A hybrid block based watermarking algorithm using DWT-DCT-SVD techniques for color images

Abstract

In this paper, we propose a robust watermarking technique which combines features of Discrete wavelet transformation (DWT), discrete cosine transformation and singular value decomposition. In this technique DWT is used to decompose the color images into various frequency and time scale. Block DCT is applied on DWT coefficients of various frequency to provide high level of robustness. DCT transformed blocks of size 4x4 are further decomposed using dual SVD technique to get singular values in which watermark is to be hidden. Algorithm is tested using Lena test image in RGB and YIQ color model. As per the results combining feature of DWT-DCT with SVD technique provides robustness against image processing and geometric attacks in YIQ color model than RGB color model.

Adaptive watermarking technique based on human visual system and fuzzy inference system

“…Remark 4: Small perturbation in image does not cause large variation in singular values [3]- [7], [10], Remark 5: Singular values own intrinsic properties of image (i.e. luminance information is contained in singular values, whereas geometric information is maintained by corresponding singular vectors) [1], [3]- [7], [11].…”

Robust Multiple Watermarking Technique for Medical Applications using DWT, DCT and SVD

“…The discrete wavelet transform (DWT) [7] is a multiresolution description which consists of decomposing image into four frequency representations. These four representations are called subbands LL (approximation), LH (vertical), HL (horizontal) and HH (diagonal).…”

Exploring DWT–SVD–DCT feature parameters for robust multiple watermarking against JPEG and JPEG2000 compression

“…In [29], Santhi et al performed SVD on the watermark such that W = U w S w V w and then added S w to the singular value matrix S derived from to a block matrix in the DWT-DCT domain. The algorithms developed in [30][31][32][33] shared the basic similarity.…”

“…The developed algorithms differ not only in embedding targets but also in processing procedures. Among those starting from the use of DWT to decompose the host image into LL, LH, HL, and HH subbands, Wang et al [28] and Santhi et al [29] chose the LL subband for watermark embedding. Divecha and Jani [30] selected the HH subband, while Murty and Kumar [31] picked the HL subband.…”

“…In fact, several attempts [28][29][30][31][32][33] have been made on the development of robust watermarking algorithms in a hybrid domain involving the DWT, DCT and SVD. The developed algorithms differ not only in embedding targets but also in processing procedures.…”