Embedding Dual Watermark in the G-Channel Image Using DCT and 2-D Walsh Code

Abstract

This paper uses two dimensional Walsh coding for blind twofold watermarking algorithm which mixes robust copyright watermark and authentication information into various color images to maintain the copyright ownership of the host images. First watermark i.e. fingerprint image is embedded in the DCT low frequency coefficients of the G-channel image of host image where as the second watermark i.e. mobile phone number is embedded in the DCT block of G-channel image by modifying the low frequency band, excluding the DC components. During the decoding process the primary fingerprint image is used as authentication matching for extraction of mobile phone number. The new algorithm uses extra security by the use of biometric authentication process of decoding. The two dimensional Walsh coding and decoding is used to maintain the robustness. The new algorithm produces more similarity structure index measure (SSIM) and peak signal to noise ratio (PSNR) values. Various JPEG quality factors with and without Walsh coding are tested below which watermark can be recovered. The algorithm was successfully tested on a variety of 512X512 colour images and also verified against various attacks through StirMark attacks and their NC values.

Keywords: Watermarking, DCT, G-channel, colour images, 2-D Walsh code 

Cite this Article

Mishra BP, Pratihari HN, Das P. Embedding Dual Watermark in the G-Channel Image Using DCT and 2-D Walsh Code. Current Trends in Signal Processing. 2015; 5(2): 27-34p.