International Journal of Managing Public Sector Information and Communication Technologies (IJMPICT)Vol. 4, No. 2, June 2013
originalimage.DWTbased compression offers scalability so image can bedividedinto
four sub bands in every level of decomposed and by choosing ofthe sub-bandtodevelop a hybridwatermarkingscheme forimprovingtherobustness, imperceptibility and capacityand help todevelop a new hybrid method.In the first method (ref. 9) the original image is segmented intoblocks then find outspatial frequency of each block select reference image undercertaincondition then apply DWT, DCT and SVD transformation.In the second method (ref. 3) theoriginalimage is segmented into blocks then find out spatial frequency of each block selectreference image under certain condition then apply DWTand SVD transformation.In theproposed method,sincehigh band is considered so it fulfills the requirementsimperceptibilityand robustness.Most of the domaintransformation watermarking techniqueworkswith DCT,DWT, SVD and their mixing algorithm such as DWT-DCT, DCT-SVD and so on.Inthis paperwe proposed adigital watermarkingtechniqueusingDWT-DCT andSVDtransformation. Thismethod isprovided a good imperceptibility and high robustness againstvarious kindsprocessingattacks.Therest ofthe paper is organized as follows:Section2, focusesonoverview of transforms for watermarking. Section 3,gives details theproposed methodologyandwatermarkingalgorithms. In section 4,gives experimental results andcompares. Insection 5,conclusion is drawn.
Asstatedearlier that transform domain based watermarkingscheme is always abetter choice thanspatial domainbasedwatermarkingscheme. Thiscan be doneby using different transformationlikeDCT, SVD andDWT.In this section, we will briefly describetheDCT, DWT and SVDtransformationsin below.
2.1. Discrete Wavelet Transform (DWT)
The basic ideaofDWTin whicha one dimensional signalisdividedintwo partsone ishighfrequency partandanotheris low frequencypart.Then the low frequency part is split into twoparts andthe similarprocess will continue until the desiredlevel.The high frequency partof thesignal is contained by the edge componentsof the signal.In each level of theDWT (DiscreteWavelet Transform)decomposition an imageseparatesinto fourparts these areapproximationimage (LL) as well as horizontal(HL), vertical (LH) and diagonal (HH)fordetail components.Inthe DWT decomposition input signal must be multiple of 2
. Where, n representsthe number of level.To analysis and synthesis of the original signalDWT provides the sufficient informationandrequireslesscomputation time.Watermarks are embeddedinthese regionsthat helptoincrease the robustness of thewatermark. Aone level DWTdecomposition processis shown inFigure 1.
Figure1.One level DWTdecomposition process
2.2. Discrete Cosine Transform (DCT)
The DCT is the most popular transform function used in signal processing. It transforms a signalfrom spatial domain to frequency domain. Due to good performance, it has been usedin JPEGstandard for image compression. It is afunctionrepresents atechniqueapplied to image pixels in