The electronic era of today rely on almost all sorts of online communication to a great extent which indeed and in turn encounters many menaces and malware attacks. The solution in disguise came in the forms of cryptography and steganography. Although, these have a long history, their updated and digital forms, derivatives, is implemented to ensure secure communication. Here proposes a yet another study in steganography which pleases the intention for which it is shaped. The study puts forward three techniques Coding method, Diagonal traversing and Random pixel traversing, respectively. The underlying idea in these methods is that confidential information is embedded in the cover based on the key generated. The performance is evaluated by means of several image metrics. The routine augments the enigmatic effect of the data camouflage, making it intricate for any invader to haul out the clandestine data.
PDF Abstract XML References Citation
How to cite this article
The modern era digital communication bloomed with the evolution of internet. As it is nothing but the interconnection of computer networks globally, it employs state-of-the-art modern day optical and wireless technologies. This has paved the way for extensive researches in the field of information security to mainly detect and correct datum sabotage (Cheddad et al., 2010; Salem et al., 2011; Schneier, 2007; Stefan and Fabin, 2000). The first and foremost functions of information security are information protection, control access and administer users. Some threats for the aforementioned include attacks erudition, rapid exposure to weaknesses, disseminated attacks and intricacies of patching. Information security makes certain the veracity, discretion, ease of use and off the record facts. The main goal here is to cater to the quandaries in the administrative, technical and physical fields in secure applications.
Cryptography concept dates back to 2000 BC, through hieroglyphics- an Egyptian practice. In modern world, cryptography has become a combat zone of top computer scientists and mathematicians (Schneier, 2007; Zaidan et al., 2010). Because, today, the decisive issue in business, online communication, war etc is the capability to safely hoard and transmit perceptive data. Cryptography is a significant classification of security system. It is characterized by plain text (original text), encryption (encoding), cipher text (modified text), decryption (decoding), key (tool with which plaintext is turned to cipher text). According to the keys used, cryptography can be classified as public key cryptography and private key cryptography (Rajagopalan et al., 2012; Salem et al., 2011; Schneier, 2007; Zaidan et al., 2010). The two ciphers used in this mechanism are block ciphers and stream ciphers. In former, the operation is done on blocks of ciphers while in the latter operation is done bit by bit. The time taken for encryption and decryption is the disadvantage of cryptography. The effective solution to this problem will be steganography (Al-Azawi and Fadhil, 2010; Al-Frajat et al., 2010; Xiang et al., 2011; Zanganeh and Ibrahim, 2011; Zhao and Luo, 2012; Zhu et al., 2011) and watermarking (Abdulfetah et al., 2010; Zeki et al., 2011) and its counter attack explained by Qin et al. (2010) called steganalysis.
Steganography- We cant say that this sounds alien. It has been in use since very ancient times; term coined from Greek and is nothing but secret message in disguise. putting it simple, hidden writing. Now it is used in digitalized version. So, what exactly does it mean? The phenomenon by which one digit file is hidden or embedded in other (Amirtharajan and Rayappan, 2012a-d; Amirtharajan et al., 2012; Bender et al., 1996; Cheddad et al., 2010; Janakiraman et al., 2012a, b; Rajagopalan et al., 2012; Thenmozhi et al., 2012). The files can be text (Al-Azawi and Fadhil, 2010; Xiang et al., 2011), video(Al-Frajat et al., 2010), audio (Zhu et al., 2011) and image (Amirtharajan and Rayappan, 2012a-d; Amirtharajan et al., 2012; Gutub, 2010; Hmood et al., 2010a, b; Janakiraman et al., 2012a; Padmaa et al., 2011; Thanikaiselvan et al., 2011; Thenmozhi et al., 2012; Zanganeh and Ibrahim, 2011). As it does so, it is mainly attributed to images.
The main modules of steganography are:
|•||Cover image (in which desired message is hidden)|
|•||Secret message/image (confidential information)|
|•||Key (tool used to do embed/retrieve the secret known only to the sender and receiver)|
|•||Finally steganographic algorithm (the procedure or method of Steganography)|
So, what is the unique feature of steganography or why it is needed? Of many, important uses are to combat fraud detection, authentication, traitor tracing, copy control, data integrity and many more (Stefan and Fabin, 2000; Zaidan et al., 2010). What a stego image should thrive for? The answer will be robustness, capacity and imperceptibility. That is more of secret information should be hidden and the resultant should survive the attacks and hence, of course, should be free of human artifacts (Luo et al., 2011; Mohammad et al., 2011).
Watermarking is a vital sub discipline of information hiding (Abdulfetah et al., 2010; Stefan and Fabin, 2000; Zeki et al., 2011). Mostly, it takes audio or image files as carrier and the secret data to be conveyed is termed as watermark. The main intention of watermarking is preservation of copyright and the substantiation is accomplished by cross correlation. The attack on watermarking is done by image processing. The mechanism does not make the grade when the payload (watermark) is swapped or removed. Watermarking is a trait of the cover, i.e., cover is imperative than message. But its choice is constrained. It can be done in Spatial domain (color separation, bit flipping) and Frequency domain (i.e., embedding in high frequencies). Unlike Steganography, here exists 1: many communications but the challenge is that an interloper cant replace or confiscate the message. In watermarking, prominence is laid on avoiding deformation of cover and every method should be as stout as possible.
Aforementioned methods have given an overview about data security through various cover objects and there is a good scope for random image steganography; hence this paper proposes three such methods to improve the randomness and imperceptibility.
Method 1: In this method, two keys are used; one for defining the number of secret bits and the other to locate the area to embed. Thus, using the given key code array, text data is embedded in an image and if the code contains binary one the secret data is embedded otherwise no embedding takes place.
Method 2: Here, the text data is entrenched in cover images pixels in random fashion. Both the size of the secret data and cover image decides the selection of those random pixels.
Method 3: Embedding follows the diagonal traversing style in this routine as in Fig. 1. That is traversing is followed from the leftmost pixel value to the rightmost one of the cover image as per the key. Figure 2 describes the proposed methodology, where the embedding adopts any of the methods 1, 2 or 3.
|Coding algorithm: Embedding algorithm:|
|Fig. 1:||Embedding in diagonal traversing path|
|Fig. 2:||Block diagram of proposed method|
|Random pixel traversing algorithm:|
|Fig. 3:||Flow chart for coding method 1|
|Fig. 4:||Flow chart for random pixel traversing method|
|Fig. 5:||Flow chart for diagonal traversing method|
Diagonal traversing method:
RESULTS AND DISCUSSION
Four gray cover images Lena, Baboon, Gandhi and Temple of size 256x256 pixels is considered for testing the performance all the implemented methods in Matlab. Figure 6a-d are used as cover images are tested for full embedding capacity and the stego images are given in Fig. 7a-d, respectively for coding method 1. The obtained MSE, PSNR, MSSIM, Relative entropy and payload readings of all the stego objects are tabulated in Table 1-3, respectively.
Results of random pixel traversing method: Figure 8a-d are used as cover images are tested for full embedding capacity and the stego images are given in Fig. 9a-d, respectively for Random pixel traversing method.
|Table 1:||MSE, PSNR, MSSIM and relative entropy values for coding method for K1 = 3 and K2 = 11010|
|Table 2:||MSE, PSNR, MSSIM and relative entropy values for random pixel traversing method: K = 4 and K2 = 11010|
|Table 3:||MSE, PSNR, MSSIM and relative entropy values for diagonal traversing method: K = 4 and K2 = 11010|
|Fig. 6(a-d):||Cover image of (a) Lena (b) Baboon (c) Temple and (d) Mahatma Gandhi by using Coding methods|
|Fig. 7(a-d):||Stego image of (a) Lena, (b) Baboon, (c) Temple and (d) Mahatma Gandhi, by using Coding methods|
|Fig. 8:(a-d):||Cover image of (a) Lena, (b) Baboon, (c) Temple and (d) Mahatma Gandhi, by using Random pixel traversing methods|
|Fig. 9(a-d):||Stego image of (a) Lena, (b) Baboon, (c) Temple and (d) Mahatma Gandhi, by using Random pixel traversing methods|
|Fig. 10(a-d):||Cover image of (a) Lena, (b) Baboon, (c) Temple and (d) Mahatma Gandhi, by using Diagonal traversing method|
|Fig. 11(a-d):||Stego image of (a) Lena, (b) Baboon, (c) Temple and (d) Mahatma Gandhi by using Diagonal traversing method|
To evaluate the performance of the proposed methods MSE, PSNR, MSSIM, Relative Entropy and capacity of the stego images is calculated using the following formulae. Figure 10a-c and d are used as cover images are tested for full embedding capacity and the stego images are given in Fig. 11a-c and d, respectively for Diagonal traversing method.
Mean square error: The average squared difference between a original image and resultant (stego) image is called Mean Squared Error (MSE). It dampens small variation between the two pixels but reprimands large ones:
where, oij represents the pixels in the original image and sij represents the pixels of the stego-image.
Peak signal-to-noise ratio: The higher the Peak Signal-to-Noise Ratio (PSNR), the nearer the stego image is to the cover. A higher PSNR value associates to a high quality image:
Mean structural similarity index: To interpret the overall image quality we use Mean Structural Similarity Index (MSSIM) which is expressed by:
where, C1 = (K1L)2 L = 255.
where, μO is the estimate of the mean intensity of the cover (N = 255), σ0 is the standard deviation
Here σOS is correlation coefficient.
Relative entropy: The parameter that concerns about stego systems security. when P(el), P(e2), ......., P(em) exhibits a particular intensity probabilities. The entropy of an image is expressed by:
If Pc denotes probability distribution of the cover image and ps denotes probability distribution of the stego image. Then the relative entropy is expressed by:
Steganography is the science of hiding information. Complexity of a stego-image should be higher and it can be improved by embedding the pixel in disorder way; since it requires entropy as a paramount parameter, randomization plays a vital role in steganography. Imperceptibility and embedding capacity are improved with the help of three methods namely random pixel traversing method, diagonal traversing method and coding algorithm. In this paper the observed value of MSE, PSNR, MMSIM, ENTROPY are outstanding than the existing methods and also provides a surpassing security. It offers elevated capacity while retaining a fine stego image eminence. Even if someone looks into the indiscrimination, envisioning for each pixel causes nightmare to the prowler. Thus, this paper offers everything what a steganographic scheme should take.
- Thanikaiselvan, V., S. Kumar, N. Neelima and R. Amirtharajan, 2011. Data battle on the digital field between horse cavalry and interlopers. J. Theor. Applied Inform. Technol., 29: 85-91.
- Zeki, A.M., A.A. Manaf and S.S. Mahmod, 2011. High watermarking capacity based on spatial domain technique. Inform. Technol. J., 10: 1367-1373.