Robust Video Watermarking Using Secret Sharing, SVD, DWT and Chaotic Firefly Algorithm

Bhargavi S, Venkata Dasari, Damodaram Avula
2018 International Journal of Intelligent Engineering and Systems  
Watermarking is the process to protect and discourage the illegal sharing of multimedia content through digital data sharing websites over INTERNET and also to prove ownership of the multimedia content. Though various watermarking solutions are available to safeguard the digital media, at the same time many methods exist to weaken the strength of watermark hence ownership is disproved and can be shared illegally. Strength of watermark can be increased to make it to robust against these methods
more » ... t the cost of perceptual quality. So, a compromised approach is required. This paper proposes a video watermarking solution to maintain robustness as well perceptual quality through optimization. This method adds secretly shared watermark bits to singular values of the discrete wavelet coefficients with proper scaling factor, which is selected by using the Chaotic Firefly Algorithm optimization method. Despite many watermarking methods exist on combination of Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD), these are not effective against all filtering attacks as these are not using all frequency coefficients either for embedding or secret sharing. This method generates secretly shared watermark based on singular values in DWT domain to make the system more robust against filtering attacks and video compression techniques. This solution leverages chaotic firefly algorithm to compromise robustness and perceptual quality of the input along with SVD and DWT. The performance of method is measured by evaluating it on a dataset comprised of 140 videos of various genre and compared its performance with state of art methods. Eventually, proved the method is performing well when compared to state of art methods with respect to robustness as well as quality at the cost of computation. The stated results in experimental section shows us that watermark can be retrieved even if watermarked audio undergoes several attacks like compression at lower bit rates, frame drop attack and resize attack etc.
doi:10.22266/ijies2018.0228.18 fatcat:f73msivhrrcltothf7372tybny