26-03-2011, 11:34 AM
Presented by:
A.Karthic
[attachment=11043]
New approach in steganography by integrating different LSB algorithms and applying randomization concept to enhance Data Security
ABSTRACT:
In this paper we propose a new technique in steganography to hide the information to be hidden with digital image. Here we take BMP image to hide the data. We take most common Steganography algorithm LSB as base and made some changes in it based upon our innovative idea that is integration of first LSB algorithm and second LSB algorithm. We consider here security as main aspect. So, we also apply randomization concept for selecting the pixel in image to hide data. Random values are generated upon secret key. Hackers cannot find random numbers unless they have key with them. Detailed visual and statistical analysis of the algorithm reveals that it yields satisfactory results. When compared with other existing algorithms, it is easy to prove that the difficulty of decoding the proposed algorithm is high.
INTRODUCTION TO STEGANOGRAPHY:
Similar to cryptography, steganography provides a means of communicating secret messages. While cryptography scrambles a message so that it cannot be understood, steganography hides the very existence of the message by hiding it inside a carrier file of some type. An eavesdropper can intercept cryptographic message, but may not even know a steganography message exists. Encryption and steganography achieve the same goal via different means. Encryption encodes the data so that an unintended recipient cannot determine its intended meaning. Steganography, in contrast attempts to prevent an unintended recipient from suspecting that the data is there. Combining encryption with steganography allows for a better private communication.
The goal of steganography is to avoid drawing suspicion to the transmission of the secret message. On other hand, steganalysis is a way of detecting possible secret communication using against steganography. That is, steganalysis attempts to defeat steganography techniques. It relies on the fact that hiding information in digital media alters the carriers and introduces unusual signatures or some form of degradation that could be exploited. Thus, it is crucial that a steganography system to ascertain that the hidden messages are not detectable.
METHOD DETAILS:
A. Message & Image Manipulation
The secret message to be hidden in the cover is converted to a bit stream which allows equivalent bit manipulation of the cover, depending on the values in the bit stream of the message. The current system demands that once the bit stream is obtained, it is broken into 8-bit blocks. Similarly, the image is broken into 8-pixel blocks, with each pixel of the block corresponding to a bit in the 8-bit block.
B. Traditional Least Significant Bit Substitution
LSB substitution is the most popular method used for Steganography due to its ease of application and less perceptual impact. The current embedding process uses LSB Steganography as the basis to implement a more robust technique. The secret message is converted to a bit stream and each bit of the message is embedded into the LSB of the pixels of the image. This ensures that the pixel value changes almost by one, which does not result in a significant change in the image quality perceptually. The word almost used here is significant as probabilistically there is 50% chance the LSB to be changed is already the one desired, and, hence no change to the image is made.
First LSB algorithm
1 0 0 1 0 0 0 1
Here only the last bit of the pixel is modified to hide the data. It is implemented highly because of its simplicity and good picture quality.
Second LSB algorithm
1 0 0 1 0 0 0 1
Here last two bits are subjected to change to increase the amount of data to be hidden. Eventually the picture quality is less than our first LSB algorithm.
C.INTEGRATION OF ALGORITHMS
The improvement of LSB information hiding algorithm mainly aims at the images of BMP type. Research revealed that, the reaction of human eyes to Red, Blue and Green is different. According to the brightness formula:
I = 0.3R + 0.59G + 0.11B
And the theory of the human visual cells sensitivity of colour, human eyes are most sensitive to the green, the next is to the red, and the least is to the blue [6]. Therefore, the different least bits of brightness components of the red, green, and blue of each pixel can be replaced by the hiding data. And according to the basic principle of the Least Significant Bits Information Hiding algorithm, the effect of replacing the least bit on original data is only 1, the second least bit replacement’s effect is 2, and the third’s is 4, and so on, the nth is 2n−1 . The higher the bit, the greater effect is. By taking advantage of the less eyesight relevance to the lower bits, the data to be hidden are embedded into the lower (first few least) bits of each pixel. If the least bit is changed in green component of the colour, the resulting effect on the brightness is 20 × 0.59 = 0.59. If the same change is made to red component, the effect on the brightness is 20 × 0.3 = 0.3. If the first 2 least bits of the blue component are altered, the effect on the brightness is (20 + 21) × 0.11 = 0.33. So, all
Effects are not greater than the maximum change in brightness the traditional LSB algorithm brings about: 20 × 0.59 = 0.59, which cannot be perceived by human eyes. Thus, for each 3-byte in the bitmap, 1+1+2=4 bits can be used for replacement. This means an increase in the hiding efficiency by about 17%, a great improvement by this improved LSB algorithm, compared with using the tradition LSB method. Moreover, there is no impact on the hiding result.
