STEGANOGRAPHY BASED ON DWT TRANSFORM

Rastislav Hovank, Peter Fori, Duan Levick Department of Electronics and Multimedia Telecommunications, Technical University of Koice, Park Komenskho 13, 041 20 Koice, Slovak Republic. Tel: +421-55-6338659; Fax: +421-55-6323989 E-mail: [email protected], [email protected], [email protected] Abstract Steganography is a technique to embed additional information into the cover media so as not to arouse an eavesdroppers suspicion. We propose a new method of steganography technique based on DWT transform. The proposed method has ability to hide secret message in a digital image. The secret message is embedded into the image by changing wavelet coefficients. The quality of the stego image is of the proposed method is very close to that of the original one. 1 Introduction Nowadays, daily communications of all kinds over the Internet have become incredibly popular. However, message transmissions over the Internet still have to face all kinds of security problems. While aim of cryptography is protecting the content of messages, steganography is the technique for hiding additional information in cover data. Steganography is the art of hiding information in ways that prevent the detection of hiding message. Whereas cryptographic techniques try to conceal the contents of a message, steganography tries to hide the fact that a communication even exists. In steganography, the object of communication is the hidden message and the cover data are only the means of sending it. 2 Generic embedding and extracting scheme Many different approaches and techniques for information hiding can be used. Figure 1 shown generic embedding and extracting scheme. The input to the embedding scheme is hiding message, the cover data and an optional public or secret key K. The output is stego data, also called stego object. Inputs to the generic extracting scheme are the tested data, the secret or public key, and depending on the method, the original cover data. The output is the extracted message.
Hidden message Cover data

M
Cover data Stego data Embedding Tested data

I
Extraction

Extracted message

IM K
a)

IT

~ M

Secret or public key

Secret or public key b)

Figure 1 Generic embedding and extracting scheme

Steganography techniques can be divided into various categories and in various ways. The basic and most common used partitioning of hiding techniques is the spatial domain, frequency domain and parametric domain steganography. Another way for categorization of steganography methods is based on the condition whether or not they use the original data for extraction of hiding message from tested data. Three types of steganography can be identified. Their difference is in the nature and combination of inputs and outputs: pure steganography secret key steganography public key steganography Steganography have to guarantee these requirements [1]: robustness the embedded information is said to be robust if its presence can be reliably detected after the image has been modified but not destroyed beyond recognition. undetectability embedded information is undetectable if the image with the embedded message is consistent with a model of the source from which images are drawn. perceptual transparency it is based on the properties of the human visual system or the human audio system. The embedded information is imperceptible if an average human subject is unable to distinguish between carriers that do contain hidden information and those that do not. security the embedding algorithm is said to be secure if the embedded information cannot be removed beyond reliable detection by targeted attacks based on a full knowledge of the embedding algorithm and the detector, and the knowledge of at least one carrier with hidden message. The above mentioned requirements are mutually competitive and cannot be clearly optimized at the same time. If we want to hide a large message inside an image, we cannot require at the same time absolute undetectability and large robustness. A reasonable compromise is always a necessity. On the other hand, if robustness to large distortion is an issue, the message that can be reliably hidden cannot be too long. This observation is schematically shown in the Figure 2 [2].

Capacity
Steganography Secure steganography Digital watermarking

Undetectability

Robustness

Figure 2 Conflicting requirements

Steganography is used to imperceptibly convey information by embedding it into the cover data. Steganography typically relates to covert point-to-point communication between two parties. Thus, steganographic methods are usually not robust against modification of the data, or have only limited robustness.

We can distinguish these attacks on steganography systems [3]: Stego-only attack only the stego-data are available for analysis. Known cover attack the original cove-object and stego-object are both available. Known message attack at some point, the hidden message may become known to the attacker. Analyzing the stego-object for patterns that correspond to the hidden message may be beneficial for future attacks against that system Chosen stego attack the steganography tool and stego-object are known. Chosen message attack the steganalyst generates a stego-object from some steganography tool or algorithm from a chosen message. The goal in this attack is to determine corresponding patterns in the stego-object that may point to the use of specific steganography tools or algorithms. Known stego attack - the steganography algorithm is known and both the original and stego-object are available. 3 Proposed method Embedding process contains two-dimensional discrete wavelet transformation of cover image, permutation of secret message, modification of transformation coefficients and finally inverse 2D DWT of modified coefficients. Inputs of embedding process are cover image I and secret message. Output of this process is stego image IM. Secret key contains information about choice of DWT transformation coefficients and permutation random vector. Before modification of coefficients, pseudorandom permutation of watermark is used for increasing security of embedded message.
Cover image

2D DWT

Choice of coefficients

Stego image

Secret key

Modification of coefficients

IM
2D IDWT

Secret message ...01100101011

Pseudorandom permutation

MP
...10101101100

Figure 3 Embedding process

We denote secret messages bit as m(x) and i(x) as DWT coefficient. Modified coefficient im(x) can be expressed in the form: im ( x ) = i ( x ) + if m( x ) = 1 (1) im ( x ) = i( x ) if m( x ) = 0 (2) where is a real number.

The extraction of secret message from the tested image consists in comparison of the selected DWT coefficients of the tested and cover image. ~ if it ( x ) i ( x ) 0 m(x ) = 1 (3) ~ if it ( x ) i ( x ) < 0 m(x ) = 0 (4) where it ( x ) is DWT coefficient of tested image, i ( x ) is DWT coefficient of cover ~ ~ image and m( x ) is extracted bit of the secret message M . 4 Experimental results Verification of proposed methods has been performed on gray scale image Lena by using 1kB secret binary message. Some experimental results are shown in Table 1. Watermark was embedded into DWT coefficients of horizontal detail subband on third level wavelet decomposition. PSNR of stego image is 46,2 dB.
Table 1 Experimental results Attack / Agreement of the extracted messages [%] without attack 100 Gaussian noise =0, =0.005 68,2 Median filter 3x3 91 Cropping of image 87,9 salt&pepper noise 0.025 68,9 Decreasing brightness -30 100 JPEG Q 50 99 Rotation 1.5 58,1 Increasing brightness +30 100 JPEG Q 70 100 Resize 92,5 Gama correction 0,5 82,3 JPEG2000 1,5 bit/pixel 99 Wiener filter 3x3 95,7 Stirmark 63,2

Conclusions In this paper the steganography method based on DWT transform was presented. Experimental results showed relationship between the lengths of the secret messages and the PSNR of the stego image. Moreover, difference between original cover image and stego image is visually unnoticeable and the embedded message could be extracted properly. Acknowledgements The work presented in this paper was supported by Grant of Ministry of Education and Academy of Science of Slovak republic VEGA under Grant No. 1/1057/04. References [1] LEVICK, Duan - HOVANK, Rastislav - KLENOVIOV, Zita: Technika digitlnych vodoznakov, princpy, systmy a pouitie. In: Slaboproud obzor. ro. 59, . 1 (2002), s. 1-5. [2] Fridrich, J.: Applications of data hiding in digital images. Tutorial for the ISPACS98 Conference in Melbourne, Australia, November 4-6, 1998 [3] Katzenbeisser,S. and Petitcolas,F.A.P, Information hiding techniques for Steganography and Digital Watermarking, Artec House,INC., 2000, ISBN 1-58053035-4