Visual encryption by pixel property permutation

Abstract

Data security is a challenging problem in the sphere of storage and transmission of data. With the development of heavy computational resources, various encryption algorithms are getting more prone to subsequent attacks or threats either via brute force, cipher attacks or statistical attacks, and side-channel attacks. Thus, for better security there is a need for more robust encryption of data especially unstructured data. This paper has tried to propose a novel image encryption technique known as Pixel Property Permutation (PPP) on the basis of Visual Cryptography. In this algorithm, any image RGB or greyscale is first converted to Black and White and compressed of (200 * 200) pixels. Although this is not mandatory, one can even use the RGB image however this guideline is assumed only for simplicity purposes. The pixel values of the image are extracted and stored. These pixel values are then permuted with one another using random permutations. Thus, the encrypted image is obtained. Now, using the encrypted image and the original image a key is generated (superimposing one on the other). The key and the encrypted image are both subjected to the 128-bit AES Encryption (Roeder, Symmetric-Key Cryptography, [1]). The decryption process is also very similar. The two AES encrypted images are decrypted using the same key obtained earlier. In the next step we use the decrypted AES image and the cipher key image to recover the original image by superimposing one on top of another. This process of PPP uses AES encryption algorithm as an added layer of defence which is not essentially necessary but encouraged, this technique is cost effective and can be used to secure highly sensitive data.