Key-based scrambling for secure image communication

Transcription

1 University of Wollongong Research Online Faculty of Engineering and Information Sciences - Papers: Part A Faculty of Engineering and Information Sciences 2012 Key-based scrambling for secure image communication Prashan Premaratne University of Wollongong, prashan@uow.edu.au Malin Premaratne Monash University Publication Details P. Premaratne & M. Premaratne, "Key-based scrambling for secure image communication," in Emerging Intelligent Computing Technology and Applications, P. Gupta, D. Huang, P. Premaratne & X. Zhang, Ed. Berlin: Springer, 2012, pp Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: research-pubs@uow.edu.au

2 Key-based scrambling for secure image communication Abstract Secure image communication is becoming increasingly important due to theft and manipulation of its content. Law enforcement agents may find it increasingly difficult to stay afloat above the ill intentions of hackers. We have been able to develop an image scrambling algorithm that is very simple to implement but almost impossible to breach with a probability less than 5x This is possible due to the fact that a user may purchase or acquire rights for an intended image by specifying a 'key' that can form a sequence of numbers 10 to 100 in length. The content provider uses this sequence as a base in developing another key sequence to scramble the image and transmit it to the user through regular channels such as an attachment. Since the user is the only party apart from the provider to possess the key for descrambling, any third party will not be able to descramble it successfully as will be shown in this paper. Keywords key, image, scrambling, secure, communication Disciplines Engineering Science and Technology Studies Publication Details P. Premaratne & M. Premaratne, "Key-based scrambling for secure image communication," in Emerging Intelligent Computing Technology and Applications, P. Gupta, D. Huang, P. Premaratne & X. Zhang, Ed. Berlin: Springer, 2012, pp This book chapter is available at Research Online:

3 Key-Based Scrambling for Secure Image Communication Prashan Premaratne 1 and Malin Premaratne 2 1 School of Electrical Computer and Telecommunications Engineering, University of Wollongong, North Wollongong, NSW, Australia 2 Department of Electrical and Computer Systems Engineering at Monash University, Victoria, Australia malin@ieee.org, prashan@uow.edu.au Abstract. Secure image communication is becoming increasingly important due to theft and manipulation of its content. Law enforcement agents may find it increasingly difficult to stay afloat above the ill intentions of hackers. We have been able to develop an image scrambling algorithm that is very simple to implement but almost impossible to breach with a probability less than 5x This is possible due to the fact that a user may purchase or acquire rights for an intended image by specifying a key that can form a sequence of numbers 10 to 100 in length. The content provider uses this sequence as a base in developing another key sequence to scramble the image and transmit it to the user through regular channels such as an attachment. Since the user is the only party apart from the provider to possess the key for descrambling, any third party will not be able to descramble it successfully as will be shown in this paper. Keywords: Image scrambling, image communication, image shuffling, key generation. 1 Introduction Digital images are increasingly sent over networks as documents, commercial items or law enforcement material. Due to the heightened activities of hackers all over the world, these images can easily end up in the hands of unscrupulous third parties who might profit/extort or modify them without the knowledge of the legitimate receiver. To safeguard the image information, research has been carried out in mathematics, cryptology and in information theory over the years. Previously, image watermarking, visual cryptology, information sharing and image scrambling has been proposed to counter image theft. Image scrambling process is an important image encryption method which has been used in watermarking for data hiding. The objective of image scrambling has been to generate a non-intelligible image which prevents human visual system or computer vision system from understanding the true content. An authorized user is empowered to descramble the image using information regarding scrambling method and the variables in order to decipher the image. Image scrambling has been proposed as a way to mitigate such issues way back in 1960 when the first documented system to do so emerged [1]. Their approach D.-S. Huang et al. (Eds.): ICIC 2012, CCIS 304, pp , Springer-Verlag Berlin Heidelberg 2012

4 260 P. Premaratne and M. Premaratne involved scrambling, concealing or encoding information and unscrambling and decoding the received images using line screens and grids consisting of opaque and transparent lines. Over the years, image scrambling has evolved into two streams; one based on matrix transformation to shift coordinates and another to permuting coordinates of pixels. Most of the scrambling approaches are based on Arnold Transform or combination of Arnold Transform with other techniques [2-4]. These are also applicable only to equilateral images. If images are not equilateral, then they have to be padded with values to make them equilateral [5]. Since most of these techniques do not use a key that provides additional security, Zhou et.al. proposed Fibonacci P-code based scrambling algorithm which required two parameters to be known by the receiver side to descramble the images [6]. Even though, this is certainly a favorable development over the others, two numbers would not provide adequate protection and the system is very vulnerable to attack. Others [7-12] have attempted scrambling using random sequences based on chaos or pseudo random number generation based on parameters. Zhou et. al proposed an algorithm using an M-sequence to shuffle image coordinates using two parameter key [13]. The M-sequences is a maximum length sequence that has been used in spread spectrum communications. It is a pseudo random noise sequence. In this approach, the authorized user is given the shift parameter r and the distance parameter p which are used to generate the 2-D M-sequence to descramble the scrambled image. Gu, et. al. presented an image scrambling algorithm based on chaotic sequences [14]. The chaotic sequence was generated using three parameters and the algorithm typically had to be iterated 100 times to generate the non linear sequence. This introduced high complexity and the resulting scrambled image histogram was modified in the process. Even though these attempts are promising, having one or two parameters controlling the entire pseudorandom sequence generation was very vulnerable to attack. 2 Key Based Scrambling In image scrambling and descrambling, it is imperative to have simple algorithm to shuffle the pixel values fast and reorder it to reveal the original. However, such simple requirements have most of the time resulted in low-secure solutions. Our approach proposes a solution which provides both simplicity and utmost security using a user defined sequence to safeguard the content. In our approach, we build a pseudorandom sequence using the user defined 10 to 100 long positive integer value sequence. Since these values are used for image pixel row and column shuffling, the values usually have a lower and an upper limit of 1 and 200. These limits restrict that a small image such as 256x128 will not be shuffled by a value such as 300 in which case, the modulo operations simply will result in switching row value by 44 (300 = ) rather than 300. This provided sequence is used to generate a longer sequence that is as long as the maximum dimension of the image (length or width). We have few options to generate the longer sequence from the user provided sequence. One option would be to periodically insert the user provided

5 Key-Based Scrambling for Secure Image Communication 261 sequence in order to generate the longer sequence. In order to shuffle the image well, we can increment each value of the shorter sequence whenever it is repeated in the longer sequence. This process is illustrated in Fig. 1. User defined short sequence Generated Shuffling Key Sequence Increment all values Increment all values Increment all values Fig. 1. Key Sequence Generation Using User Provided Key Once the shuffling order sequence is generated, which is the key in this process, sequence values are read and the rows are switched (if the first value of the sequence is 78, then row 78 of the image is copied into row 1 and row 1 is copied into row 78). Once all rows are switched according to the key sequence, columns are switched using the same sequence. At this stage the amount of scrambling achieved is visually not acceptable as indicated by Figures 2(b). Thus this process is now followed by circular shifting of rows and then the columns using the key sequence. The result is Fig. 2. (a) Original Image of Lena (b) Scrambled Image Using Row-Wise and Column Wise Shuffling. (c) Final Scrambled Image: Result of the Row-Wise Followed by Column Wise Circular Shifting of Image (b). (d) Descrambled Image of (c) Using the Correct Sequence.

6 262 P. Premaratne and M. Premaratne visually acceptable and is shown in Fig. 2(c). Table 1 lists the image scrambling process as discussed above. The generated key sequence can again be used by the authorized party to unscramble image by undoing the circular shifting and shuffling. The unscrambled image of Fig.2(c) is shown in Fig. 2(d). 3 Robustness of Our Approach Since the user defined sequence which can be from 10 to 100 values long positive integers in the range of 1 to 200, statistically, the probability of estimating such a sequence will be (1/200) 100 = 5x Hence we can conclude that estimating such a sequence to unscramble the image will be practically impossible. Due to this robustness, the user can expect multiple images of different sizes scrambled and communicated by the same content provider. In the event of communication channel or unauthorized party adding noise to the transmitted image, the descrambling process will not be affected. Table 1. Summary of the proposed scrambling process Step1 Use the user provided key to generate a key sequence that is of the length of the maximum dimension of the image Step2 Use the key sequence to switch the rows Step3 Use the key sequence to switch the columns Step4 Use the key sequence to circular shift the rows Step5 Use the key sequence to circular shift the columns 4 Discussion We have proposed a highly secure yet simple image scrambling algorithm that seems to address most of the concerns of having a random sequence to scramble and descramble an image. Since the sequence generation is not governed by few numbers of parameter as has been reported before [7-12], a sequence that is almost 100 values long will almost be impossible to crack. Since the algorithm does not affect the pixel values of the image, its histogram and the content remain unchanged. Any additive noise will not affect the descrambling process. The algorithm equally applies to color images. Since the authorized user determines the key, same key can be used multiple times for multiple images when dealing with the same content provider. References 1. Renesse, R.L.: Hidden and Scrabbled Images- a Review. In: Conference on Optical Security and Counterfeit Deterence Techniques IV, vol. 4677, pp SPIE (2002) 2. Huang, H.: An Image Scrambling Encryption Algorithm Combined Arnold and Chaotic Transform. In: Int. Conf. China Communication, pp (2010)

7 Key-Based Scrambling for Secure Image Communication Fang, L., YuKai, W.: Restoring of the Watermarking Image in Arnold Scrambling. In: 2nd International Conference on Signal Processing Systems (ICSPS), pp (2010) 4. Liu, Z., Chen, H., Liu, T., Li, P., Xu, L., Dai, J., Liu, S.: Image Encryption by Us-ing Gyrator Transform and Arnold Transform. Journal of Electronic Imaging 20(1) (2011) 5. Kong, T., Zhang, D.: A New Anti-Arnold Transformation Algorithm. Journal of Software 15(10), (2004) 6. Zhou, Y., Joyner, V.M., Panetta, K.: Two Fibonacci P-code Based Image Scram-bling Algorithms. Image Processing: Algorithms and Systems VI 6812, , 1 12 (2008) 7. Che, S., Che, Z., Ma, B.: An Improved Image Scrabbling Algorithm. In: Proc. Second Int. Conf. Genetics and Evolutionary Computing, pp (2008) 8. Fridrich, J.: Image Encryption Based on Chaotic Maps. In: Proc. IEEE Conf. Systems, Man, and Cybernetics, pp (1997) 9. Yeo, J., Guo, C.: Efficient Hierarchical Chaotic Image Encryption Algorithm and Its VLSI Realisation. In: IEE Proceedings Vision, Image and Signal Processing, vol. 147(2), pp (2000) 10. Liping, S., Zheng, Q., Bo, L., Jun, Q., Huan, L.: Image Scrambling Algorithm Based on Random Shuffling Strategy. In: 3rd IEEE Conference on Industrial Electronics and Applications, vol. 4677, pp SPIE (2008) 11. Ville, D., Philips, W., Walle, R., Lemahieu, I.: Image Scrambling Without Band-width Expansion. IEEE Trans. Circ. Sys. Video Tech. 14(6), (2004) 12. Liu, S., Sheridan, J.T.: Optical Information Hiding by Combining Image Scrambling Techniques in Fractional Fourier Domains. In: Irish Signal and Systems Conference, pp (2001) 13. Zhou, Y., Panetta, K., Agaian, S.: An Image Scrambling Algorithm Using Parameter Based M-sequence. In: Proc. Seventh Inter. Conf. Mach. Learning and Cybern., pp (2008) 14. Gu, G., Han, G.: The Application of Chaos and DWT in Image Scrambling. In: Proceedings of the Fifth International Conference on Machine Learning and Cybernetics, pp (2006)