Image Secret Sharing - Visual Cryptography and Secret Image Sharing - page 380

a plurality of ways. Images of different qualities can be extracted, depending
on the need for quality as well as the computational resources available. For
instance, images with loss of contrast are reconstructed by merely stacking
the shares; a simple yet effective bit-wise operation can be applied to restore
the halftone image; or images of perfect quality can be restored with the aid
of the auxiliary look-up table. Visual cryptography has been extended to al-
low for multiple resolutions in terms of image quality. Different versions of
the original image of different qualities can be reconstructed by selectively
merging the shares. Not only this, a spatial multiresolution scheme has been
developed in which images of increasing spatial resolutions can be obtained
as more and more shares are employed.
This idea of progressive visual cryptography has recently been ex-
tended [14] by generating friendly shares that carry meaningful information
and that also allows decryption without any computation at all. Purely stack-
ing the shares reveals the secret. Unlike [13] and [7] which require compu-
tation to fully reconstruct the secret, the scheme proposed in [15] has two
types of secrets, stacking the transparencies reveals the first, but computa-
tion is again required to recover the second-level secret. Fang's scheme is also
better than the polynomial sharing method proposed in [26] by Thien and
Lin. The method proposed in [26] is only suitable for digital systems and the
computational complexity for encryption and decryption is also a lot higher.
Currently, one of the most robust ways to hide a secret within an image is
by typically employing visual cryptography. The perfect scheme is extremely
practical and can reveal secrets without computer participation. Recent state-
of-the-art watermarking [8] can hide a watermark in documents that require
no specific key in order to retrieve it. We take the idea of unseen visible
watermarks and apply a secure mask to them and incorporate it for use within
the VC domain, thus improving the overall security that is currently one of
its weaknesses.
Weir et al. [31] also provide a mechanism for secret sharing using color
images as a base. The technique relies on visual cryptography as a mechanism
for sharing the secret. Many smaller secrets can be embedded within a color
image and a final share can be created in order to reveal all of the secrets. The
color image is visually similar to the original image before embedding due to
the high Peak-Signal-to-Noise Ratio (PSNR) achieved after embedding.
Another recent novel application for color image sharing and using color
images for secret sharing was presented by Weir and Yan [30]. Using Google
Maps, along with its Street View implementation, personally identifiable in-
formation can be obscured using visual cryptography techniques and can be
accurately recovered by authorized individuals who may need the information.
Specifically for law enforcement agencies who many find this type of informa-
tion helpful for a particular case. This type of practical application is very
important for the progression of visual cryptography, which presents a unique
way of using these techniques in a real-world situation.