A Filtering Based Approach to Adaptive Steganography N Santosh Arjun, IEEE Student Member, Dept of Electronics and Communication engg, Osmania University College of Engineering Hyderabad, India – 500 007.
[email protected] Abstract- An ideal steganographic technique embeds message information into a carrier image with virtually imperceptible modification of the image. Adaptive steganography comes closer to this ideal since it exploits the natural variations in the pixel intensities of a cover image to hide the secret message. The objective of adaptive steganographic methods in the literature is to embed a message, such that the obtained steganographic images appear unmodified. These approaches are limited due to use of local image information and the lack of parametric control of embedding capacity. In this paper a new method of adaptive steganography is proposed, that uses both global and local image features. The novel approach allows for a high embedding capacity and enhanced security. A detailed algorithm is presented along with results of its application on some sample images. A comparison between the features of existing adaptive steganographic methods and the new method are also presented.
I.
INTRODUCTION
Maintaining the secrecy of digital information when being communicated over the Internet is presently a challenge. Given the amount of cheap computation power available and certain known limitations of the encryption methods it is not too difficult to launch attacks on cipher-text. A message in cipher text is almost certain to arouse suspicion on the part of an interceptor while an “invisible” message created with steganographic methods is more likely to slip through. A block diagram of a basic image steganographic system is depicted in Figure 1. The transmitter consists of a cover image C (a carrier), message M is the data for confidential transmission. M can be plain text, cipher, etc. anything that can be embedded in a bit stream. The cover image C is used to embed the message by using a stego-encoder controlled by stego–key K. K is a shared secret with the intended recipient whose knowledge of the key enables them to decode the message from the stego-image. The decoding parameters are also understood to be known as a shared secret. In this paper our main concern is in explaining the steganographic technique, how these shared secrets are communicated confidentially is outside the scope of this paper. Adaptive steganography attempts to secure greater stealth for the message by ensuring that the changes introduced into the cover image remain consistent with the natural noise model associated with cover images. The most important aspect of steganography is that the carrier medium should not attract attention. When the existence of hidden information is known, the main goal of steganography is
Atul Negi, IEEE Senior Member, Dept of Computers and Information Sciences University of Hyderabad Hyderabad, India – 500 046
[email protected] defeated [1]. This is very dangerous from a security point of view. There is a tradeoff between the invisibility (imperceptible to naked eye) and the embedding capacity (amount of information that can be hidden) of the stegoimage, for a given cover image [2]. If too much information is hidden in an object, then the invisibility becomes frail and features for steg-analysis and detection get exposed. In this paper, our approach aims to reduce the risk of detection, while keeping a high embedding capacity. In section II the basis for new method and its implementation is presented, followed by the algorithm and results. In section III a comparison between existing steganographic methods and the new method are presented. II. ADAPTIVE STEGANOGRAPHY USING FILTERING Adaptive Steganography reduces modifications to the image, and adapts the message embedding technique to the actual content and features of the image. In general, to keep a good degree of stealth ness, Adaptive methods embed message bits into certain random clusters of pixels (avoiding areas of uniform color) selecting pixels with large local standard deviation or image blocks containing a number of different colors. The main advantage of adaptive steganography is that the changes made to the cover image take into account the sensitivity of the human visual system and also various statistical parameters generally being used by steg-analysis algorithms. The main challenge posed to existing adaptive steganography techniques [3,4,5,6] is that the methods so far developed doesn’t seem to have a way to
Fig. 1. Basic steganographic model.
control the amount of information that is to be hidden, for a given cover image. This problem is overcome in the method presented in this paper. The proposed approach utilizes the sensitivity of the human visual system to adaptively modify the intensities of some pixels in a high frequency components spatial image (HFSI) of the cover image. The modification of pixel intensities depends on the magnitude of the pixels in HFSI and also on the local features of the cover image. If the contrast of the image is large (e.g., an edge), the intensities can be changed greatly without introducing any distortion to human eyes. On the other hand, if the contrast is small (e.g., a smooth), the intensities can only be tuned slightly. In this method, first the cover image is passed through a filter to separate the high and low frequency components of the image. The inverse transform of both the images is computed. Now the pixels values of HFSI are modified depending on the magnitude of the pixel i.e. more the magnitude more the Least Significant Bits (LSB’s) of that pixel are changed and also the local features of cover image are considered. Now both the LFSI (Low Frequency components spatial image of cover image) and HFSI are added to form the stego – image. At the receiver the reverse process is to be done to recover the message. A. Algorithm Step 1: Let the cover image is represented by c(x,y). It is then passed through a filter with transfer function h(x,y) to separate high and low frequency components.
F [c ( x, y )] = C ( X , Y ) Where C(X,Y) represents Fourier Transform of the cover image. In this paper capital letters representation for pixel is used for frequency domain and small letters for spatial representation.
C ( X , Y ) H ( X , Y ) = LO ( X , Y ) + HI ( X , Y ) Where LO(X,Y), HI(X,Y) represent low frequency and high frequency components of cover image respectively, obtained after passing through the filter with cut off as stated above. Step 2: Inverse transform of both the frequency components is found out, known as HFSI (High Frequency components Spatial Image) and LFSI (Low Frequency components Spatial Image) separately.
F −1[ LO ( X , Y )] = lo( x, y ) and
image. Let the message is represented as m(x,y) and the embedding function as M[ ].
mhi( x, y ) = M [hi( x, y ) + m( x, y )] Step 4: Both the modified HFSI and unmodified LFSI are added to form stego – image.
Steg ( x, y ) = mhi ( x, y ) + lo( x, y ) Step5: At the receiver LFSI is subtracted from stego – image leaving modified HFSI image.
mhi( x, y ) = steg ( x, y ) − lo( x, y ) Step 6: Now the message is decoded from the Modified HFSI image using the stego – key
m( x, y ) + hi ( x, y ) = M −1[mhi ( x, y )] B. Adaptive embedding based on pixel magnitude in HFSI image and local features of an cover image. In the HFSI image, the message is embedded using an algorithm which manipulates LSB’s according to their magnitude in the filtered image and also taking into consideration the local features of cover image. Step 1: First a 3Χ3 block is moved over the cover image centered at location (x,y). If all the pixel values under the block in c(x,y) are the same, then the central pixel of the block in hi(x,y) is not used for message embedding. All the remaining pixels of hi(x,y) are modified in the following way. Step 2: For a given pixel in hi(x,y), we find the first nonzero MSB bit, which is the Kth bit from LSB side as shown in Figure 2. (K-2) bits of that pixel are replaced by message bits and (K-1)th bit is modified such that the error between original pixel and the modified pixel value is reduced. For example, if hi(x,y) magnitude is either 2, 3, a single bit (the least significant bit) is replaced by the message bit. The Kth bit cannot be used for embedding message information since it is used at the receiver to mark the number of message bits for a given pixel in the stego-image steg(x,y). The Kth bit for a given magnitude and number of bits that can be modified are summarized in table 1. This same table is also used at the receiver for decoding the message from the stegoimage.
F −1 [ HI ( X , Y )] = hi ( x, y ) Where lo(x,y) and hi(x,y) are the spatial components of low and high frequencies in the cover image respectively. Step 3: Now message is embedded into HFSI image. The number of bits modified in a pixel is made to depend up on its magnitude and also on the local features of the cover
Fig.2. Embedding Algorithm
Table 1. Number of Bits that can be used for Hiding Information for a given Magnitude of a pixel in HFSI Image. Magnitude of Kth MSB Number of Bits that Pixel in HFSI bit which can be used for Image (M) is one. Hiding Information (k-2) bits. 0
