An Efficient Image Steganographic Algorithm Using CMYK Color Model

Transcription

1 ISSN (Online): ISSN (Print) : International Journal of Research and Innovations in Science & Technology, SAINTGITS College of Engineering, INDIA Technical paper An Efficient Image Steganographic Algorithm Using CMYK Color Model M. Shobana Assistant Professor, Electronics and Communication Engineering, Karpagam College of Engineering divyashobana.m@gmail.com Copyright 2015 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Steganography assumes a key part in the mystery information (computerized) correspondence. The pattern of consolidating mystery picture in Cover picture is begat as Image-Image Steganography. Most presumably all pictures are part into red, green and blue layers, But here sequestered from everything reason, the spread's layers picture are spitting into Cyan, Magenta, Yellow and key (Black). On the establishment of these four shading layers, three calculations have been arranged and inspected in an effective way. In this calculation pixel intensities stand in a critical spot to choose its inserting limit. To assess the bore of the got picture its PSNR and MSE has been computed. Keywords: Network Security, Steganography, CMYK, Information hiding. 1. Introduction Data stowing away is a typical expression around a broad range of troubles in front of embeddings messages in computerized content. The word embeddings can identify with either making the substance barely observable or keeping up the mystery's survival information. The developing interest of system among open masses and the abundant attainable quality of every single advanced material has driven industry pros to pay a particular sympathy toward information security [2]. Information concealing strategy is a novel method for clandestine correspondence field. As of now, three essential advances are utilized: cryptography, watermarking, and cryptography. Information is emitted to pass on a couple data about the screen question, for example, ownership and rights in watermarking. Despite the fact that new system has been the focal point of quite a bit of advancement exploration for cryptography and Steganography methodologies are critical for fortifying mystery data security[3]. While cryptography dissipates the mystery data with the goal that mystery message not able to recognize by the aggressor, Steganography cloak the information over any spread medium so that it can't be judged. Of late, in the coliseum of system security, the private information concealing strategies have enthralled much research. Steganography contains spread, mystery, information, stage picture and key as the essential structural engineering. Spreads articles utilized can be of diverse sorts like picture/sound/video/content, or IP datagram. The expert branch of data concealing procedure is Image Steganography, which utilizes open channels for hiding the shrouded data. It covers the mystery information utilizing the spread picture which has massive deftness [4]. The areas of picture Steganography are generally ordered into two sorts: spatial and transform techniques. The spatial space, utilizes modification of the LSB for concealing the mystery information. While transform area utilizes wavelet space to cover the individual information. This calculation concentrate on computerized picture, as spread medium over spatial area[5]. The striking point of preference of picture coding systems is it can't be experienced through the stripped eye. Every point of interest of an advanced picture is termed as pixel and every pixel is determined by a number whose worth lies between 0 to 255 and it can be given by its 8-bit parallel identical [6]. As the photo component of the picture's changed LSB has no visual information so that a human eye can't distinguish the progressions of bits in the picture. Along these lines, uprooting the possibility of this inadequacy of human vision in mystery correspondence, the mystery information will be incubated in those spaces alone[7]. In Steganography approaches the LSB substitution (Spatial-area) is a standard system which has numerous steganalysis routines that are right now utilized. Another 25

2 Steganography Spatial space LSB coordinating, rolls out simply little improvements in LSB substitution strategy, however it is hard to name the shrouded data. In the blink of an eye all inclusive visually impaired identification is connected for identifying the LSB coordinating techniques. Alternate qualities are non-traceable, strength and limit of the discharged information[8]. The measure of incognito data that can be embedded in the transporter article, utilizing the stage calculations to prevent up from undesirable gatherings to watch the concealed information alludes to mental capacity[9]. Strength is the force of a stego item to deal with the alteration amid recouping and wrecking concealed data by the enemies. A decent innovation ought to be rich but then keep the implanted message when qualities are adjusted[10]. 2. Existing System In this assault the shading picture is connected as the screen object. The shading picture is isolated into red plane, green plane, blue planes. There are two channels particularly information channel and index channel. The information channel is utilized to hold the information (i.e. Message bits). In light of th e two LSB of the index channel, the information channel will be taken, if LSB is "00" then no inserting will happen l, on the off chance that it is "01" then installing of information will be in information channel2 also, in the event that it is "10" then inserting of information will be in information channel1and in the event that it is "11" implanting will be on both information planes. In this mode the information channel is using so as to get taken an installing procedure. In related works they proposed three systems for taking the above two determined channels. In the first place system they characterized red as the file channel. Green and blue go about as the information divert in which information is implanted. In the second technique, User will characterize record channel the rest to channel play as the information channel. In the third technique, the record divert is separate in a cyclic way[1]. 3. Proposed Method From the historical backdrop of picture Steganography so far we have bargain just with the greyscale and RGB (red, green, blue) type of images. But in this paper pictures are considered in the CMYK shading model. The CMYK shading model is a subtractive shading model, utilized as a part of shading printing, and is additionally used to clarify the printing system itself. CMYK alludes to the four inks utilized as a part of shading printing: cyan, red, yellow, and key (dark). The "K" in CMYK alludes the key in light of the fact that in four-shading printing, cyan, fuchsia, and yellow printing plates are mindfully keyed, or unified, with the dark's key plate. A few sources suggest that the "K" in CMYK originates from the last letter in "dark" and was chosen in light of the fact that B as of now means blue. At the point when two RGB hues are joined just as they manufacture the shades of the CMYK model, called as subtractive primaries. Green and blue hues produce cyan (C) shading, red and blue shading produce maroon (M) shading, and red and green shading produce yellow (Y) shading. Dark is blended to the model in light of the fact that it can't be created with the 3 subtractive primaries. The K (key) remains for dark shading. Figure.1.Color Composition of CMYK In this system the spread picture is splitted into four distinctive shading layers specifically cyan, magenta, yellow and dark rather than part it into three shading layer(rgb).in this technique the mystery message is additionally a picture which can be on any format. This pixels of the mystery message is further changed over into twofold values such that every pixel of eight bits. The message bits will get install in the four layers in view of three methods. The calculations are planned on the premise of pixel force and to enhance the message's irregularity implanting process. The layer which holds the message bits is termed as message layer. The shading layer which does not hold any message bits is termed as spread layer. Selection of message and cover layer is rely on the calculation. 26

3 3.1 4-bit based image Steganography In this approach, one temporary decimal value(d) is assigned for each picture element in the cover image. The decimal value is varies from 0 to 15.for example, the first pixel s d value is 0,for 2 nd pixel is 1 and so on,then for 17 th pixel s d value is again starts from 0.Using d value the message bits can be embedded in any of the four color layers. D value assigned to each Pixel Table 1: Selecting message layer and cover layer binary equivalent Pixel is splitted in to cyan Magenta Yellow key * * * * * * * * * - * * * * - * * * - * * * * * * * * * * - - * * - * * - * * * * * * - * * * * * * * * In this above table,* denotes the message layer and - denotes the cover layer. 3.2 LFSR based image Steganography In this method, 4-bit LFSR will produce random 4-bit binary value for all the pixels in the cover object.with the help of the binary value, message layer and the cover layer can be decided. LFSR is intialized with some 4-bit value after that it will generate the random numbers Figure 2 Working of LFSR The above figure specifies the generation of random numbers by the LFSR. 3.3 Color intensity based image Steganography In this approach, the pixels of the four color layers will be transmitted to its binary equivalents. For each picture element in the icon, the layer which delivers the least count of ones in its MSB is not supposed to hold message bits and the remaining color layers are viewed to be message layers. On examining the three algorithms given in the proposed method color intensity based image Steganography will lead us to raise the count of embedding bits in the cover image without disturbing its quality. By applying any of the three algorithms, message and cover layer is decided for each pixel. Each message layer will be roughly 8-bit binary value. In this layer the message is normally covered in the 1st four LSB of the 8-bit value and the number of embedding bits will be equal to the number of 1 s in the 1st four MSB of the 8-bit value. The number of one s representing the pixel intensity. If the intensity is high the embedding bits will also cause increased. 27

4 4. Algorithm bit based random image Steganography Embedding method Variables used: Covert Image (D), Carrier image(c) and Stego-image (I) 1. Convert secret image pixels to binary 2. Convert the RGB image(cover image) in to CMYK image 3. Set apart the cover image into cyan, magenta, yellow and Black layers 4. When every bit traversed in D perform the steps given under 4.1 For each bit in D, assign a three bit value of (0 15), starts from 0 increments the value by 1 until it reaches 15, if it reach again go back to 0 and repeat increment 4.2 Using the table (1) select the message layer and data layer 5. For each pixel traversed in D, do the following: 5.2 Embed E-bits in the LSB of the message layer -If all secret data are then embedded Proceed to next step Else Repeat this step 6. Save the resulting stego image as an RGB image Extraction method Variables used: Stego Image (I) and Covert Image (D) 1. Convert the RGB image(stego image) in to CMYK image 2. Stego image I is splitted into cyan, magenta, yellow and Black layers. 3. Message (Image) size is passed as the key (k) 4. Do the step 3 which is presented in the embedding side fork times to get message layer and top layer 5. For each pixel traversed in I, do the following: 5.2 Extract E-bits in the LSB of the message layer 6. Pull through the result as Secret Image (D) 4.2. LFSR based image Steganography Embedding method Variables used: Covert Image (D), Carrier image(c) as inputs, Stego image (I) with covert image embedded in it as output. 1. Convert secret image pixels into binary 2. Convert the RGB image(cover image) in to CMYK image 3. Set apart the carrier image into Cyan, Magenta, Yellow and Black planes. 4. Initialize the 4-bit LFSR 5. Generate 4-bit value of LFSR for each pixel presents in the cover image 6. On the basis of LFSR value message layer and cover layer is selected using the table 7. For each pixel traversed in D, do the following: 7.1 Consider E = number of 1 s in the MSB of message layer 7.2 Embed E-bits in the LSB of the message layer -If all secret information are then implanted Proceed to next step Else Repeat this step 8. Save the resulting stego Image as RGB image 28

5 4.2.2 Extraction method Variables used: Stego Image (I), key (k) as input and Covert image (D) as output. 1. Convert the RGB image(stego image) in to CMYK image 2. Stego image I is splitted into cyan, magenta,yellow and Black layers. 3. Message (Image) size is passed as the key (k) 4. Do the step 4,5 and 6 which is given in the embedding side for k times to get message layer and top layer 5. For each pixel traversed in I, do the following: 5.2 Extract E-bits in the LSB of the message layer 6. Save the result as Secret Image (D) 4.3. Color intensity based image Steganography Embedding method Variables used: Covert image (D), Carrier image(c) as inputs, Stego image (I) with covert image embedded in it as output. 1. The secret image pixels are converted into binary 2. Convert the RGB image(cover image) in to CMYK image 3. The carrier image is split into Red, Green and Blue 4. Consider the layer which has a minimum number of 1 in the MSB as cover layer and remaining layer as message layer 5. For each pixel traversed in D, do the following: 5.2 Embed E-bits in the LSB of the message layer -If all secret data are then embedded Go to next step Else Repeat this step 6. Save the resulting stego Image as RGB image Extraction method Variables used: Stego Image (I) as input and Covert image (D) as output 1. Convert the RGB image(stego image) in to CMYK image 2. Stego image I is splitted into cyan, magenta,yellow and Black layers. 3. The message size is given as the key (k) 4. Perform the step 4 which is given in the embedding side for k times to get message and cover layer. 5. For each pixel traversed in I, do the following: 5.2 Extract E-bits in the LSB of the message layer 6. Save the result as Secret Image (D) 5. Results and Discussion To measure the efficiency of this approach using CMYK images, the result of the three algorithms using RGB image has been compared to the result of the CMYK image. Its result (Stego image) is analyzed using PSNR and MSE. Here Cover image is 256x256 Lena and its Secret image is in the form of greyscale and its size is 165X165. As a result,stego images, has obtained and computed the PSNR and MSE value for the stego images using the following equations: (1) Where M and N are rows and columns of the matrix of the image s pixels and R denotes the maximum error occurs in the stego image. Table 2 compares the values of PSNR and MSE in the existing and proposed method. In existing method, the same algorithm is implemented in the RGB image. Figure 3 is the Cover image used for embedding (2) 29

6 message in it. Figure 4 is the Stego-image obtained using Method 1. Figure 5 is the Stego-image obtained using Method 2. Figure 6 is the Stego-image obtained using Method 3 Figure 3 Original Image Figure 4: Stego-Image(using Method 1) Figure 5 Stego-Image(using Method 2) Figure 6: Stego-Image(using Method 3) Table 2 Comparison of PSNR and MSE between Existing and Proposed Method Stego Image (Lena) Red Green Blue PSNR MSE PSNR MSE PSNR MSE Method 1 Existing Proposed Method 2 Existing Proposed Method 3 Existing Proposed Conclusion In this paper, secret message (greyscale image) is hidden in the Cover image. Initially the cover image is in the RGB format, for security purpose then it is converted to CMYK format. After embedding it is again converted to RGB image. In this case if any attackers break this image in to RGB layers they will not get any message properly. Using this algorithm we can achieve both security and high embedding capacity. References [1] Amirtharajan R, Mahalakshmi, V, Sridharan, N, Chandrasekar. M,Rayappan.J.B.:Modulation of hiding intensity by channel intensity - Stego by pixel commando. International Conference on Computing, Electronics and Electrical Technologies (ICCEET) (2012). 30

7 [2] R.Amirtharajan,Dr. R. John Bosco Balaguru.:Tri-Layer Stego for Enhanced Security A Keyless Random Approach. Internet Multimedia Services Architecture and Applications(IMSAA).December.pp.1 6(2009). [3] Fabien A.,P. petitcolas, Ross.J. Anderson, and Markus G. Kuhn.: Information Hiding A Survey. proceeding of the Institute of Electrical and Electronics Engineers (IEEE).vol. 87 n. 7.July.pp (1999). [4] Abbas Cheddad, Joan Condell, Kevin Curran, Paul Mc Kevitt.: Digital image steganography: Survey and analysis of current methods Signal Processing. 90.pp (2010). [5] W. Bender, D. Gruhl, N. Morimoto, A. Lu.:Techniques for data hiding.ibm Syst (3&4) (1996). [6] Zaidan, B.8., A.A. Zaidan, A.K. Al-Frajat and H.A. Jalab.:On the differences between hiding information and cryptography techniques:an overview. Applied Sci., 10: (2010). [7] S. Katzenbeisser, F.A.P. Petitcolas.:Information Hiding Techniques for Steganography and Digital Watermarking.Artech House. Norwood.MA(2000). [8] E.K. Chan, L.M. Chen.:Hiding data in images by simple LSB substitution, Pattern Recognition.Vol. 37.No. 3.pp (2004). [9] Gutub, A.A.A.:Pixel indicator technique for RGB image Steganography. J. Emerging Technol. Web Intell 2: 56-64(2010) [10] M. Padmaa, Y. Venkataramani and Rengarajan Amirtharajan.:Stego on 2n:l Platform for Users and Embedding.Information Technology Journal, 10: (2011). 31