Frame selection based video watermarking technique based on DWT, K-SVD, rail fence to address quality loss of data

The advances in Internet and multimedia have led to many challenges like copyright protection and ownership identification. Multimedia media data such as audio, images and videos are getting distributed illegally without the knowledge of owner; this has led to development of many techniques including watermarking. The proposed work deals with application of a watermarking technique on videos that includes identification of frames from the videos to be watermarked along with embedding of an encrypted binary watermark on frames followed by testing of the watermarking technique by imposing various attacks to check the validity of the proposed technique. The watermarking is done by using hybrid approach of DWT and KSVD and rail fence is used as encryption to the watermark before embedding into multimedia data. The major challenge after insertion of the watermark on multimedia data is the loss of quality, which is addressed in proposed work. The results are formulated by calculating values of peak signal to noise ratio (PSNR).