An Efficient Construction of Extended Visual Sharing Scheme using Error Diffusion

Abstract

Visual Cryptography (VC) is a cryptographic technique for encrypting visual information like written materials, textual images, handwritten notes print and scanned images etc. in a perfectly secure way. The information is coded into shares/meaningful shares and distributed. The advantage of this cryptographic technique is that the decryption is performed by the human eye. It also ensures security, in a way that hackers cannot perceive any clues about a secret image from individual shares. Most of the existing color visual cryptography schemes are not  atisfactory in terms of producing either meaningless shares or meaningful shares, leading to suspicion of secret encryption. This paper introduces an efficient extended method of encrypting color visual information into n-color images producing n-meaningful shares of the covert images of high visual quality. This is based on two principles - retaining original pixels and error diffusion. Original pixel retainment synchronizes the positions of pixels that carry visual information throughout the color channels of the color image so that the original pixel values of the visual information are retained even after encryption. This ensures high quality shares that are visually identical to that of originals. Error diffusion is used to construct the shares in
such a way that the noise introduced by the preset pixels are diffused away to neighbors during encryption which results in the shares pleasant to human eyes. The effectiveness of the proposed method is validated using PSNR and WSNR values.