Abstract:
In this article, a visual cryptography method has been presented which exhibits significant improvement over previous approaches in terms of computation cost, volume of information and the quality of the recovered images. By means of this algorithm, JPEG images, without having to be decompressed and then encrypted, are directly converted to distinct shares and placed within other images as covered images. In this way, it is no longer necessary to decompress the JPEG image and to encrypt each share separately and this cuts down on computation overhead. Another advantage of our work is that, contrary to previous methods which were used to conceal two pieces of share images in other images with 9 times the size, here we were able to conceal them in images with a smaller size (i.e. 4 times the size). Also in contrast to lower quality of covered images in previous methods, in this approach, no tangible change of quality is noticed in covered images and thus, they don’t raise the curiosity of the invader.

Keywords: Visual cryptography, Secret image, Pixel, DCT coefficients