A SECURE AND EFFICIENT MECHANISM FOR MULTIPLE SECRET SHARING IN VISUAL CRYPTOGRAPHY

Abstract

Cryptography is the art of protecting information by conversion of the data into a disordered code that can be sent across a public or private network and deciphered with the correct key which is kept with the designated receiver. An array of cryptographic techniques has been proposed for providing data security. However, most of the traditional cryptographic methods require complex algorithms for encryption and decryption. Visual cryptography is a technique for sharing images between two or more participants. It achieves the goals of security such as confidentiality and authentication without any complex computations. Here an image that has to be kept confidential (secret image) is divided into certain number of images called shares. The concept of share eliminates the use of key and encrypted data used in conventional cryptographic methods. The secret image can be reconstructed just from the stacking of the shares. Traditional visual cryptography methods produce random shares which are susceptible to attackers. Some methods have been proposed to generate innocent-looking shares so that attacker does not get doubtful by looking at the random pattern of the share used in conventional visual cryptography methods. The shares look like valid images and the adversary cannot judge whether they are part of secret image. However many of these techniques use additional data structure and take much time in encoding the secret. In this dissertation entitled "A SECURE AND EFFICIENT MECHANISM FOR MULTIPLE SECRET SHARING IN VISUAL CRYPTOGRAPHY", a secure and efficient visual cryptography technique is proposed. The proposed addresses the major challenges in producing meaningful shares for secret sharing process. This technique overcomes the problems of privacy invasion, high computation time and storage requirement. The proposed technique uses the concept of meaningful shares to enhance the security of the scheme. It uses an efficient algorithm to improve performance and needs no additional data structure to reduce storage overhead. The technique is extended to share multiple secrets together. This reduces the number of shares produced and the storage requirement is further optimized.