Analysis of robust watermarking techniques: A retrospective

Abstract

In this paper we focuses extensively on review of achievement of various non-blind invisible watermarking techniques. The main intention of the proposed techniques is to embed color logo into color host images or videos using transform domain techniques. And this concept is extended to video watermarking with multiple logos such as gray and color logos, to evince the supreme capacity of watermarking algorithm with least perceptual degradation of host video. In addition, watermarking algorithm was subjected to 12 different intentional and unintentional attacks such as frame averaging, dropping and swapping, compression and noise attacks, and common image processing attacks, which revealed that the hybrid video watermarking technique with multiple logos to be superior. Moreover, it is concluded that the projected technique is more robust under various subjected attacks. The most important objectives of digital watermarking techniques are capacity, robustness and imperceptibility. In all the proposed watermarking schemes, imperceptibility and robustness are estimated by measuring Peak Signal to Noise Ratio (PSNR) and Correlation Factor (CF). And this paper conveys the potency of the existing algorithms against various attacks.