A hybrid semi-blind gray scale image watermarking algorithm based on DWT-SVD using human visual system model

Abstract

To achieve good imperceptibility and robustness, a hybrid image watermarking algorithm based on discrete wavelet transform (DWT) and singular value decomposition (SVD) is proposed using the characteristics of human visual system model for copyright protection and authenticity. In the proposed watermarking algorithm, one level DWT is applied to selected image blocks to obtain four sub-bands of each block and then the U component of low frequency sub-band (LL) obtained after SVD transformation is explored under different threshold values for embedding and extracting the watermark. The experimental results show that HVS model based hybrid image watermarking scheme is imperceptible and robust against several image processing operations like JPEG compression, median filtering, sharpening, cropping and addition of Gaussian noise. Peak signal to noise ratio (PSNR) and bit correction rate (BCR) are used to measure the quality of watermarked image and extracted watermark respectively.