A Spread-Spectrum Watermarking Scheme with Adaptive Embedding Strength and PSNR Guarantee
Ying Huang, Hu Guan, Baoning Niu, Shuwu Zhang · 2018
Watermarking is a crucial technique for copyright protection of digital images. The most common spread-spectrum watermarking schemes suffer from host signal interference (HSI), which degrades the decode performance. Although this condition can be improved by increasing the embedding strength, it might make the embedded watermark perceptible. This paper solves these problems by proposing a novel blind spread-spectrum watermarking scheme with adaptive embedding strength (SSAES). A notion called the error threshold is introduced to analyze how HSI and distortions compromise robustness and imperceptibility of the spread-spectrum scheme, which motivates us to come up with our solutions. SSAES is HSI free by utilizing the priori knowledge of HSI and features an adjustable embedding strength derived from a theoretical equation related to peak signal-to-noise ratio (PSNR), which strikes a perfect balance between robustness and imperceptibility. SSAES can be easily applied to other watermarking frameworks. The experimental results demonstrate our scheme can ensure that the PSNR of each watermarked image is more than 40 dB and is superior to the state of the art method under the same watermarking framework in any simulation scenarios include image processing attacks and geometric attacks.