A Novel Robust Approach of Optimized Hybrid Watermarking Scheme for Color Images Using DWT, DCT, Permutation, and PSO

Abstract

This paper presents a novel digital image watermarking technique for color images that integrates discrete wavelet transforms (DWT), discrete cosine transforms (DCT), permutation scrambling, and particle swarm optimization algorithms (PSO). In the trade-offs of resilience and imperceptibility, it is a method that will be extended to improve both for efficient watermarking. To improve the security, we first randomize the watermark image using a permutation and then perform DCT on it to obtain its coefficients. The coefficients are embedded through the optimal scaling factor estimated by the PSO into the DWT sub-bands of the host color image. This optimization provides a trade-off with a reasonable amount between the fidelity and robustness of the watermarked image. The proposed approach surpasses 49 dB in peak signal-to-noise ratio (PSNR), hence indicating superior imperceptibility of the watermark. Moreover, it has significant resilience against several conventional image processing methods, with normalized correlation coefficients (NCC) consistently around 1.0. An optimizing approach using PSO ensures little visual distortion and maintains robustness, overcoming limitations observed in alternative approaches. Experimental results validate the technique's efficacy in improving both the invisibility and strength of digital image watermarking.