24-01-2013, 12:22 PM
AMBTC-Compressed Image Using Genetic Algorithm
1AMBTC-Compressed Image.pdf (Size: 661.9 KB / Downloads: 28)
Abstract
In this paper, we present an image-hiding scheme based on genetic algorithm. The secret messages are embedded into a compressed image of AMBTC. Genetic algorithm is enveloped to find the best substitution of AMBTC bitmap. The proposed scheme provides high visual quality of the stego-image. The enhanced system of the proposed scheme increases embedding capacity while retaining good quality of the stego-image. Experimental results show that the proposed scheme outperform the comparative schemes.
INTRODUCTION
As the problems of the illicit interception and illegal copying of digital media are becoming more and more serious the technique of information hiding is thus proposed to securely transmit and protect digital media without incurring the attention of unintended recipients when sending data over the Internet. The technique is different from cryptography [5]. In
contrast, information hiding is about transmitting the given secret data through a cover medium in a manner that the existence of secret data is unnoticed [2]. Although there is a trade-off between the hiding capacity and the quality of the stego-medium, these two factors are considered to be the important measurements in developing data hiding systems. Among the spatial domain image hiding methods, the least significant bit (LSB) method is the simplest one [10]. In 2002, Jo and Kim proposed a watermarking scheme [6] based on vector quantization. Several image hiding methods that embed secret data in frequency domain had been proposed. In [8], a steganography method based on the discrete wavelet transformation (DWT) was presented. In addition, a scheme that embeds information in the middlefrequency of the quantized DCT was provided in [1]. In 2002, Tsai et al. proposed an image hiding technique using block truncation coding [9]. In this scheme, the secret message is embedded into the AMBTC compressed image. In 2000, Pan et al. proposed a datahiding scheme [7] for two-color images
PROPOSED SCHEME
In the proposed scheme, secret data is mainly hidden in the bitmap of the AMBTC compressed code. The proposed scheme is based on GA to explore the best AMBTC code combinations for embedding secret bits. After obtaining an AMBTC compressed block code, the genetic- based embedding procedure is adopted to embed secret data. The secret data is a random bit stream of „0‟ and „1‟. The hiding capacity of each AMBTC compressed block in the proposed scheme is not fixed. Let RV denote the integer value of the secret data. The embedding rule is to modify the AMBTC bitmap for satisfying RV = g mod 2h, where h is the number of embedding bits per block. The embedding procedure is finished when all bits of secret data are embedded into the AMBTC compressed code. The point is how to alter the bits in the AMBTC bitmap so that g can satisfy the above equation with minimal MSE distortion constraint. The MSE distortion is defined as the mean square errors between original host image and the stego-image. To systematically find the best solution, the mechanism of GA has been developed. The flowchart of the genetic based embedding procedure is shown in Figure 1.
EXPERIMENTAL RESULT
To evaluate the performance of the proposed scheme, we have
performed several simulations of the AMBTC encoded image
data-embedding based on GA and compared the experimental
results with those two related schemes. We compared the
proposed scheme with Jo and Kim‟s watermarking scheme
since both schemes were designed for data-embedding in the
compressed domain. In both schemes, one secret data bit was
block-wise embedded in an image. In Jo and Kim‟s
watermarking scheme, four 512×512 grayscale images,
namely, Lena, Man, Couple, and Woman were used to test the
performance. The original image was partitioned into 16384
blocks whose size was 4×4 each. The same test images and
block size were used in theproposed scheme.
CONCLUSION
In the proposed scheme, GA has been developed to find the
best substitution for the AMBTC bitmaps. The stego-image is
visually indistinguishable from the AMBTC-decoded images.
According to the experimental results, the image quality of the
proposed scheme outperforms that of Jo and Kim‟s scheme.
The average PSNR of the proposed scheme is 33.46 dB while
it is 29.03 dB in Jo and Kim‟s scheme. The hiding capacity of
the enhance system is larger that of Tsai et al‟s scheme while
retaining good image quality.