A DWT Ordering Scheme for Hiding Data in Images Using Pixel Value Difference

Abstract

A scheme is proposed to hide data in images based on a prioritised ordering of the content of the host image. The embedding process uses the watermark strength to determine the ordering of the 16 regions resulting from the second level wavelet transform DWT decomposed content of the host image. The DWT decomposed sub-bands of the cover image were analysed, transformed using the Pixel Value Difference (PVD), and then ranked in terms of their ability to withstand changes that do not imperil the visual quality (PSNR) of the watermarked image depending on which an N×N-sized watermark stream is hidden in the 8 highest ranked subbands. Using the proposal, a 256×256 Lena image had the capacity to accommodate up to 96800 bits of a watermark logo while still maintaining an appreciable visual quality manifested by a PSNR value of 33.66 dB.

A Visual Complexity-sensitive DWT Ordering Scheme for Hiding Data in Images

“…The exquisite performance regarding the ease and quality of the embedding and extraction of data presented thus far and in (Al-Asmari et al however, not generalised for all the sub therefore there is the need to cull from the 16 sub that make up the best for both the watermark embedding Simple cluster Figure 16 shows the ordering of the sub-bands of the DWT decomposed host image representing the simple normal and complex clusters based on the In this figure, the rightmost letters (LL, HL, HL and DWT decomposed host image, while the remaining two letters specify the bands as explained earlier in 12 indicate the priority accorded band in the watermark embedding nce regarding the ease and quality of the embedding and extraction of data et al, 2012) are, for all the sub-bands and therefore there is the need to cull from the 16 sub-bands best for both the watermark embedding and extraction. Hence, the need to prioriti of the sub-bands based on the best choices of K. In deciding on the orderings in Fig the LL region of the DWT decomposed sub host image because it has been proven to have an adverse effect on the quality of watermarked images and extraction of second level decomposed cover images ones that are on the quality of watermarked images and extraction of hidden data in already marked ima Asmari et al, 2012;Salama et al, 201 The ordering suggested here guarantees high embedding capacity, good visual quality watermarked images and complete recovery of the hidden data, i.e., the watermark. Figure 17 shows the 10 sample images from the Corel 1000A dataset that were used together with the images located at the centre of the simple, normal and and extraction.…”

“…A consequence arising from the recent growth and advancements in the field of digital communication technology is the ease with which information can be produced, exchanged and copied. This has made the need for state-of-the-art algorithms to protect data and multimedia such as text, images, video, sound, etc., against un-authorized copying and tampering both mandatory and de rigueur (Al-Asmari et al, 2012).…”

“…In Al-Asmari et al (2012), we proposed an embedding process based on an ordering of the DWT decomposed content of the cover images as determined by the watermark strength. The proposal exploits the relationship between the strength of the watermark (or gain factor), K and the imperceptibility of the resulting watermarked images in order to prioritize the ordering of sub-bands of the second order decomposed content of the cover image for the watermark embedding procedure.…”

“…To accomplish this, our attention is mainly focused on understanding and utilizing the notion of visual complexity of images. Using the Corel 1000A dataset, we generalized the SND visual complexity space (Le et al, 2012(Le et al, , 2011 correlated it with the watermark strength dependent DWT-based watermark embedding in (Al-Asmari et al, 2012). Based on these generalizations, we deduced a relationship between visual complexity of an image and the watermark strength required for effective data hiding.…”

Reversible and Embedded Watermarking of Medical Images for Telemedicine

“…In [4], the difference expansion technique is applied to the high-frequency subbands of integer wavelet transform (IWT) in order to increase the peak-signal to noise ratio (PSNR) of the watermarked images. In [5], a discrete wavelet transform ordering scheme is proposed to increase the watermark embedding capacity. To further increase the capacity, an effective way is to define a region of interest (ROI).…”