Novel Steganography Method for Secure Data Transmission

Novel Steganography Method for Secure Data Transmission

Krishnadeva K1, Sripa Vimukthi Bannakkotuwa1, Gayani Sandamali Wijesinghe1, Lakmal Rupasinghe1, Jenny Krishara1 ,Chamodi De Silva1, Thilini Weerasooriya1, Hasini Perera1

1 Department of Information Systems Engineering, Faculty of Computing, Sri Lanka Institute of Information Technology,

New Kandy Rd, Malabe, Sri Lanka

Abstract Steganography is the art of hiding information to prevent the detection of hidden messages. Cryptography can be defined as the conversion of plain text into a scrambled code. Both are two general ways of sending vital information in a secret way. A message in scrambled code aka cipher text, might arouse suspicion on the part of the recipient while an invisible message created with steganographic methods will not. But many steganographic systems are weak against visual and statistical attacks. Systems without these weaknesses offer only a relatively small capacity for a steganographic message. In this paper, we present a new steganographic technique to hide data inside an image. The newly planned technique withstands visual and statistical attacks, yet it still offers a large steganographic capacity. This method does not modify the image. Message to be hidden will be converted into a bit stream and divided into several blocks. The bit pattern of each block is identified inside the image and their location is stored in an index table. This index table represents the locations of data inside the image. Later this table will be encrypted and transferred along with the image.

Keywords Steganography, Cryptography, Steganalysis, Algorithm, Encryption, Decryption, Stego image, Data hiding, LSB.

1. INTRODUCTION

   Data is the raw form of information, stored in columns and rows in databases, network servers, and personal computers. This information is generally from personal files and intellectual property for market analytic which are confidential. With the rapid development of the internet and the technology, personal and sensitive information is at risk. Hence, data security has become one of the most significant factors in information communication technology and data from unauthorized disclosure and modification.

   Cryptography and Steganography are studies focused on data security. Cryptography enhances security by providing confidentiality, integrity, authentication and non-repudiation for sensitive information. Cryptography uses encryption algorithms which jumble information by using a key which is known to both parties involved in the communication. Any third party cannot interpret the message due to data encryption. Yet, there is a possibility of a third party getting suspicious that there is a message embedded inside the cover medium. Steganography plays a major role in avoiding such situations. Steganography is the mechanism used to hide the existence of the message in the covered medium [1]. The common modern technique of steganography exploits the property of the media itself to convey a message. Plaintext, still imagery, audio and video, and IP datagram can be used for digitally embedding the

   message. Digital images are the most favored because of their frequency on the Internet [2]. This research is concerned with providing a high-level security mechanism for data transmission considering images. Both cryptography and steganography are combined to achieve this purpose.

2. SIMILAR TECHNOLOGIES

1. Random Pixel Selection

   In this algorithm, data is hidden randomly. Data is hidden in some randomly selected pixel. Random pixel is generated by using Fibonacci algorithm.

   A novel data embedding method using random pixels selection, this embedded data method uses a multilevel histogram shifting technique. In the embedding process, random pixels in natural images were selected to be used in hiding data. The result showed that the embedded data was distributed in a more irregular manner and can better evade the detection of Statistical Steganalysis tool. This method obtained better Stego image quality. In comparison to another similar work, this approach provided better security while offering low distortion [9].

   Replacing Intermediate Bits Using this technique, any bit or any intermediate bit from a given byte (of a pixel value) can be substituted by the bit of data to be hidden. Bits are replaced according to any random sequence from LSB to MSB position. Raster Scan Principle Pixels from alternate horizontal lines are used for replacing the secret information. A simple LSB scheme can be used for pixels of the first horizontal line. Then the second line is skipped. Again the third line is used to hide secret information and so on.

   1. Random Scan Principle

      The sequence, in which pixels are drawn, they are used to hide secret information. Again any simple data hiding algorithm like LSB can be used to hide secret information. By this method, data can be hidden in random pixels in an image.

   2. Color Based Data Hiding

      In this scheme, one fixed color is used to hide secret data. Intensity values of this fixed color are converted into binary format, and the secret information is hidden in this binary data.

   3. Shape Based Data Hiding

      Any shape can be taken to hide the data in an image. Confidential information can be hidden only in the pixels which are available in shape, instead of hiding secret

      information in the whole image. We can use any shape having any dimensions.

2. Pixel Value Differencing Steganography

   The pixel-value differencing (PVD) scheme provides high imperceptibility to the Stego image by selecting two consecutive pixels and designs a quantization range table to determine the payload by the difference value between the consecutive pixels. Besides, it offers the advantage of conveying a large number of payloads, while still maintaining the consistency of an image characteristic after data embedding [10].

   PVD is designed in such a way that the pixel modification does not violate gray scale range interval. The selection of the range intervals is based on the characteristics of human vision sensitivity to gray value (0-255) varies from smoothness to contrast. It provides an easy way to produce a more imperceptible result than simple LSB replacement methods. The embedded secret message can be extracted from the resulting Stego image without referencing the original cover image. Moreover, to achieve secrecy protection of hidden data, a pseudo-random mechanism may be used. If secret data is stored randomly it is difficult to understand by the intruder. PVD embedding is used for edged areas to increase image quality. It is also used to hide message into grayscale as well as in color image [11].

3. Intensity Based Steganography

   In this method, all the three-color planes will be converted into binary values. For each pixel in the image, the plane which has the minimum number of ones in its MSB will act as index plane and the other two color planes are considered as data planes. Compared to method 1 and method 2 in the existing work, this approach will help us to embed number of message bits in the cover medium [13].

   RGB is the most common and simplest model. The technique is more secure. The third party cannot easily detect the presence of hidden data. One of the main advantages is its capacity because it embeds a large amount of data as compared to previous techniques. The main benefit of the proposed algorithm is if we utilize all the bytes of cover-image to hide the data bits, then the algorithm has a very high capacty of data hiding. Average number of bytes changed in image for embedding data and Average number of bits changed per pixel

4. LSB Based Steganography Method

Least Significant Bit method is the most popular as well as the easiest amongst the steganography methods. In this LSB technique, the secret message can be hidden within a cover image. The main concept of this method is to get the bit representation of the secret message and the bit pattern of the cover image separately and then replace the least significant bit (last bit) of the cover image by the bits of the message. The number of bits from the secret message that can be stored within a cover image will be differing according to the type of the image. When a 24-bit image is used as the cover image of this process, there are three components such as Red, Green, and Blue and therefore three bits from the secret message can be stored in each pixel. At the same time, if its an 8-bit cover image, only one bit can be modified. But theres a huge

difference in those two cases since the change in the 24-bit image cant be easily detected by the human while the change of the 8-bit image can be easily detected due to the color changes [16].

1. RESEARCH OBJECTIVE

   Most of the existing stenographic methods apply the Least Significant Bit (LSB) as a technique of hiding a message inside an image. In this method, the binary representation of the message overwrites the LSB of each byte in the cover image. It is important that the Stego-image does not contain any detectable artifacts. Least significant bits are the less important bits in a byte. Therefore changing these bits has no high impact in modifying the original image. It is indiscernible to the human eye. Modern steganalysis techniques are capable of detecting the embedded message by performing analysis statically or by identifying the signature of stenographic tools. View presents in this paper are to provide an advanced solution for steganography that can be used for the high reliable communication when transferring images. The primary concern of this method is to use an image for data communication instead of embedding the data inside the image. This method will not modify the image. It generates an index table to represents the locations of data inside the image.

   This research is focused on introducing an advanced steganography method for transferring images (JPEG, BMP, GIF, and PNG). An image can be illustrated as an array of numbers which represents light intensities at various points (pixels). Each pixel is derived from three primary colors: Red, Green, and Blue, and each primary color is represented using

   1 byte. 24-bit images use, 24 bits (3 bytes) per pixel to represent a color value. A message can be embedded by modifying these bits in an image.

2. SYSTEM OVERVIEW AND DESIGN

   A standard approach for hiding information in digital images include LSB insertion, masking and filtering algorithms and transformations. Each of these techniques can be applied, with varying degrees of success to different image files. Most stenographic techniques involved in changing properties of the cover source, but these changes can be detected. Statistical analysis is the widely-used method for this purpose to identify the difference between random values and real image values. Using this technique, it is also possible to detect messages hidden inside JPEG files with the DCT method, since it also involves bit modifications [3]. In contrast, the image is not modified in the proposed method, but it provides high security by standing against visual and statistical attacks. Instead of modifying the image, an index table will be generated to point out the locations of data inside the image. The encrypted image and the encrypted Index table are transmitted separately. Hence, intercepting one transmission is insufficient to access the data.

   This overall concept has been achieved using several phases. In this section, we mainly focus on 3 main phases: encryption phase, transmission phase, and decryption phase. In the Encryption phase, the inputs are given: sensitive data, secret key and career image. Delivering appropriate index table is the

   main responsibility of this phase. In Transmission phase, the data is sent to the receiver using transmission media: email, the web or portable storage. In the decryption phase, the data is retrieved using the career image and image table using the correct key.

   Finally, a parameter is added to the index table to indicate the recovery option is selected. If the image is modified during the transmission, data is retrieved by the Recovery module using both index tables.

3. METHODOLOGY

   In the proposed algorithm an image is used to provide high security to the message, but the message is not embedded inside the image. Instead, it generates an index table to represents the message inside the image. This index table contains locations of each byte in a message inside the image. Later, index table has to be transmitted along with the image to the receiver for retrieval of the message. Implementation of the proposed steganographic system consists of 3 major modules:

   • Encryption Module

   • Decryption Module

   • Recovery Module

   Cover image and the data file are the inputs of the Encryption module. It will accept any images regardless of its type. Then it converts the data file into index table which will represent the locations of each byte of the data file in the image. In the process of generating the index includes the following procedure.

   First, the availability of the inputs are verified: cover image and data file. Second, the career image is converted to a byte array and check all the possible 8-bit patterns are contained in the image. If not available, an appropriate image handler module is used to add missing patterns randomly. Image handler module is a sub module of the encryption module. The purpose of this module is to ensure the suitability of the career image. Third, the data file is converted to a byte array, and get the hash value of the data file. Then search the locations of each element in the data array, within the image byte array. Store the matching coordinates in a separate array. Furthermore, file name, file type and the hash value are added to the index table. Finally, the index table is encrypted using a proper key and compressed.

   Decryption module retrieves data using the index table and career image. Career image plays a huge role in this process. If the correct career image is not selected, the data cannot be retrieved. Hash verification is done for verification of data transmission. In this module, following operations are performed. First, the availability of all inputs are verified: career image and encrypted index table. Second, the index table is decompressed and decrypted. Then the hash value and the file name is retrieved from the index table. Simultaneously, the image is converted into a byte array and retrieve the data in the locations mentioned in index table and reconstruct the data file. Finally, the hash value is verified and generate the data file.

   In addition to the Encryption and decryption modules, recovery module is used for higher reliability. Encryption is an optional module, which is used to retrieve data when the image is corrupted during the transmission. The image is divided into two parts and generate two individual arrays of the index table. Next, a single index table is generated by combining them.

   First, the normal retrieval operation is performed. If fails, the second index table is used. In this case, a small portion of the image is used to retrieve data. Alternatively, if the portion used by the recovery module is damaged, data cannot be retrieved. The proposed method uses two new algorithms shown in figure 1, to generate the index table.

   1. Index table generating algorithm

   2. Pattern checker algorithm

   Figure 1: Indx table generating and pattern checker

   1. Index table generating algorithm

      The purpose of this algorithm is to generate index table. The initial step is to convert the message into a byte array. Then a search operation is performed to identify the same bit pattern of each element in byte array inside the image. Identified locations are stored sequentially in another array. This newly created array is converted into a string and is written to a text file.

   2. Pattern checker algorithm

   Availability of possible 8 bits patterns in the image is checked using this algorithm. It converts the image into a byte array and searches for all patterns to identify any missing patterns. Then it inserts missing patterns randomly. The Same procedure is done at the receivers end to generate same image array. This module is to ensure that the system supports all possible characters in the message.

4. RESULTS AND DISCUSSION

   Based on the testing results, the functionality and the performance of the proposed method are identified. To evaluate the system performance on encryption and decryption, the duration was recorded respect to the various file sizes. A comparison was carried out using the existing steganography tools and the results are shown in Table 1.

   Table 1: Test results in comparison with existing Steganography tools

   Table 2: Comparison between the LSB technique and proposed method

   Table 3: Comparison of features and facility: S-Tools and proposed method

   In other steganographic techniques, the career image size should be larger than the data file size [4], but in the proposed method the file size is independent of the image size. A comparison is done between the proposed method and the LSB technique on important characteristics. The results are shown in Table 2. The researcher also compares the proposed method with the S-tool which is considered as one of the leading Steganography tools in the market. The comparison results are shown in Table 3.

   The proposed system handles image in byte level by converting them into byte arrays with providing the capability of accepting any image. Therefore, this indicates the image size and message size are independent variables. This process increases the efficiency in transmitting a large message by using the cover image. By eliminating the image transmission, the proposed system withstands against most of the cyber- attacks. This is clearly shown in Table 4 which compares the weaknesses of both the proposed tool and the S-tool.

   Table 4: Comparison of attacks/ vulnerabilities: S-Tools and proposed method

   1. Encryption Process

      Figure 2: Test results for decryption process for difference data sizes

   2. Decryption Process

   The proposed technique is comparatively slow in encryption and fasts in decryption as shown in figure 2 and figure 1 respectively. The delay in the encryption is due to the search algorithm which generates index table. Following records shown in figure 2 and figure 1 are calculated using an image of 75KB size.

   Figure 3: Test results for decryption process for difference data sizes

5. STEGANALYSIS ADVANCEMENT Steganalysis is the discovery of the existence of hidden

   information. The goal of steganalysis is to discover hidden information and to break the security of its carriers [5]. LSB method is widely used for image steganography [6].

   In LSB method, least significant bits of the cover pixels are overwritten by the secret bit stream and steganalysis program needs to read back the LSBs of the image. Paris Analysis, a higher-order steganalysis method implemented [7], detects secret messages embedded in digital images. This method is used to estimate the length of the secret message. Other steganalysis methods [8] use different approaches to work against any steganographic embedding algorithm.

   All steganalysis methods are based on a common characteristic. They detect secret message by using the modification done to the image during the embedding process. In this proposed method it avoids the image modification which eliminates the steganalysis attack.

   There are two approaches for steganalysis, steganalysis method specific to a particular steganographic algorithm and steganalysis technique works with any steganographic embedding algorithm, even with an unknown algorithm. The proposed system generates an index table instead of embedding data into the image. Both index table and image are a must to retrieve data. Therefore, steganalysis can be eliminated by ensuring that the hacker does not receive both image and index table. Index table contains information about data and it required to be transferred. But transmission of the image is not mandatory since no modifications are done. Image transmission can be avoided by selecting the images which both parties are having in their possession. But existing methods cannot avoid transmission of the image due to the embedding of data into the image.

6. CONCLUSION

   In this paper, a new steganographic technique for images is presented. The main concern of this technique is to use the image for secure data communication instead of embedding the data inside the image. The new technique withstands visual and Statistical attacks, yet it still offers a large steganographic capacity. It provides more benefits and flexibility. Since the technique handles both image and message as byte arrays, any type of images including JPEG, BMP, GIF, PNG, etc. can be used as a cover image and this can be extended to audio, video files. The size of the image is not depending on the size of the

   message. Even for a large message, a small image can be selected which provides efficiency in transmission. Due to the feature that it does not modify the image, existing steganalysis techniques cannot be used to retrieve the message by any unauthorized person. The figure 2 and 3 shows the result of the Guillermos Chi-Square Steganography test done for the image with data and the image with no data. Its clear that it is not a difficult task to identify whether an image contains data or not with the presence of such tests. In the proposed system the data is not embedded within the image. For the successful retrieval of the message, both index table, and the image is required. If the image and the index table are transferred separately, intercepting one communication is insufficient for an attacker to get the message. Selection of image which both parties are having, will avoid the image transmission and completely eliminate steganalysis attacks because there is no way that the attacker can get the image.

7. FUTURE WORK

This project can be extended to a level such that it can be used for different cover media types such as audio and video in the future. The searching algorithm which is used to generate the index table can be improved. The security of proposed system is better than current techniques, but it can be leveled up to a certain extent by varying the carriers.

REFERENCES

1. Johnson, N.F. & Jajodia, S., Exploring Steganography: Seeing the Unseen, Computer Journal, February 1998

2. Jamil, T., Steganography: The art of hiding information is plain sight,

   IEEE Potentials, 18:01, 1999

3. Artz, D., Digital Steganography: Hiding Data within Data, IEEE Internet Computing Journal, June 2001

4. Guillermito. (September 27th 2004). A few tools to discover hidden data (1st ed.)[Online]. Available: http://www.guillermito2.net/

5. Pierre Richer, Steganalysis: Detecting hidden information with computer forensic analysis, SANS/GIAC Practical Assignment for GSEC Certification,version 1.4b

6. Andrew D. Ker, Steganalysis of LSB Matching in Grayscale Images,

   IEEE SIGNAL PROCESSING LETTERS, VOL. 12, NO. 6, JUNE 2005

7. J. Harmsen and W. Pearlman, Higher-order statistical steganalysis of palette images, in Proc. SPIE Security Watermarkng Multimedia Contents, vol. 5020, E. J. Delp III and P.W.Wong, Eds., 2003, pp. 131 142.

8. Dr. Monisha Sharma and Mrs. Swagota Bera, A REVIEW ON BLIND STILL IMAGE STEGANALYSIS TECHNIQUES USING FEATURES EXTRACTION AND PATTERN CLASSIFICATION METHOD, International Journal of Computer Science, Engineering and Information Technology (IJCSEIT), Vol.2, No.3, June 2012

9. Dipesh Agrawal1 Analysis of Random Steganography Techniques Using Random Pixels, International Journal of Modern Trends in Engineering and Researche-ISSN No.:2349-9745, Date: 2-4 July, 2015,

10. D.-C. Wu and W.-H. Tsai, A stenographic method for images by pixel- value differencing, Pattern Recognition Letters, vol. 24, no. 9-10, pp. 16131626, 2003.

11. Mehdi Hussain1, 2, Ainuddin Wahid Abdul Wahab1, Nor Badrul Anuar1, Rosli Sallepand Rafidah Md Noor Pixel Value Differencing Steganography Techniques: Analysis and Open Challenge,Malaysia 2015 International Conference on Consumer Electronics-Taiwan (ICCE- TW)

12. Avinash K. Gulve1 and Madhuri S. Joshi2,An Image Steganography Method Hiding Secret Data into Coefficients of Integer Wavelet Transform Using Pixel Value Differencing Approach,Government College of Engineering, Aurangabad, Maharashtra 431 005,

    India,Jawaharlal Nehru College of Engineering, Aurangabad,

    Maharashtra 431 005, India, 23 December 2014

13. Mandep Kaur, Surbhi Gupta, Parvinder S. Sandhu, Jagdeep Kaur, A Dynamic RGB Intensity Based Steganography Scheme, World Academy of Science, Engineering and Technology International Journal of Computer, Electrical, Automation, Control and Information Engineering Vol:4, No:7, 2010

14. Mehdi Hussain, Mureed Hussain, Pixel Intensity Based High Capacity Data Embedding Method, IEEE, 2010

15. M.Shobana , R.Manikandan, Efficient Method For Hiding Data By Pixel Intensity, International Journal of Engineering and Technology (IJET), 2013

16. Chyquitha Danuputri1, Teddy Mantoro2, Mardi Hardjianto1, Data Security Using LSB Steganography and Vigenere Chiper in an Android Environment, 2015 Fourth International Conference on Cyber Security,

    Cyber Warfare, and Digital Forensic

17. Dalila Boughaci, Abdelhafid Kemouche and Hocine Lachibi, LRIA- USTHB, Stochastic Local Search Combined with LSB Technique for Image Steganography