Comparison of Visual Cryptographic Algorithms for Quality Images Using XOR

Transcription

1 Comparison of Visual Cryptographic Algorithms for Quality Images Using XOR Sathiya K 1, Senthamilarasi K 2, Janani G 3, Akila victor 4 1,2,3 B.Tech CSE, VIT University, Vellore Assistant Professor, SCSE, VIT University, Vellore Abstract: The core objective of this paper is to increase the visual quality of the secret image which is obtained after superimposed using XOR operation in the two-half toned shares and to compare the visual cryptographic algorithms. In this paper, for encoding we are using halftone technique. We can obtain the halftone image by applying three algorithms Floyd, Jarvis, Stucki algorithm and reverse halftone image by pixel reversal technique in halftone image. On applying XOR operation to shares, the method is expected to give the good quality of the decoded image. Keywords: halftone, Floyd, Jarvis, Stucki, pixel reversal. 1. INTRODUCTION In cryptography, Encryption is the technique which converts the data into a cipher text that cannot be easily understood by illegal people. Decryption is the method of converting encrypted data into its original. Visual cryptography (VC) is used to hide secrets in images and decrypted by the human visual scheme if the key image matches. It is unfeasible to recover that secret from one of those shares. We can use the techniques for any number of shares. Here we implement it for two shares. It can be used in number of applications like Remote Electronic Voting, Bio metric authentication, Banking Customer Identification, Message Concealment etc value of p. Now consider the superposition of the two shares as shown in the last column of Fig. 1. If a pixel is white, the superposition of the two shares always shows one black and one white sub pixel. If p is black, it yields two black sub pixels. There is a contrast loss in the decoded image. To reduce the contrast loss we applied XOR. For example the input image fig2 (a) converted into halftone image fig2 (c) and reverse halftone image fig2 (d). Secret image pixels fig2 (b) are embedded in share1 fig2 (e) and share2 fig2 (f). When we overlap the shares we get the output as shown in fig2 (g). Figure 1 VC Principle For exemplifying VC, [2] we consider the two out-of-two visual scheme. Here each pixel p of the Secret Image (SI) is encoded into a pair of sub pixels in each of the two shares. If pixel p is white, one of the two rows under the white pixel in Fig1 is selected. If pixel p is black, one of the two rows under the black pixel is selected. The selection is performed by each row has 1/2 probability to be chosen. So, in the selected row the first two pairs of sub pixels are assigned to share 1 and share 2, respectively. An individual share gives no clue as to the Figure 2 Visual Cryptography Examples Volume 2, Issue 2 March April 2013 Page 273

2 2. Background Figure 7 Pixel Reversal Halftone image and secret information pixels will be in share1 as well as reverse halftone image and secret information pixels will be in share2 as shown in fig3 which is proposed method. Figure 3 Proposed Method On applying error diffusion technique we obtained halftone image. In Error diffusion, error is distributed to neighbouring pixels which have not yet been processed. For selecting the pixel position we are using void and cluster algorithm. [1] For encoding, we are converting the secret pixel p into 2 x 2 matrixes. Each of the two shares, only two pixels represents the secret information pixels. Fig 8 is explained as the position of the secret information pixel (A, B) should be same in both shares. If the pixel value is zero (white), a matrix M is randomly selected from the matrix collection R 1. Figure 4 Floyd s Error Diffusion If the pixel position is one (black), a matrix M is randomly selected from the matrix collection R 2. It can also be found that if p is white, either pixel A is white in two shares or B is white in two shares. Otherwise p is black. Figure 5 Jarvis Error Diffusion Figure 6 Stucki Error Diffusion To get the reverse halftone image, we are applying the pixel reversal in the halftone image. Pixel reversal is a technique which converts the white pixels to black and vice versa. (0 becomes 1 and 1 becomes 0). => Superimpose two halftone cells (a), (b) from R 1 if p=0 or (c), (d) from R 2 if p=1. Figure 8 Pixel Selection Volume 2, Issue 2 March April 2013 Page 274

3 If we add those two shares, we can decode the secret image. This is the key for decryption. The decoded image is noticeably identified, even though some contrast loss is perceived. When we superimpose the shares with XOR we get the secret image in better visual quality. 3. Results Richa Fruit 3.3Visual Quality Table 3 (VISUAL QUALITY OF DECODE IMAGE) Einstein Text Logo 3.1 Histogram Error Table 1 (SECRET IMAGE VS FINAL IMAGE) 3.4 Time Efficiency Table 4 (TIME EFFICIENCY) 3.2 Error Diffusion Table 2 (INPUT IMAGE VS HALFTONE IMAGE) Volume 2, Issue 2 March April 2013 Page 275

4 (c) Halftone of Red channel (d) Halftone of Green channel (e) Halftone of Blue channel (f) share1 Figure 9 Time efficiency 4. Conclusion Here, we utilized the error diffusion and pixel reversal method. On applying the error diffusion we get the halftone image and reversing the pixels we get the reverse halftone image. Superimposing the shares we get the secret image, for better quality we are using the XOR operation. In terms of histogram error, among three algorithms give the same result but within the algorithm XOR operation will give the better result. Stucki algorithm gives the better visual quality than other algorithms. While comparing the algorithms, Floyd is little faster and the output of Stucki likely to be clean and sharp. Jarvis algorithm is good at error diffusion so it is slower than Floyd. Each algorithm is best for particular parameter. Based on our requirement, we use specific algorithm. Being a type of secret sharing scheme, visual cryptography can be used in number of applications including access control. Decryption algorithm is not required (Use a human Visual System). So a person who is new to cryptography can also decrypt the message. 5. Extended Work We are applying the same concept for colour images (bmp). Still, we are trying to improve the visual quality. (a) Input image (b) Secret image (g) share2 (h) share3 (i) Halftone image (j) Output image Figure 10 Visual Cryptography for colour images References [1] Zhi Zhou, Member, IEEE, Gonzalo R. Arce, Fellow, IEEE, and Giovanni Di Crescenzo, Halftone Visual Cryptography,vol.15, no.8, August [2] Sangeetha Devi, E... Enhanced visual secret sharing scheme via half toning technique,2012 INTERNATIONAL CONFERENCE ON COMMUNICATION CONTROL AND COMPUTING TECHNOLOGIES, [3] L. A. MacPherson, Grey Level Visual Cryptography for General Access Structures, M.S. thesis, Univ.Waterloo, ON, Canada, [4] C. Blundo, A. De Santis, and M. Naor, Visual cryptography for grey level images, Inf. Process. Lett. vol. 75, pp , Volume 2, Issue 2 March April 2013 Page 276

5 [5] G.Ateniese, C. Blundo, A. De Santis, and D. R. Stinson, Extended capabilities for visual cryptography, Theoret. Comput. Sci., vol. 250, no. 1 2, pp , [6] M. Naor and A. Shamir, Visual cryptography, Adv. Cryptol.: EUROCRYPT,Lecture Notes Comput. Sci., vol. 950, pp. 1 12, [7] William Stallings cryptography and network security, 4th Edition. AUTHOR Ms. Sathiya K is currently pursuing her B.Tech degree in Computer Science and Software Engineering. Ms. Senthamilarasi K is currently pursuing her B.Tech degree in Computer Science and Database management. Ms. Janani G is currently pursuing her B.Tech degree in Computer Science and Software Engineering. Ms. Akila victor received her MSc software Engineering Degree in 2008 from Anna University Chennai and is also a university Rank holder and M.E degree in Computer Science and Engineering in 2010 from Anna University Tirunelveli, India. Presently working as an Asst.Prof in VIT University, Vellore. Her areas of interest are Image processing, Cryptography and Software Engineering. She has presented papers in national and international conferences in various fields. Volume 2, Issue 2 March April 2013 Page 277