Optimal frame selection-based watermarking using a meta-heuristic algorithm for securing video content

Abstract

Optimal embedding factor selection is still an open challenging issue in video watermarking. To address the same, this paper introduces a modified gravitational search algorithm (MGSA) based video watermarking (VW) scheme, termed VW-MGSA. In this proposed method, a novel variant of gravitational search algorithm i.e MGSA is employed to attain multiple optimal embedding factors (MOEF). VW-MGSA embeds watermark logo into maximum entropy blocks of size 8 × 8 followed by 1-level RDWT and Schur transform. The proposed GSA variant (MGSA) was evaluated experimentally and statistically using 22 standard benchmark functions, covering unimodal, multimodal, and fixed-dimension categories. The performance has been assessed using key metrics such as mean, standard deviation, Friedman test, and convergence graphs. These results confirm that the proposed variant outperforms existing meta-heuristic algorithms. Moreover, VW-MGSA has been validated on 8 standard benchmark videos over 19 attacks and evaluated using PSNR, SSIM, and NC metrics. The experimental and statistical results confirm that VW-MGSA outperforms existing video watermarking methods. It significantly improves the balance between imperceptibility and robustness compared to existing methods, with a measured improvement of 39.63%. The improved performance of the VW-MGSA can be applied to real-world platforms like Netflix and Amazon Prime to safeguard licensed content, with watermarks aiding in tracing piracy sources.