An Improved Digital Image Watermarking Scheme using Particle Swarm Optimization Technique

Abstract

Digital watermarking has been proposed as most effective ways to protect a copyright of digitally formatted multimedia documents in networked environments by inserting secret data into host image without significantly affecting the visual quality. Robustness to different accidental or intentional attacks is the most desired watermarking requirements for copyright protection application. This study presents a robust, blind and an imperceptible image watermarking scheme based on lifting wavelet transforms (LWT), discrete cosine transform (DCT), singular value decomposition (SVD) and particle swarm optimization (PSO). Block-based DCT and SVD are performed on reference sub-images which are formed by taking alternative pixels across row and column. A binary watermark is embedded in the first reference sub-image by modifying singular values and the other sub-image is used as a reference for both watermarks embedding and extracting. Different standard grayscale images have been used as test images and the obtained experimental results show improvement in both imperceptibly and robustness against various attacks.