Video object watermarking using 3D-Walsh Hadamard transform and Arnold transform

This paper presents a new approach for watermarking of digital video MPEG-4 providing robustness against geometric distortion and JPEG compression at the same time with blind detection remains especially challenging. We propose in this work a robust video object based watermarking scheme with scrambled watermark, based on three dimensional Walsh Hadamard Transform (3D-WHT) and Discrete Wavelet Transform (DWT). Firstly, we divide the initial video to cubes non-overlapping. The 3D-WHT is applied on these cubes and the DWT is performed on the first rows of each cube. The watermark is embedded using a classical additive scheme into the coefficients of the HH subband in the DWT domain. Finally, the inverse 3D Walsh Hadamard Transform (3D-IWHT) is performed thus obtaining the watermarked video. The detection is done using scrambled watermark. Robustness of the proposed method against a variety of both intentional and unintentional video attacks has been observed through experimental results.