Multiobjective Cost Function Based Digital Vide Watermarking Technique

Abstract

: Video watermarking is a key process for copyright protection of digital data. In the video sequence, there is a huge quantity of data and these data are more susceptible to attacks. Hence, this paper intends to develop a novel video watermarking technique with the aid of multi-objective cost function. Initially, the keyframes are extracted from the input video sequence and these extracted video frames are subjected to wavelet transform for achieving the wavelet coefficients. Then, a multi-objective cost function is proposed with the help of multiple criteria like edge, brightness, the intensity of pixel, coverage, energy and frequency coefficient. The bit plane technique is utilized here to develop multiple binary images for the secret message by means of partitioning the secret messages. Further, with the aid of the multi-objective cost function, the message bit is embedded in the frame (wavelet coefficient). The embedded image is transmitted from the sender to the receiver via a communication channel and on the receiver side, the retrieval process takes place. The original image is retrieved in the receiver side with the help of a multi-objective cost function. Finally, a comparative analysis is accomplished between the proposed wavelet+Cost (wavelet transform based on a multi-objective cost function) and the existing models like wavelet model, Linear Significant Bit model (LSB) and LSB + cost model in terms of peak-to-signal noise ratio (PSNR) and correlation coefficients.