativity of two images, is certain threshold.

The threshold for the determination is set as the mean of the correlativity value for all the pels [ 4 ] . The water line is detected if it is larger than the threshold otherwise water line can non be found in the image.

2.3 Watermarking Design Issues

The demand of a watermarking system strongly relies on the peculiar applications in which it will be deployed. The chief demands which should be fulfilled of digital watermarking system are imperceptibility, security and hardiness [ 5 ] , [ 6 ] , [ 7 ] .

2.3.1 Imperceptibility

The watermarking design should be perceptually unseeable so data quality is non degraded and aggressors are prevented from happening and canceling it. It is called unperceivable if the watermarked image is perceptually tantamount to the original water line information [ 20 ] .

Measuring imperceptibility in watermarking is of import to mensurate the quality of the image the decipherer phase. Peak Signal-to-Noise Ratio ( PSNR ) is used to bespeak the quality of image with watermarking procedure. The expression is showed below

( 2.5 )

Where MA-N is image size, 255 is grey degree scope of image, M ( I, J ) and M ‘ ( I, J ) are grey degree values at pel ( one, J ) of original image and watermarked consequence image severally. Mean Square Error ( MSE ) value is the amount between original image and watermarked image in db unit. Na Li, Xiaoshi Zheng, Yanling Zhao, Huimin Wu and Shifeng Li, 2008 in their researches proves that the bigger the value of PSNR, the better the quality of image [ 8 ] .

2.3.2 Robustness

Watermark hardiness explains for the capableness of the concealed water line to last legal of day-to-day use or any image processing use from knowing and unwilled aggressors. Means that, robust water line has the handiness to defy assorted image onslaughts therefore supplying hallmark and ownership designation.

In order to measure the hardiness of watermarking algorithm, the original water line, W and watermarked image W* is calculated with the expression below [ 5 ] , [ 8 ] , [ 10 ] :

( 2.6 )

Where NC is Normalized Cross Correlation

Besides, the watermarking strategy is besides tested by assorted sort of onslaught to mensurate the hardiness of the algorithm. The watermarking is tested by geometrical onslaughts, contrast sweetening, adding noise, etc. A full list of onslaught with related parametric quantities is recorded in subdivision 2.6

2.3.3 Security

Ton Kalker defines watermarking security as “ the inability by unauthorised users to hold entree to the natural watermarking channel ” . In other words, watermark security refers to the failures of unauthorised users to change, to take, to read or to compose the water line content established by robust watermarking [ 9 ] . Harmonizing to Kerckhoffs, in Ke Luo and Xiaolin Tian diary, the security must lie in the pick of the cardinal [ 10 ] .

2.4 Existing Image Watermarking Techniques

The categorization of watermarking algorithm is done in several position points. One of the point of views is based on treating sphere spacial sphere or frequence sphere.

2.4.1 Spatial Domain Techniques

Most of the early researches in digital water line embedded the water line in the spatial sphere which is straightforward, simple and non dearly-won. Least Significant Bit ( LSB ) is the easiest technique in the spacial sphere. Techniques in spacial sphere normally portion the following features:

The water line information is applied in the pel sphere.

No transforms are applied to the screen object in water line embedding.

Combination with the screen object is in the pel sphere.

The correlativity is calculated between the expected forms with the standard signal.

By and large, spacial sphere watermarking techniques are non robust against image processing operations because the embedded water line is non distributed for the whole image and therefore contributes in easiness to destruct the water line [ 2 ] .

Least Significant Bit ( LSB ) Technique

The easiest method of water line embedding is to implant the water line into the least important spots of the screen object. In grayscale image, the most important spot ( MSB ) is in the left side and the least important spot ( LSB ) to the right of 8 spots of a pel.

Figure 2.3a shows a pel holding the grey value 130. The thought of LSB is to replace the LSB of a pel with the water line. As shown in Figure 2.3b, the value changes from 130 to 131 when the LSB is changed. It is undetectable from human eyes. If alter the spot place more closely to MSB, the image will be distorted more, as described in Figure 2.3c [ 2 ] .

Figure 2.3 ( a ) 8 spot pels with a value of 130. ( B ) The value changed to 131 after replacing the LSB. ( degree Celsius ) The value is changed to 2 after the MSB permutation

2.4.2 Frequency Domain Techniques

In order to hold a more powerful technique, frequence sphere techniques is introduced such as DCT sphere, DWT sphere, DFT sphere etc.

Discrete Cosine Transform ( DCT ) Watermarking Techniques

DCT plants by dividing images into parts of differing frequences. Merely the most of import frequences that remain are used to recover the image in the decompression procedure [ 13 ] . The DCT plants by dividing images into different frequence sets. Therefore, it is much easier to implant watermarking information into the in-between frequence sets of an image. The in-between sets are chosen to avoid the low frequence set that contained the most of import parts of image. Furthermore, the in-between set is selected without unmaskings themselves to removal through compaction and noise onslaughts in high frequence set [ 14 ] . Then, pseudo-random sequences, such as M-sequences, are added to the DCT coefficients at the in-between frequences as signatures.

Figure 2.4 DCT sphere watermarking

Discrete Fourier Transform ( DFT ) Watermarking Techniques

Fourier transform decomposes image map into a set of extraneous maps, and can transform the spacial strength image into its frequence sphere [ 2 ] .

The chief drawbacks of utilizing FFT are it has less ability to defy JPEG compaction and cropping onslaught [ 6 ] . Besides, the loss of clip information in a signal by Fourier Transform will take to the trouble in treating [ 5 ] .

Discrete Wavelet Transform ( DWT )

Recently, most of the research workers focus on implanting water line in ripple sphere because of the belongings of multi declaration analysis that it provides. The bing ripple based watermarking techniques are explained below:

P. Ramana Reddy, Dr. Munaga and Dr. D. Sreenivasa ( 2009 ) present a robust digital watermarking of images by modifying the frequence coefficients of the image based on the Human Visual System ( HVS ) of image content. The operation of implanting and extraction of the water line is done in the frequence sphere. Therefore, contributes hardiness against frequency- based onslaughts such as compaction and filtering. The water line is implanting into an image by modifying coefficients of mid frequence sets ( LH and HL subbands ) . The experimental consequence proves the watermarking strategy applied is extremely robust against assorted onslaughts such as filtering, compaction, Gaussian noise etc [ 5 ] .

Jianmin Xie and Qin Qin ( 2010 ) proposed a watermarking strategy based on the Discrete Wavelet Transform ( DWT ) . In order to better the water line is invisibleness ; algorithm choice coefficient is in high frequence subbands to add a water line. The experimental consequences show that the algorithm research paper is executable, simple and easy to implement. This technique proved to be more robust than the DCT method [ 10 ] .

Krishnan Nallaperumal, R.K Selvakumar, S. Rajapandian, K.ArulMozhi and C.Nelson Kennedy Babu ( 2006 ) proposed the image is decomposed into ripple coefficients and a ocular recognizable logo and content based watermark information is embedded in the ripple coefficients. Implanting the water line in such pels makes it possible to utilize maximal sum of water line due to human oculus insensitiveness to high information countries [ 9 ] . Munesh Chandra and Shikha Pandey ( 2010 ) give an overview of watermarking techniques and proposed a seeable watermarking algorithm for copyright protection of digital images based on DWT. The advantages are due to good clip frequence characteristics and good fiting with HVS directives [ 4 ] .

Na Li, Xiaoshi Zheng, Yanling Zhao, Huimin Wu and Shifeng Li ( 2008 ) in their paper proposed a robust algorithm of digital image watermarking based on DWT. This technique adds binary image water line into grey image. Furthermore, the host image is needed for observing at decipherer phase. Watermarking is embedded in the 3rd category HL subbands so that it has unobtrusiveness every bit good as the hardiness to maintain quality of the original image. Furthermore, Arnold Transform is applied to binary image to do watermarking more strongly robust against cropping operation [ 8 ] . Akhil Pratap Singh and Agya Mishra presents a robust watermarking technique in the grey sale image. In this paper, grey scale image is found to be simpler than other transform technique [ 15 ] .

Ke Luo and Xiaolin Tian ( 2008 ) proposed a new robust watermarking strategy based on DWT, where a water line is embedded into a host image twice in two different frequence ranges to defy different type of image processing onslaughts. First, the water line is embedded into lower frequence coefficients. Next, the same water line is embedded into the mid- frequence coefficients of the host image once more to heighten hardiness of the water line. From the consequences obtained, the water lines inserted into the in-between frequence and high frequence are normally less robust to low-pass filtering, JPEG compaction and grading, but are highly robust with regard to salt and pepper noise and rotary motion. The low frequence water lines are typically strong robust with nonlinear filtrating such as lossy compaction, average filter, etc [ 7 ] .

Mohamed A. Mohamed, Mohy El-Din A. Abou-Soud, and Mai S. Diab in their paper said, another technique for water line embedding is to utilize the correlativity belongingss of linear pseudo-random noise forms as applied to an image. A pseudo-random noise ( PN ) form of W ( x, Y ) is added to the host image of I ( x, y ) . Harmonizing to the equation:

( ten, Y ) = I ( x, y ) + k *W ( x, Y ) ( 2.7 )

Where K denotes a addition factor and the ensuing watermarked image

Increase the value of K, will increasing the hardiness of the water line at the disbursal of the quality of the watermarked image [ 21 ] . Furthermore, Jianmin Xie and Qin Qin said in their paper, if K has greater value, the stableness is better and the invisibleness is worse. On the other manus, if k has smaller value, the invisibleness is better but the stableness is worse [ 10 ] .

Mohamed A. Mohamed, Mohy El-Din A. Abou-Soud, and Mai S. Diab in their paper proposed watermarking strategies based on Haar Wavelet Transform. The Haar ripple transform are chosen because it is simple and fast. It is precisely reversible without any border effects. They proposed watermarking strategy embeds the water line in the LH and HL set. The LH and HL values are modified harmonizing to pseudo random Number ( PN ) sequence for the water line spot 0. Then reverse DWT is used to acquire the watermarked image. During watermark extraction procedure, if the correlativity of LH and HL values of watermarked image and the PN sequence is greater than the average correlativity so the water line spot is set to 0 [ 22 ] .

2.5 Discrete Wavelet Transform ( DWT ) Domain Watermarking

Discrete Wavelet Transformation ( DWT ) transform distinct signal from clip sphere into time-frequency sphere. The transmutation consequence is the coefficients that are enables spectrum analysis of the signal and besides spectral behaviour of the signal in clip [ 16 ] .

DWT has been used in digital watermarking more often than other frequence sphere techniques such as DCT and DFT because of it particular features. This is due to multi declaration features, first-class spacial localisation, more accurately to the theoretical theoretical accounts of the human ocular system ( HVS ) as compared to DCT and DFT [ 4 ] , [ 9 ] , [ 10 ] .

The human ocular system is related to the perceptual quality and measured harmonizing to the sensitiveness or acuteness of human oculus to see inside informations in an image. Based on research in human perceptual experience, it is found that the retina of the oculus splits an image into several frequence channels. Each signal in these channels is processed independently. This procedure is similar to DWT multi declaration decomposition. The multi declaration consecutive estimate enhances the declaration of an image and enhances the declaration of water line at the same time. This benefit allows utilizing higher energy water lines in parts where the HVS is less sensitive. As a consequence, implanting water lines in those peculiar parts provides to increase the hardiness of the watermarking techniques.

2.5.1 Wavelet Decomposition

For 2D images, the ripple transform is done in both horizontal and perpendicular waies. First, using 1D ripple horizontally to each row of the image giving Low Horizontal ( LH ) and High Horizontal ( HH ) and so using on all the column of the image giving ( LL, LH ) and ( HL, HH ) can be computed the 2D transform.

The LL bomber set represents the coarse-scale DWT coefficients, while LH, HL and HH sub sets represent the fine-scale of DWT of DWT coefficients. The following harsh graduated table of ripple coefficient can be determined by farther procedure the LL bomber set, until some concluding graduated table of N is approachable. When N is reached, 3N+1 bomber set is obtained, dwelling of the multi declaration subbands LL, LH, HL, and HH bomber sets. Example of an image being decomposed into 10 sub sets for three degrees is shown in Figure 2.5. [ 5 ] , [ 9 ] , [ 10 ] , [ 13 ] .

Figure 2.5 DWT Decomposition of Image

2.5.1 Wavelet Image Reconstruction

From DWT coefficients as mentioned before, the original image can be reconstructed. The ripple image Reconstruction is similar to the opposite of the ripple decomposition. The original image is obtained by concatenating all the coefficients, get downing from the last degree of decomposition. This procedure is continued through the same figure of degrees as in the decomposition procedure [ 10 ] .

2.6 Possible Attacks on Image Watermarking

The aggressors meaning at the watermarked images can be classified as unwilled or knowing. The aggressors have three schemes to get the better of watermark hardiness as:

To take plenty watermark signal

To throng the concealed communicating channel

To desynchronise the watermarked content

The purpose of the aggressors is to change, take or degrade the effectivity of the water line. An onslaught is said to be successful if the aggressors disturb any phase of the watermarking rhythm [ 2 ] .

In this chapter, different types of onslaughts will be described in order to prove the hardiness of watermarking strategies. The most popular categorization is summarized in Figure 2.5 [ 7 ] , [ 16 ] , [ 17 ] .

Figure 2.5 Attacks on Water lines

2.7 Applications of Watermarking

There is a broad assortment of applications in watermarking. Several applications are listed below.

Copyright Protection

Digital water lines are capable to be used in placing and protecting the copyright ownership of the content [ 6 ] . It besides can be used in following illicitly distribution transcripts [ 17 ] .

Identity Card / Passport Security

In the field of informations security, water lines may be used for hallmark, enfranchisement, and conditional entree such as in individuality card and passport security [ 19 ] . Information in a passport or ID card can besides be included in the individual ‘s exposure that appears on the ID card.

The interpolation of the water line provides an excess degree of security in this application. For illustration if ID card is stolen and he/she replaces the image, the failure in pull outing the water line will annul the ID card [ 2 ] , [ 18 ] .

Transcript Protection

Digital content can be watermarked to bespeak that the content can non be illicitly replicated and prevent people from doing illegal transcripts of right of first publication content [ 2 ] , [ 6 ] .


Digital water lines can be used to track the use of digital content. Each transcript of digital content can be unambiguously watermarked with metadata stipulating the authorised users of the content. Tracking application is used to observe illegal copying of content by placing the users who fake the content illicitly. The watermarking technique used for tracking is called as fingerprinting.

Fingerprint is a existent progress in placing existent manufactured objects from bogus 1s based on digital images of the original merchandise stored in a protected waiter [ 2 ] , [ 19 ] .

Share this Post!

Send a Comment

Your email address will not be published.

Kylie Garcia

Hi there, would you like to get such a paper? How about receiving a customized one?

Check it out