# INTRODUCTION igital Watermarking is the process of possibly irreversibly embedding information into a digital signal. The signal may be audio, video, or pictures for example. If the signal is copied then the information is also carried in the copy. In visible watermarking, the information is visible in the picture or video. Typically, the information is text or a logo which identifies the owner of the media. The image on the right has a visible watermark. When a television broadcaster adds its logo to the corner of transmitted video, this is also a visible watermark. In invisible watermarking, information is added as digital data to audio, picture or video, but it cannot be perceived as such (although it may be possible to detect that some amount of information is hidden). The watermark may be intended for widespread use and is thus made easy to retrieve or it may be a form of Steganography, where a party communicates a secret message embedded in the digital signal. In either case, as in visible watermarking, the objective is to attach ownership or other descriptive information to the signal in a way that is difficult to remove. It is also possible to use hidden embedded information as a means of covert communication between individuals. One application of watermarking is in copyright protection systems, which are intended to prevent or deter unauthorized copying of digital media. In this use a copy device retrieves the watermark from the signal before making a copy; the device makes a decision to About -Sri Rangapoopathi College of Engineering watermark is embedded into a digital signal at each point of distribution. If a copy of the work is found later, then the watermark can be retrieved from the copy and the source of the distribution is known. This technique has been reportedly used to detect the source of illegally copied movies. While some file formats for digital media can contain additional information called metadata, digital watermarking is distinct in that the data is carried in the signal itself. The use of the word of watermarking is derived from the much older notion of placing a visible watermark on paper. # b) Advantages Of Contourlet Domain # ? Contourlet possesses the important properties of directionality and anisotropy which wavelets do not possess and so it outperforms wavelets in many image processing applications. # ? Contourlet provides a much richer set of directions and shapes than wavelets and thus is more efficient in capturing smooth contours. # ? Contourlet is a multidirectional and multi-scale transform and is obtained by combining both directions can be specified by the user. # ? Contourlet gives more edges, it is more suitable for data hiding applications and more data can be hidden in the high frequency regions without perceptually distorting the original image. # c) Image Processing with MatLab Image processing is a very important and widely used process in which images are processed to retrieve information that is not visible to the naked eye, as well as it i s used in a large number of other applications where the standard picture is not enough. The first step is to download all the 'jpg' and 'tif' files that are to be used in this lab. Locate them as a link on the webpage. The second step is to set the directory of MatLab to the directory which contains your\ image files. Be sure that you create this directory on your U-drive. We can do this by first typing cd on the MatLab prompt. Usually the default setting is: c:\MatlabR12\work If this is so, change the directory by simply typing cd.. which changes the directory to c:\Matlab. We can repeat the step to go to C:\ drive and then type cd d:\. This will change the MatLab directory to our'd' drive. In order to read the image use the following command in MatLab imread('filename'); This command will save the image in the Image detail module. Since bitmap images are fairly large, they take a long time to convert into matrices. Hence we stick to jpeg and tiff image formats in this lab. # III. # Embedding phase # ©. Genetic Operator This function is used to produce offsprings from parent chromosomes by the set of selected chromosomes. Function f = genetic operator (parent chromosome, M, V, mu, mum, l_limit, u_limit) Where, M -Number of objective functions. V -Number of decision varaiables. Mu -distribution index for crossover. Mum -distribution index for mutation. l_limit -a vector of lower limit for the Corresponding decision variables. u_limit -a vector of upper limit for the corresponding decision variables. # iii. Selection The individuals are selected by using their rank that is obtained through their fitness value and it is continued till the cross-over area gets filled. Before that checking of every individuals contains the same number of elements is performed. Function tournament_selection (chromosome, pool_size, tour_size) # Tournament selection process In a tournament selection process n individuals are selected at random, where n is equal to tour_size. From these individuals only one is selected and is added to the mating pool, where size of the mating pool is pool_size. Selection is performed based on two criteria. First and foremost is the rank or the front in which the solutions reside. Individuals with lower rank are selected. Secondly if the rank of two individuals is the same then, the crowding distance is compared. Individuals with greater crowding distance are selcted.The size of chromosome is taken. The number of chromosome is not important while the number of elements in chromosome is important. iv. # Cross over and Mutation Cross over operation is performed by using the probability with 90%. Two individuals are selected and it is checked that the selected two individuals for cross over operation not the same. The information of the two selected parent chromosomes is collected and the cross over operation is performed for each decision variable in the parent chromosome. Function f= replace_chromosome (intermediate_chromosome,pro,pop) This function replaces the chromosomes based on rank and crowding distance. Initially until the population size is reached each front is added one by one until addition of a complete front which results in exceeding the population size. At this point the chromosomes in that front are added subsequently to the population based on crowding distance. ? C i,k = Child. ? p i,k = Selected parent. ? = Random # S. No. G4 G3 G2 G1 Fitness Rank C1 0 1 0 0 4 1 C2 1 0 0 0 8 2 C3 1 1 0 0 12 4 C4 1 0 1 0 10 3 Table 1: Rank Calculation Mutation is performed with a probability of 10% and it is based on polynomial mutation. Mutation is performed on the selected element of the chromosome and the selected element is inverted or reversed to obtain a new child as a resulting chromosome. C k =p ike +(p k u -p k t) ? k. ? ? c 1, k= 1 2 [(1-)p1,k+(1+ )p2,k] ? ? k k p( ) = 1 2 (n +1) ,if 0 1 ? ? c c ? n ©2011 Global Journals Inc. (US) Where, Ck is the child. Pk is the parent. P k u is the upper bound on the parent component. P k t is the lower bound on the parent component. ? k is small variation which is calculated from a polynomial distribution. if rk < 0.5 if rk >0.5 Where, rk is an uniformly sampled random number between (0,1). is mutation distribution index. # After Mutation Table 4: After Mutation v. Embedding Algorithm In Embedding phase, an image that is transformed into binary format is inserted in the original image as the watermark. The Embedding Algorithm is as following: 1. The original image is decomposed into its sub bands using contourlet transform. In Contourlet, the number of directional subbands at each level is set to 2n where n is a positive integer number. Forexample, if we choose to decompose an image into four levels using n= (1, 2, 3, 4) then we get 2, 4, 8, and 16 subbands Where, X: input image. Fname: filter name. n: number of decomposition tree levels. # S. No 1 n +1 k = 1-[2(1-r )] k m ? n +1 k = (2r ) k m 1 -1 ©2011 Global Journals Inc. (US) 3. The Sub band Co-efficients is selected by using a threshold value. 4. Genetic algorithm is applied to the Co efficient to increase the robustness and quality simultaneously. 5. Then the watermarks are embedded in the selected Co-efficients. In order to embed a zero bit in a Contourlet sub band, the standard deviation of its coefficients magnitude is set to zero. A one bit, is inserted with an increase in this standard deviation. The following rules are used for embedding: a. If watermark bit is zero, b.To insert one bit, the sub band coefficients will be changed as equation Where, sign is the sign function. ?i is a positive constant. # Performance Evaluation a) PSNR The phrase peak signal-to-noise ratio, often abbreviated PSNR, is an engineering term for the ratio between the maximum possible power of a signal and the power of corrupting noise that affects the fidelity of its representation. Because many signals have a very wide dynamic range, PSNR is usually expressed in terms of the logarithmic decibel scale. The PSNR is most commonly used as a measure of quality of reconstruction of lossy compression codecs (e.g., for image compression). The signal in this case is the original data, and the noise is the error introduced by compression. When comparing compression codecs it is used as an approximation to human perception of reconstruction quality, therefore in some cases one reconstruction may appear to be closer to the original than another, even though it has a lower PSNR (a higher PSNR would normally indicate that the reconstruction is of higher quality). One has to be extremely careful with the range of validity of this metric; it is only conclusively valid when it is used to compare results from the same codec (or codec type) and same content. It is most easily defined via the mean squared error (MSE) which for two m×n monochrome images I and K where one of the images is considered a noisy approximation of the other is defined as: The PSNR is defined as: Where, m is the number of watermark bits and BER represent bit error rate. Embedded watermark is extracted completely, if BER is equal to zero. In digital transmission, the bit error rate or bit error ratio (BER) is the number of received bits that have been altered due to noise, interference and distortion. It is find out by the difference in the no. of bits that are embedded in the original image and the watermarks that are extracted from the watermarked image. # SI. No. # Attacks # CONCLUSION aND FUTURE ENHANCEMENT Robustness and fidelity are two major requirements in image watermarking that have opposite effect on watermarked images. There are a few methods that provide high robustness against other common image processing attacks, simultaneously. The aim of this work is to improve the robustness against common image processing attacks. In this regard, the watermarking technique is improved by selecting the larger coefficients of third level Contourlet transform to ensure higher robustness and determine the variation of coefficients constants using Multi-Objective Genetic Algorithm. MOGA provides a good balance between the robustness and the fidelity requirements, but it increases the running time in embedding phase. To show the superiority of this method, experimental results obtained by watermarking is compared with standard image using this method. As shown in experimental results section this method shows improvement in robustness against common image processing attacks (blurring, histogram equalization, sharpening, noise addition and median filter). This algorithm can be implemented in latest transform domains such as Curvelet and D-Tree Complex Wavelet to use their potential capabilities. In curvelet selecting more coefficients in multiple directions is possible and in D-Tree complex wavelet using the coefficients in both phase and magnitude is possible. In addition, introducing new approaches for specifying the significant triplets can result to select more reliable coefficients (in term of robustness) for watermarking. 1![Figure 1: Watermarking Technique II.](image-2.png "Figure 1") ![Laplacian Pyramid (LP) and Directional Filter Bank (DFB). After Laplacian Pyramid, each copy or not depending on the contents of the watermark. Another application is in source tracing. A band pass image is passed through a series of ©2011 Global Journals Inc. (US)](image-3.png "?") 3102![Figure 3: Embedding watermarks d) Implementation Of Genetic Algorithm i. Initialization The Chromosomes are initialized by using the population size, no. of Objective functions, no. of decision variables, minimum and maximum range of decision variables. Function f = initialize variables (N, M, V, min_tange, max_range) Where, N -Population size M -Number of objective functions V -Number of decision variables. min_range -A vector of decimal values which indicate the minimum value for each decision variable. max_range -Vector of maximum possible values for decision variables.ii.Evaluation function](image-4.png "Figure 3 : 10 MayFigure 2 :") 5![Figure 5: Contourlet decomposition y = dfbdec(x, fname, n).Where, X: input image. Fname: filter name.n: number of decomposition tree levels.](image-5.png "Figure 5 :") 67113![Figure 7: Extraction Process Steps 1. The watermarked image is decomposed to sub bands LL3, HL3, LH3 and HH3 using three-level Contour let transform. 2. The significant Co efficients are extracted using the positions which are used in the step 3 of embedding algorithm. 3. For each Contour let sub band, the standard deviation of its coefficients is calculated and then is compared to a given constant threshold.The watermark bit is considered 1 (0) if the standard deviation is higher (less) than the threshold. 4. The extracted watermarks sequence is compared with the embedded watermarks sequence. The similarity of two sequences, is calculated as,](image-6.png "6 .Figure 7 : Extraction Process Steps 1 . 13 May") 68![Figure 6: Extractions of Watermarks](image-7.png "Figure 6 :Figure 8 :") 9111314101415![Figure 9: Image to be merged c) Image values](image-8.png "Figure 9 :Figure 11 :Figure 13 : 14 MayFigure 10 :Figure 14 :Figure 15 :") 17![Figure 17: Calculating PSNR with Pixel ratio PSNR of Watermarked image (45.74db), Wavelet (38.24db)](image-9.png "Figure 17 :") 2Operations of Cross overSelected IndividualsS. No.G4G3G2G1FitnessC3110012C4101010After Cross overS. No. G4G3G2G1FitnessC34111014 32011May11Volume XI Issue IX Version IGlobal Journal of Computer Science and Technologyk ? 5BERBER(Contourlet)(Wavelet)1.Blurring10/7214/722.Median13/7217/72filtering3.Histogram2/724/72Equalization4.Sharpening11/7216/72 MayRobust Image Watermarking in contourlet Domain Using Genetic Algorithm ©2011 Global Journals Inc. (US) MayRobust Image Watermarking in contourlet Domain Using Genetic Algorithm This page is intentionally left blank ©2011 Global Journals Inc. (US) * Robust Digital watermarking scheme using Contourlet transform B ChandraMohan S.SrinivasKumar International Journal of Computer Science and Network Security 2008 Korea * A novel non-redundant contourlet transform for robust image watermarking against non geometrical and geometrical attacks AT SGuiduo Duan XiHo Zhao 5th International Conference on Visual Information Engineering Xi'an, China 2008. 29 July-1 Aug. 2008 * The contourlet Transform: An efficient directional multi resolutional image representation MinhNDo MVetterli IEEE transactions on Image Processing New York, NY Institute of Electrical and Electronics Engineers 2005 * ©2011 Global Journals Inc. (US) * Directional multiscale modeling of images using the contourlet transform DD .-YPo MNDo IEEE Trans. On Image Processing 2003 * An octave-band family of non-redundant directional filter banks PHong MJ TSmith IEEE proc. ICASSP 2002 2 * On Self-Adaptive Features in Real-Parameter Evolutionary Algorithm GeorgHans KalyanmoyBeyer Deb IEEE Transactions on Evolutionary Computation 5 3 June 2001 * A Fast Elitist Multi-objective Genetic Algorithm: NSGA-II KalyanmoyDeb AmritPratap SameerAgarwal TMeyarivan IEEE Transactions on Evolutionary Computation 6 2 April 2002 * Survey on multiobjective evolutionary and real coded genetic algorithms MMRaghuwanshi OGKakde Proceedings of the 8th the 8th * Asia Pacic Symposium on Intelligent and Evolutionary Systems 2004 * A review of watermarking principles and practices MLMiller IJCox JM GLinnartz TKalker IEEE International Conference on Image Processing 1997 * An SVD-based watermarking scheme for protecting rightful ownership RLiu TTan IEEE Trans. on Multimedia 4 1 March 2002 * A Robust Digital Image Watermarking Scheme Using the Wavelet-Based Fusion DKundur DHatzinakos 1997 International Conference on Image Processing 1997 1 icip * Multimedia data embedding and watermarking technologies MDSwanson MKobayashi AHTewfik Proceed. of the IEEE 86 6 1998 * Secure spread spectrum watermarking for multimedia JCox FTKilian TLeigton Shamoon IEEE Trans. on Image Processing 6 Dec. 1997 * The contourlet transform: An efficient directional multiresolution image representation MNDo MVetterli IEEE Trans. on Image Processing 14 12 Dec. 2005 * SPMaity MKKundu A blind cdma watermarking scheme in wavelet domain. IEEEInt. Conf Image Processing Oct. 2004