A multilayer visual cryptography framework for secured secret messages transmission

Visual Cryptography acts as an encryption technique, which helps to store secret data within an image and then split this image into a decided number of slices. Each and every authorized user will receive a single slice of the encrypted image containing a portion of the cipher text. The message cannot be deciphered or read without superimposing (combining) all the slices in the proper order. Many existing papers which concentrate on the processes during and after image-slicing, though considered being effective to a certain extent, these individual concepts are a major reason for compromising the security and privacy of a cryptosystem. It is possible to detect cipher text in any secret image using standard decryption tools. Our proposed model not only strengthens the cryptosystem by integrating random key generation technique using key bunch matrix, and an improved S-DES algorithm but also makes the cryptosystem hard to crack. The generated cipher text is embedded into an image, which is then sliced into a number of parts say `n' different parts, thereby fully safeguarding the inbuilt-secret message. The secret text can only be interpreted or deciphered when all the slices of the image are securely exchanged between the authorized users and superimposed in the correct order. The earlier limitation is solved as image slices cannot be identified by tools as containing partial cipher text (acts as junk bits). Also, it needs permission from all authorized users for interpreting cipher text into plain text. Thereby, our proposed cryptosystem has increased the overall efficiency and strengthened the authentication and security barrier.