Performance Analysis of Image Watermarking for Different Sub-bands Using LWT and Arnold Map

Abstract

In this paper, blind image watermarking is proposed for grey-scale images using LWT and Arnold maps. A comprehensive analysis of robustness and imperceptibility for different sub-band is analyzed, and a robust sub-band against different attacks is determined for designing a system robust against intentional attacks or combined attacks. The importance of embedding a watermark in several sub-bands has been examined to increase the robustness against various image attacks while retaining a reasonable level of imperceptibility. During the study, robustness is analyzed and watched against the number of attacks such as compression attacks, noisy attacks, de-noising attacks, and geometric attacks. It is moreover seen that higher sub-bands are seen to offer good imperceptibility, and robustness performance depends on the nature of attacks. It has also been noticed that entire attacks affect the watermarked image in a different way. A standard image dataset is used to test the suggested concept, and it is discovered that sub-band 1 performs admirably for strength (robustness) and imperceptibility against different image attacks.