Enhancing Visual Cryptographic Shares using RSA Algorithm

Enhancing Visual Cryptographic Shares using RSA Algorithm

Rima Saha

Department of Computer Science & Engineering SJB Institute of Technology

Bangalore, India rimsaha42@gmail.com

Abstract-Visual cryptography is a cryptographic technique which allows visual information to be encrypted in such a way that decryption becomes a mechanical operation that does not require a computer.The demonstration of visual secret sharing is such that an image is broken into n shares so that someone with all n shares only could decrypt the image , while someone with n-1 shares reveals no information about the original image. each share is printed on a separate transparencies , and decryption is performed by overlapping the shares . When all shares are overlaid , the original image would overlaid . By taking the advantage of this property , third person can easily retrieve the secret image if shares are passing in sequence over the network . This project presents an approach of enhancing visual cryptographic security using public key encryption . RSA algorithm is used to provide double security . RSA is a cryptosysystem which is used for secure transmission . In cryptosystem the encryption key is kept public and private key is kept secret . Thus secret shares are not available in their actual form for any alteration by the adversaries to create any fake shares . The scheme provides more secret shares that are robust , again a number of attacks and the system provides a more security for handwritten text , images and printed documents over the public network .

Keywords—Visual Cryptography; Encryption; Information Security; VC shares

1. INTRODUCTION

   In todays information age, information sharing and transfer has increased exponentially. The threat of an intruder accessing secret information has been an ever existing concern for the data communication experts. With the rapid advancement of network topology, multimedia information is transmitted over the Internet conveniently. Various confidential data such as military maps and commercial identification are transmitted over the Internet. While using secret images, security issues should be taken into consideration because hackers may utilize weak link over public network to steal information that they want. To deal with security problems of secret images, we should develop some secure appropriate algorithm by which we can secure our data on internet. With this system visual information (pictures) can be secure over the internet with the help of Visual Cryptography.

   The proposed scheme combines the advantages of both Visual

   Cryptography as well as Public Key Cryptography. This scheme enhances the security of VC shares by encrypting with Public Key Cryptography [14][15], which provides the strong security to the transfer of secret information in form of images, printed text and hand written material.

   Visual Cryptography (VC) is a special encryption technique used to encrypt images in such a way that it can be decrypted by the human visual system if the correct key image

   Shambhawi

   Department of Computer Science & Education SJB Institute of Technology

   Bangalore, India shambhawi1512@gmail.com

   are used. The technique was proposed by Moni Naor and Adi Shamir[6] in 1994. According to them Visual Cryptography is a method of encrypting a secret image into shares such that stacking a sufficient number of shares reveals the secret image. Shares are binary images usually presented in transparencies. Unlike conventional cryptographic methods, VC needs no complicated computation for recovering the secret image. The act of decryption is to simply stack shares and view the secret image that appears on the stacked shares. Visual Cryptographic technique is being used for secretly transfer of images in army, hand written documents, financial documents, text images, internet- voting etc.

   VC shares exist in their actual form during the transmission over network. However, directly third person cannot guess the secret information with any single share, but there is a possibility of retrieval if hackers are able to collect all the shares passing in sequence over the network. Thus to get rid of this problem, we need to enhance the security of shares. For the same purpose we have used Public Key Cryptography in addition to Visual Cryptography so that even if hackers are able to get all the shares but they cannot retrieve the original secret without the access of private key.

2. RELATED WORK

   Various researches have been carried out in this area to increase the security & visual quality of the secret image. Some of them are as follows:

   Néelima Guntupalli et al [5] presented survey of various Visual Cryptographic Schemes and established the conceptual knowledge about Visual Cryptography.

   Yogesh Bani, Dr. B.Majhi, Ram S. Mangrulkar [13] proposed a novel approach for Visual Cryptography using Data Hiding by Conjugate Error Diffusion watermarking technique. Two shares have been generated and then embed into the cover image x with the help of watermarking. Secret and cover images have been revealed after overlapping shares. Cover image consume extra storage space. Intruder can attack on the shares to reveal the secret, which causes disturbance in the pixels of original image and the receiver will not get the actual secret. At the receiver end

   the cover image and secret both will be revealed, so the quality will be very poor of retrieved secret image.

   Debashish Jena, Sanjay Kumar Jena [4] implemented Data Hiding using Conjugate Ordered Dithering (DHCOD) algorithm for generating the shares. A dithered halftone image generated by the cover image was the first share. For second share, some noise was added to the secret image and converted it to the binary image after that using share 1 and binary image they generated the second share. The secret image has been revealed with the simple AND operation of share 1 and share 2. Share generation process is made complicated by this method.

   B. Padhmavati, P. Nirmal Kumar, M. A. Dorai Rangaswamy [2] generated shares first by Visual Cryptography VC (2, 2) scheme. Then both shares were embedded into the cover images with the help of watermarking. For reveal of secret image, the extraction process was used to extract the shares from the embedded images. At last both shares were overlapped and revealed the secret image. Two cover images have been used to hide the shares which require extra memory space.

   M. Nakajima, Y. Yamaguchi [7] suggested Extended Visual Cryptography for natural images. Three input pictures have been taken; one is secret and other two for encryption. The encryption process is based upon determining the arrangements of transparent sub pixels on two images (used to conceal the existence of third secret image) according to the pixel transparencies, t1, t2 and tT. Where, tT is the transparency of target image. The secret picture is reconstructed by printing the two output images on transparencies and stacking them together. The problems with this technique are network overload due to two extra images and poor quality of revealed image.

   Wei-Qi-Yan, Duo Jin, Mohan S Kankanhalli [12] suggested a solution for superimposition of two shares. Some alignment marks are used in Walsh transform domain. It is always beneficial to use the scheme developed by this author, because in VC decryption stacking of two shares is mandatory and without exact alignment retrieval is not possible.

   Abhishek Parakh and Subhash Kak [1] suggested Recursive Hiding scheme for 2 out of 3 secret sharing. Secret bit is ivided into 3 pieces p1, p2, p3.

   For 0 p1=p2=p3 as 000,111, 222

   For 1p1p2p3 as 021,102,120,210,201etc

   Shares of smaller message are used to create shares of larger message. This scheme helps in decreasing the network load. Per pixel 9 bits expansion if the image size is multiple of 3, 16 bits expansion if image is multiple of 4 and so on this is not acceptable after a limit. Currently the efficiency of this system is

   33% which will decrease as the size will not be exactly in multiple of 3.

   Vaibhav Choudhary et al [11] discussed an Improved Pixel Sieve Method for Visual Cryptography used an additional sieve to generate shares. In this scheme Secret is hidden properly using this scheme but efficiency of this scheme cannot be evaluated as decryption algorithm and the results of retrieval have not been shown in the paper.

   Ujjwal Chakraborty et al [10] proposed two schemes for (2, 2) and (2, 3) visual cryptographic encryption. The first scheme considers 4 pixels of input image at a time and generates 4 output pixels in each share. The second scheme considers 2 pixels (1 block) of input image at a time and generates 3 output pixels in each share. The dimension of revealed image is increased by 1.5 times in horizontal direction and re mains same in vertical direction.

   Shyamalendu Kandar & Arnab Maiti [9] has proposed a technique of k-n secret sharing on color images. At the time of dividing an image into n number of shares, they have used random number generator. Minimum k numbers of shares are sufficient to reconstruct the image. If k numbers of shares are taken then the remaining shares are (n-k). In an image if certain position of a pixel is 1, then in (n-k) +1 number of shares in that position of that pixel there will be 1. In the remaining shares in that position of the pixel there will be 0. A random number generator is used to identify those (n-k) +1 number of shares. Secret is not properly hidden and it is easy to guess the contents in all three shares. If intruder is able to get the information about randomness, secret image can be retrieved.

   Chandramathi S, Ramesh Kumar R, Suresh R & Harish S [3] in 2010 concluded from the overview of all existing VC schemes that researchers should focus on good quality of reconstructed image & to increase security with minimum pixel expansion.

   P. S. Revenkar, Anisa Anjum and W. Z Gandhare [8] evaluated the performance of various Visual Cryptographic Schemes, which help in choice of best scheme according to the available bandwidth or color of secret image or level of security required. Following parameters have been used to evaluate the performance:

   No. of Secret images Pixel Expansion

   Image Format

   Type of shares generated

   As we have observed that conventional cryptography is not used to protect the shares. In some cases cover images are used to

   carry the secret share which is an extra overload on network. This limitation forced us to use Public Key cryptography whic h provides shares with change in actual information.

3. METHODOLOGY OF THE PROPOSED SCHEME The proposed scheme generates the VC shares using

   basic Visual Cryptography model and then encrypt both shares using RSA algorithm of Public Key Cryptography so that the secret shares will be more secure and shares are protected from the malicious adversaries who may alter the bit sequences to create the fake shares. During the decryption phase, secret shares are extracted by RSA decryption algorithm & stacked to reveal the secret image. As shown in Fig. 3.1, complete scheme is divided into following four phases:

   Secret Image

   Share 1

   Share 2

   Phase1

   Encryption using RSA

   Encryption using RSA

   Phase 2

   Decryption using RSA

   Decryption using RSA

   Phase3

   Decoded Secret Image

   Phase4

   Fig.3.1 Methodology of the Proposed Scheme

   1. PHASE-1 Generating shares of secret image: In this phase Visual Cryptography Encryption is implemented. It consists of generation of shares from secret image using VC (2,

      1. scheme. The secret image is first converted into a binary image then each pixel in the secret image is broken into 8 sub pixels, 4 pixels in each share by selecting the random pixel encoding scheme out of three given in Fig.3.2.

         Fig. 3.2. Pixel encoding schemes

         BIN_IMG, R_size, C_size

         .jpg image/.bmp image

         The algorithms for conversion of image into binary and share generation are given below:

         Image-conversion Input:

         Output:

         IMG=read () BIN_IMG=Convert_to_binary (IMG) [R_size C_size]=Cal_size (BIN_IMG)

         Algorithm 3.1 Image Conversion

         Share_generation

         Input: BIN_IMG, R_size, C_size

         Output: SHARE1 SHARE2

         For i=1 TO R_size Do

         For j=1 TO C_size Do

         Pix_enc_scheme=Rand_select() SHARE1=Pix_enc_scheme(BIN_IMG(i,j)) SHARE 2=Pix_enc_scheme(BIN_IMG(I,j))

         Done

         Done

         Algorithm 3.2 Share Generation

   2. PHASE-2 Encrypting the generated Shares: This is the second phase of our approach which will encrypt shares generated from the first phase. We have used RSA for encryption in this step. First we have generated the key for RSA and then performed the encryption. Results of this phase are encrypted shares.

   3. PHASE-3 Decrypting the Shares using RSA: This process takes place at the destination of the document/image/text. We again convert the encrypted shares in their actual form using RSA decryption algorithm, which were encrypted at the sender end.

   4. PHASE-4 Visual Cryptographic decryption: In this phase Visual Cryptographic decryption is performed. We have decrypted the original secret image by applying the binary XOR operation on both decrypted shares.

4. EXPERIMENTAL RESULTS

Proposed scheme has been implemented in MATLAB

1. To run this scheme minimum hardware configuration is required with no extra specifications. The experiments have been run in Windows 7 on a Sony VAIO laptop with Intel i5 2.4 GHz processor.

   To test the performance of this scheme number of experiments has been conducted with varying image sizes, types & ke ys but every time secret image is retrieved with good visual quality. The confidentiality of shares is also tested by super imposing the encrypted shares before reaching to the destination. Results of some experiments are shown in Fig.4.1, Fig.4.2 & Fig.4.3.

   1. Binary Input Image1

   2. Image1_share1 C. Image1_share2

1. Image1_Encrypted share2 E. Image1_Encrypted share2

   1. Revealed Image from Encrypted shares

   2. Image1_Decrypted share1 H. Image1_Decrypted share2

      1. Revealed Image1from decrypted shares

         Fig. 4.1 Experiment-1

         1. Binary Input Image2

         2. Image2_share1 C. Image2_share2

1. Image2_Encrypted share2 E. Image2_Encrypted share2

   1. Revealed Image from Encrypted shares

   2. Image2_Decrypted share1 H. Image2_Decrypted share2

      1. Revealed Image2 from Decrypted shares

         Fig. 4.2 Experiment-2

         1. Binary Input Image3

         2. Image3_share1 C. Image3_share2

1. Image3_Encrypted share2 E. Image3_Encrypted share2

   1. Revealed Image from Encrypted shares

   2. Image3_Decrypted share1 H. Image3_Decrypted share2

      These experiments have been conducted taking secret images of different sizes as an input shown by A in Fig.4.1, Fig.4.2 & Fig.4.3, B & C sho w share1 & share2 of the secret image generated by the Visual Cryptographic encryption phase.

      D & E show the encrypted share1 & encrypted share2. These are the results of second phase, in which the Visual Cryptographic shares have been encrypted using encryption algorithm. G & H show th decrypted share1 & decrypted share2, the results of decryption phase using RSA. I shows the original secret image revealed by overlapping the decrypted share1 and decrypted share2. The Visual Cryptographic decryption is used to retrieve this secret image.

      Performance of the scheme has been evaluated to test whether retrieval of input images have been possible by any opponent having all the shares at the same time by stacking encrypted shares. In Fig.4.1, Fig.4.2 & Fig.4.3 F shows the result of stacking encrypted shares which prove that opponent cannot retrieve secret image without having secret key. Table I shows that system is enough efficient with the varying size of input and random selection of key.

      Table I. System Performance

      8X9	25,163202	77929 63201	Not Retrieve	Retrieved
      0X9	53,511493	0418 11493	Not Retrieve	Retrieved
      0X9	(13, 92341	56229 2341	Not Retrieve	Retrieved

      V. CONCLUSION & FUTURE SCOPE

      1. Revealed Image3 from decrypted shares

Fig. 4.3 Experiment-3

We have tested this scheme on different types of input images with change in size of the image and keys of RSA. But the entire time secret image is retrieved with good visual quality. The confidentiality of shares is also tested by super imposing the encrypted shares before reaching to the destination. In all the cases it has been observed that if any intruder will be successful to get the encrypted shares from network, he or she cannot retrieve the original secret image without availability of private key.

It has been observed that there are many possible enhancements and extensions exist as the visual quality & size of revealed image. The major areas of future scope are:

We can use colour image in place of binary image and then generate the shares using Visual Cryptography.

Compression of encrypted shares to reduce bandwidth requirement

More sophisticated public key encryption to reduce key size. Size of image

Variations in format of Input image

REFERENCES

1. A. Parakh and S.kak A Recursive Threshold Visual Cryptography Scheme. Department of Computer Science, Oklahoma State University Stillwater, OK 74078.

2. B. Padhmavati, P. Nirmal Kumar, M. A. Dorai Rangaswamy A Novel Scheme for Mutual Authentication and Cheating Prevention in Visual Cryptography Using Image Processing. Department of Computer Science & Engineering, Easwari Engineering College, Chennai, DOI: 02, ACS.2010.01.264, 2010 ACEEE.

3. Chandramathi S., Ramesh Kumar R., Suresh R. and Harish S. An overview of visual cryptography International Journal of Computational Intelligence Techniques, ISSN: 09760466 & E-ISSN: 09760474Volume 1, Issue 1, 2010, PP-32-37

.

1. Néelima. Guntupalli et al, An Introduction to Different Types of Visual Cryptography Schemes, International Journal of Science and Advanced Technology (ISSN 2221-8386), Volume 1 No 7 September 2011, PP 198- 205.

2. M. Naor and A. Shamir Visual Cryptography. Advances in Cryptology EUROCRYPT 94. Lecture Notes in Computer Science, (950):112, 1995.

3. M.Nakajima and Y.Yamaguchi Extended Visual Cryptography for Natural Images. Department of Graphics and Computer Sciences, Graduate School of Arts and Sciences, the University of Tokyo 153-8902, Japan.

4. P. S. Revenkar, Anisa Anjum, W. Z. Gandhare Survey of Visual Cryptographic Schemes. International Journal of Security and Its Applications Vol. 4, No. 2, April, 2010.

5. Shyamalendu Kandar & Arnab Maiti K-N Secret Sharing Visual Cryptography Scheme For Color Image Using Random Number. International Journal of Engineering Science and Technology (IJEST),

   ISSN 0975-5462, Vol. 3 No. 3 Mar 201,PP 1851-1857

6. Ujjwal Chakraborty et al, Design and Implementation of a (2, 2) and a (2,

   1. Visual Cryptographic Scheme International Conference [ACCTA- 2010], Vol.1 Issue 2, 3, 4, PP 128-134

7. Vaibhav Choudhary An Improved Pixel Sieve Method for Visual Cryptography International Journal of Computer Applications, (0975 8887) Volume 12 No.9, January 2011.

8. Wei-Qi Yan, Duo Jin, Mohan S Kankanhalli Visual Cryptography for print and scan applications School of Computing, National University of Singapore, Singapore 117543

9. Y. Bani, Dr. B. Majhi and R. S. Mangrulkar, 2008. A Novel Approach for Visual Cryptography Using a Watermarking Technique. In Proceedings of 2nd National Conference, IndiaCom 2008.

10. Behrouz A. Forouzon, Cryptography & Network Security 4th Edition.

11. http://gdp.globus.org/gt4-tutorial/multiplehtml/ch09s03.html Last accessed on 28.07.12