An Efficient Discrete Wavelet Transform Based Hybrid Image Watermarking Algorithm Using Human Visual Model

Abstract

More than ever with growing of multimedia technology, the digital data are exchanged in the internet, which can be duplicated by unauthorized users. To avoid this problem the watermarking technology has brought this paper. Performance improvement with compare to existing algorithm is obtained by proposing a new watermarking algorithm based on Human Visual Model (HVM) and Discrete Wavelet Transform (DWT) for securing the digital data and copyright protection. DWT is applied to the input image and at each level of DWT sub bands to embed a watermark image in selected coefficients of the sub band. HVM integrate the weight factor effect of human visualization by considering into the eye’s vision is less sensitivity area, depending on brightness, frequency band and texture areas of the image sub band. In the proposed approach robustly and imperceptibly, DWT and HVM are used for obtainable weight factor, according to human eye perceptual and to determine the optimal strength at which the threshold to embedding reaches the perceptual invisibility of watermarked image to various attacks. Performance is evaluated such as Peak Signal to Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM) and Quality Index (QI) is used to evaluate and achieve the imperceptible watermarked image. Results of PSNR values of hybrid image watermarking are between 49.73db to 36.24db. Experimental results show that our hybrid image watermarking process has enhanced robustness and displays the effectiveness of presenting images watermarking system.