Block SMRT and knapsack optimization-based sequency selector for robust, imperceptible, and payload-efficient color image watermarking for binary watermark

Abstract

Watermarking is a generic strategy for overcoming numerous issues associated with multimedia security. Performance of a watermarking system in terms of imperceptibility, robustness and payload is highly dependent on the positions used for embedding. The choice of sequency packets plays an important role in the performance when Sequency-based Mapped Real Transform (SMRT) is used for embedding. A novel sequency selector for robust, imperceptible and payload-efficient color image watermarking of binary watermark is proposed using SMRT and two-level optimization. SMRT of block partitioned Cb channel of cover image is input to sequency selector where results of first-level optimization, based on imperceptibility and robustness, along with payload is optimized using knapsack optimization in the second level. The binary watermark is LSB embedded in SMRT coefficients present in each packet of the optimal combination of sequencies chosen by the sequency selector. Simulation results show better performance compared to existing optimized and non-optimized embedding techniques.