Robustness of Modified Non-Separable HaarWavelet Transform and Singular Value Decomposition for Non-blind Digital Image Watermarking

Abstract

Security is a matter of significant concern for image media. An effective way to protect images is by digital watermarking. This paper introduces a non-blind digital watermarking scheme using modified non-separable Haar wavelet transform (NSHWT), singular value decomposition (SVD), Arnold's cat map, and Rabin-p cryptosystem to embed a binary watermark image into a color cover image. Aside from robustness, security is also prioritized in the scheme. High robustness is achieved using two transform domain techniques, discrete wavelet transform (DWT) and SVD, while security is heightened with the double encryption by Arnold's cat map scrambling and Rabin-p cryptosystem. A disadvantage of the discrete wavelet transform in digital image watermarking is the calculation complexity and the high cost of changing memory and time. The proposed algorithm uses a modified NSHWT instead of the traditional method of DWT. Therefore, the load of the process on the hardware can significantly be reduced while still maintaining the advantages of DWT. The algorithm is objectively evaluated in terms of imperceptibility, robustness, and embedding capacity for both binary and color watermarks, as well as efficiency. It has been concluded that the proposed scheme performs competently compared with other recent watermarking techniques based on DWT and SVD.