Keywords- Stenography, spatial domain techniques, secret information, Least Significant Bit, cover image

Transcription

1 DATA HIDING IN DIGITAL IMAGES USING STAGNOGRAPHY PARADIGM: STATE OF THE ART 1 PRIYANKA MATHUR, 2 SHAMBHU ADHIKARI 1 Assistant Professor, 2 Student (BTech),School of Engineering and Technology, Ansal University,Gurgaon,India 1 priyankamathur@ansaluniversity.edu.in, 2 shambhu2049@gmail.com Abstract- With the ever increasing pace of communication over internet and digital communication, there is great demand for secure transfer of information and data. The important data is always vulnerable to the attacks of hackers worldwide. In the present scenario stenography plays a vital role. It is techniques by which secret information can be hidden in any image or audio or video file which is being transferred over the communication channel. Least Significant Bit Steganography techniques are the commonly used data hiding techniques using images as cover image. This paper presents a state of the art of the data hiding techniques in grayscale image using LSB and its modified versions. Further a comparative analysis of the same is presented. Keywords- Stenography, spatial domain techniques, secret information, Least Significant Bit, cover image I. INTRODUCTION Intruders are successful in hacking information because of the form that data can be read and comprehend. Steganography is a system of hiding information in e-media. The word steganography in Greek means covered writing (Greek words stegos meaning cover and grafia means writing )[1]. Steganography become more important as more people join the cyberspace revolution. Steganalysis is the art and science of detecting messages hidden using steganography[2]; this is analogous to cryptanalysis applied to cryptography. In comparison to cryptography, it does not prohibit others from gaining the hidden information but it is to keep others from speculating that information exists. Steganography is the skill of concealing data in a way that prevents the recognition of hidden messages. The steganography consists of a cover image, message to be hidden in any form and finally the stego or final image. A cover image is the image which is used as a hide the message to be sent and a stego image is the image which is formed by the combination of normal image and secret message[3].cover image may include image, audio or video in which we can embed audio, text,image or any video as hidden data. Image Steganography is classified into two categories; spatial domain and transform domain [4] as shown in fig.1. According to spatial domain, the secret data or secret message is directly embedded into the LSBs of image pixels. One of the most known examples of spatial domain method is LSB (Least Significant Bit) insertion [4].while the schemes of transform domain embed the secret data within the cover image that transformed, such as the Fourier transform, discrete cosine transform, or the wavelet transform. Since most images have compressed by manipulating transform domain coefficients, the transform domain techniques add a fair amount of strength against the destruction of the secret data due to lossy image compression [5]. The work in this paper concerns about the spatial domain. Fig.1. Types of stagnography techniques. We have done a comparison of simple LBB and pseudorandom LSB in this paper. In simple LSB, secret information is embedded in the least significant bits representing pixels or samples of the cover file[6]. In psudo random LSB the pixel where secret data is to be inserted is picked up randomly.also we have proposed adding the bit arbitrarily in 1st or 2 nd LSB to increase the security. Using pseudorandom LSB without changing the quality of image security is enhanced. The rest of the paper is organized a follows: Section II deals with related work of researchers. In section III architecture of LSB and pseudorandom is explained in detail. Section IV discusses the results and discussion followed by conclusion and future work in section V. 98

2 II. PROLOGUE ASSESSMENT Many efficient image steganography algorithms have been proposed till date but still there is a lot of space for research in this vast and growing field. Arvind Kumar and Km. Pooja [6] presented paper on Image steganography-a Data Hiding Technique and discussed how digital images can be used as a carrier to hide messages. A comprehensive survey of the most recent LSB Steganographic techniques is presented by Priya Thomson in [26]. Secret data is embedded in the least significant bits GIF images which is palette based format and stores image colors in alpha channel or a lookup table [9]. The data embedding capacity of GIF images is less and is vulnerable to attacks. Amitava Nag et al [3] presented a paper on An Image Steganography Technique using X-Box Mapping. This paper represents a LSB technique in spatial domain using X-Box mapping where four X-Boxes with sixteen different values are used which are mapped to the four LSB s of the image. K. Pramitha et al [27] in his paper used MOD-4 Embedding Algorithm based on Image contrast.lsb technique is used for embedding the characters of the secret text message into image file using LSB but before embedding the entire text is chopped up and each segment are randomized at bit level.kshetrimayum Jenita Devi et. Al [24] in her work proposed choosing pixel values randomly to insert data so that security can be enhanced. Mamta et.al[21] also reviwed some of the most coomonly used LSB insertion techniques. Ekta et al. [12] Compared stagnographic techniques taking into account its features and limitations. Shamim Ahmed Laskar[13] analysed random selection of pixels for data hiding. This method modifies LSB 1 and LSB 2 of the cover image according to PRNG sequences was done by etal. Zakir Khan [15]et al. introduced bit complementation for stagnography where 4 of the 8 bits representing pixels are taken as target bits so as to replace with it the secret data bits or its complement depending on its matching decision. represents the message by six binary bits by using LSBraille method (Braille method of reading and writing for blind people) instead of using the ASCII encoding format. In this method, three bits of the message are hidden in a single pixel, and a true image is composed of three layers (Red, Green, and Blue) layer. Two bits are embedded in the Blue layer, and one bit is embedded in the green layer of the same pixelet al. Marwa M. Emam et al [16]represents the message by six binary bits by using LSBraille method (Braille method of reading and writing for blind people) instead of using the ASCII encoding format. In this method, three bits of the message are hidden in a single pixel, and a true image is composed of three layers (Red, Green, and Blue) layer. Two bits are embedded in the Blue layer, and one bit is embedded in the green layer of the same pixel. Dr. Nanhay Singh et al [18] modified LSB ALGORITHM uses 99 the pseudo-random number generator to spread the secret message over the cover in a random manner. This algorithm will be more immune to statistical attacks without affecting its efficiency significantly. III. IMAGE STEGANOGRAPHY TECHNIQUE 3.1. LSB Substitution In LSB steganography techniques generally the least significant bits of the cover image is used to hide the message. The simplest of the LSB steganography techniques is LSB replacement[12]. LSB replacement steganography changes the last bit of each of the data values to accommodate the message that needs to be hidden. For example Consider an 8-bitgrayscale bitmap image where each pixel is stored as a byte representing a gray scale value. Suppose the first eight pixels of the original image have the following gray scale values: To hide the letter B whose binary value is , we would replace the LSBs of these pixels to have the following new grayscale values: Note that, on average, only half the LSBs need to change. The difference between the cover (i.e. original) image and the stego image will be hardly noticeable to the human eye. This is the major advantage of using LSB techniques that intruder cannot identify if something is hidden in the cover image as it is not noticible by human eye. 3.2 Psedurandom LSB Substitution In this technique, a random key is used to choose the pixels randomly and embed the message. This will make the message bits more difficult to find and reduce the realization of patterns in the image. Data can be hidden in any pixel not necessarily in a sequence. This makes it almost impossible to recover the secret message without knowing the selection pattern[24]. Sender and receiver must share the same random key selection so that data can be recovered. In this process of encoding method, a random key is used to randomised the cover image and then hide the bits of a secret message into the least significant bit of the pixels within a cover image. The random-key is usually used to seed a pseudo-random number generator to select pixel locations in an image for embedding the secret message [6] In this process of extraction, the process first takes the key and then random-key. These keys take out the points of the LSB where the secret message is randomly distributed. Decoding process searches the

3 hidden bits of a secret message into the least significant bit of the pixels within a cover image using the random key[22]. In decoding algorithm the random-key must match i.e. the random-key which was used in encoding should match because the random key sets the hiding points of the message in case of encoding. Then receiver can extract the embedded messages exactly using only the stego-key. In this paper we have modified this pseudorandom technique. In addition to the random selection of pixels whose LSB has to be replaced, we have used 1 st and 2 nd LSB for insertion. It makes the system more secure as compared to previous method of 1 st LSB substitution. By changing the initial two bits of image bitplane does not make noticeable to human eye. Therefore without loosing the integrity of the cover image, secret data can be easily inserted. IV. RESULTS AND DISCUSSION We have implemented both the LSB and Pseudorandom modified LSB on different set of cover and secret data. To compare the results obtained we have applied the statistical analysis for performance evaluation. In our experimental procedure we have used different set of cover and secret image. The experimental work of the proposed method was conducted using Matlab 2012(a). We formed stego images by embedding secret image onto cover image using proposed pesdurandom modified normal LSB replacement method.and then compared the same images with normal LSB method. The difference of the stegoimage can hardly be distinguished after embedding. It is observed that human visual system(hvs) can hardly differentiate the original image and the stegoimage and also the stego-images does not generate any suspicion as shown in fig.2 and fig.3. As in present proposed method a pseudorandom number generator is used to choose random pixels in an image by permuting the pixel indices with a secret key. The concept behind the proposed method is to devise a technique that enables secure data communication between sender and receiver. Furthermore, the secret messages were also retrieved successfully without encountering any loss of data. Most importantly, the modification of the cover image is not perceptible on the stego image at all and thus arouses no suspicion to third parties. Fig.2.(a) Cover image (b) Secret image (c) Stegno image (d) Recovered secret image using Pseodo random modified lsb. Fig.3.(a) Cover image (b) Secret image (c) Stegno image (d) Recovered secret image using lsb PERFORMANCE MEASURE: PSNR AND MSE : The Mean Squared Error (MSE) and the Peak Signal to Noise Ratio (PSNR) are the two metrics used to compare the Image quality. A Peak Signal-to-Noise Ratio (PSNR) value is calculated to measures the quality of stego images. It is defined as the ratio between the maximum possible power of a signal and the power of corrupting noise that affects the fidelity of its representation[11]. This ratio measures the quality between the original and a compressed image. The higher value of PSNR represents the better quality of the compressed image. PSNR is defined via the mean squared error (MSE) for two m*n monochrome images I and K where one of the image is considered a noisy approximation of the other. It is defined as follows in eq. (1): PSNR= 10.log10 (MAX 2 /MSE) (1) The MSE represents the cumulative squared error between the stego and the original image. The distortion in the image can be measured using MSE. Lower the value of MSE, lower is the error. The MAX is the maximum possible pixel value of the images. For example, if the pixels are represented using 8 bits per sample, then the MAX value is

4 If the value is larger among all other techniques, then it will results in the better quality of image. It is defined as the average squared difference between a reference image and a distorted image. The smaller the MSE, the more efficient the image steganography technique[9]. MSE is computed pixelby pixel by adding up the squared differences of all the pixels and dividing by the total pixel counta shown in eq.(2) (2) MSE = [I(i, j) I (i, j)] Fig.4. Comparison between PSNR values in db of LSB and proposed pseudorandom modified LSB Table 1 shows the results of MSE and PSNR values obtained on set of images using proposed Pseudorandom modified LSB method. Table 2 shows the results obtained after LSB substitution on same set of images. Further fig.4 and fig.5 shows graph obtained which clearly indicated that pseudorandom modified LSB performs better than simple LSB. Thus without altering the image integrity we are able to securely transmit and hide the image data in cover image. Table1: MSE and PSNR values for LSB and Pseudorandom Modified LSB Table 2: MSE and PSNR values for LSB Fig.5. Comparison between MSE values of LSB and proposed pseudorandom modified LSB CONCLUSIONS Stegnography is a procedure employed for the security of data. The message to be protected is implanted in the cover file which can either be text, image or audio file. In steganography, LSB technique is applied in spatial domain to embed data in cover file. In this paper, we have compared the spatial domain technique The LSB Technique and Pseudo- Random Modified LSB Technique are applied on images to obtain secure stego-image which shows that PSNR of our Pseudo random modified LSB encoding is higher than PSNR of LSB encoding. The image resolution doesn't vary much when we insert the message into the image and the image is protected with the secret key. So, it is not possible to damage the data by any unauthorized person. This paper mainly enhances the security of the hidden data inserted in image deals with increasing the security of the message and increasing PSNR and reducing the distortion rate. REFERENCES [1] Anderson, R.J. & Petitcolas, F.A.P., On the limits of steganography, IEEE Journal of selected Areas in Communications, May [2] S.M. Masud Karim, Md. Saifur Rahman, Md. Ismail Hossain, A New Approach for LSB Based Image Steganography using Secret Key, Khulna University, Khulna 9208,Bangladesh,proceedings of 14th International Conference on Computer and Information Technology(ICCIT 2011)22-24 Dec [3] Amitava Nag, Saswati Ghosh, Sushanta Biswas, Debasree Sarkar, Partha Pratim Sarkar, An Image Steganography technique using X-Box Mapping,IEEE-International Conference on Advances in Engineering, Science and Management (ICAESM-2012) March 30-31,2012 [4] Shilpa gupta, geeta gujral and neha aggarwal Enhanced least significant bit algorithm for image steganography ijcem international journal of computational engineering & management, vol. 15 issue 4, july 2012 [5] Mohammad Pooyan, Ahmad Delforouzi LSBbased Audio Steganography Method Based on Lifting Wavelet Transform International Symposium on Signal Processing and Information Technology, PP ,2007 IEEE [6] Arvind Kumar, Km. Pooja, SteganographyA Data Hiding Technique, International Journal of Computer Applications, Vol.9-No.7, November

5 [7] Wai Wai Zin, LSB Based Random Byte Data International Journal of Science Technology & Embedding, International Conference on Computer Engineering Volume 1 Issue 12 June 2015 ISSN Applications, ICCA 2012, February 28th-29th, Yangon, (online): X Myanmar [18] Archana garg, Harmanjot Singh Dhaliwal Lsb Technique [8] M. Hussain, A survey of image steganography For Stegnography For Data Security: Review, techniques, International Journal of Advanced Science International Journal of Engineering Research & and Technology, vol. 54, pp , Technology (IJERT) Vol. 2 Issue 4, April [9] K.Nitin K and N. Ashish V, Comparison of Various [19] Surbhi Singhal, Rajkumar Singh Rathore, Detailed Images Steganography Techniques, International Journal Review of Image Based Steganographic Techniques of Computer Science and Management Research, Vol. 2, IJCST Vol. 6, Issue 3, July - Sept 2015 Issue 1, PP , January [20] Amrit Preet Kaur, Gagandeep Singh, Review on Various [10] Jasleen Kour and Deepankar Verma Steganography Image Steganographic Techniques, international journal Techniques A Review Paper International Journal of for advance research in engineering and technology, Emerging Research in Management &Technology ISSN: Volume 3, Issue II, Feb 2015 ISSN (Volume-3, Issue-5)May Pp [21] Mamta. Juneja, and Parvinder S. Sandhu, An Analysis of [11] Ravi K Sheth, Rashmi M. Tank Image Steganography LSB Image Steganography Techniques in Spatial Techniques International Journal of Computer Domain, International Journal of Computer Science and Engineering and Sciences (IJCES) Volume-1 Issue-2, Electronics Engineering (IJCSEE) Volume 1, Issue (2013) [12] Ekta Dagar, Sunny Dagar, International Journal of [22] Preeti, Ayushi, A survey and comparison of Emerging Engineering Research and Technology steganography techniques,proceedings of International Volume. 2, Issue 2, May 2014, PP conference on sustainability, Engineering,Information [13] Shamim Ahmed Laskar 1 and Kattamanchi Hemachandran Technology, management and the environment, DAV, International Journal of Computer Engineering and institute of manegement, Faridabad, pp ,28 nov Technology (IJCET), ISSN (Print), ISSN (Online) Volume 4, Issue 2, March April [23] Ajay Kumar Nain, Shailender Gupta, Bharat Bhushan, (2013) Rashmi Chawla, An adaptive pseudorandom stego-crypto [14] Z.Wai1, S. Than, Data Hiding Technique Depended on technique for data communication,international Journal Pseudorandom Sequences, International Journal of of Computer Networks & Communications (IJCNC) Scientific Engineering and Research (IJSER) Vol.5, No.4, July Volume 1 Issue 2, October 2013 [24] P.sanjay kumar jena, Kshetrimayum Jenita Devi, A [15] Zakir Khan, Mohsin Shah, Dr. Muhammad Naeem, Danish Sesure Image SteganographyUsing LSB Technique and Shehzad, and Toqeer Mahmood, LSB Steganography Pseudo Random Encoding Technique, MTech thesis, NIT using Bits Complementation, Int'l Conf. on Chemical Rourkela (2013). Engineering & Advanced Computational Technologies [25] P. Thomas, Literature survey on modern image (ICCEACT 2014) Nov , 2014 Pretoria (South steganographic techniques, International Journal of Africa) Engineering Research and Technology, vol. 2, pp [16] Marwa M.Emam, Abdelmgeid A. Aly, Fatma A. Omara 111, 2013., A Modified Image Steganography Method based on LSB [26] K. Pramitha, Dr. L.Padma Suresh, K.L.Shunmuganathan, Technique, International Journal of Computer Image Steganography Using MOD-4 Embedding Applications ( ) Volume 125 No.5, Algorithm based on Image Contrast, International September 2015 Conference on Signal Processing, Communication, [17] Dr. Nanhay Singh, Ayush Singhal, Mohit Singh Bora Computing and Networking Technologies (ICSCCN An Advanced Approach for Implementation of Audio 2011), pg Steganography: Modified LSB Algorithm, IJSTE - 102