الفهرس 
introductionOnly 14 pages are availabe for public view
 |  | from | 119 |  |  |
| | | from | 119 | | |
Abstract
Data hiding is the art and science of hiding a secret message in a host cover/media without any perceptual distortion of the host. The composite cover is usually referred to as the stego-cover. Data hiding is also found useful as a general tool to send side information in multimedia communication for achieving additional functionalities or enhancing performance. Data compression is one of the factors that affect the process of data hiding and recovering. However the property of redundancy in cover/media could be exploited by cover compression in data hiding process. This will help to transmit the stego-cover in a smallest possible size with an efficient form and maintaining the confidentiality of the payload message.
At the moment, the exchange of information between users of network is no longer limited to text messages, but evolved to include most types of multimedia (e.g. image, video, audio). Certainly, it is very easy to explain an important topic with an image or a video without the need to write a long text. Therefore, there are a huge number of digital images available on the World Wide Web. The rapid development of information technology increases the illegal attempts to penetrate the secret data. In this respect, the digital image has been an appropriate media to handle the data hiding process.
Many robust data hiding techniques depending on the wavelet transform of digital image are available. Some reversible techniques need to merge with the wavelet compression process of images. The challenge is how to hide as many data as possible in cover-image without conflicting with data loss through compression process.
The work included in this thesis aims to develop an algorithm of data hiding. The algorithm uses the discrete wavelet transform to overcome and avoid the problem of loss of confidential data when the cover image influenced by wavelet-based compression process. The research involves studying the quality of stego-image after embedding a large amount of data, and evaluating the effect of hiding process on compression ratio considering the confidentiality of data during the process. To achieve the above mentioned aim, the thesis presents two integrated reversible data hiding methods integrated with image compression technique. In the first integrated method (EZHS), a reversible data hiding method using integer wavelet transform with embedded zero-tree wavelet for gray-scale and color images is applied. It merges
VI
lossless data hiding technique that depends on integer wavelet transform with the EZW image compression algorithm. Such method can recover the secret message again from the last pass of bit-stream of EZW robustly. This is due to shifting parts of histogram of high frequency subbands in the invertible integer-to-integer wavelet domain of EZW to make space for data hiding. The effect of embedded secret message into EZW coefficients on both of the compression ratio of coding algorithm and stego-image quality is discussed. Experimental results show high visual quality of stego-images at high payloads. The compression ratio of coding algorithm is affected by about 2% on average because of hiding process. Both of the secret message and the original cover image are reconstructed without any loss.
Following the same approach of combining traditional lossless data hiding method with embedded zero-tree wavelet compression algorithm, to compare between results of compression ratio and quality of stego-images, the second method is integrated searching for the best. The second integrated method (EZTE) merges a lossless data hiding method for digital images using integer wavelet transform and threshold embedding technique with EZW compression algorithm. Data are embedded into the least significant bit-plane (LSB) of high frequency integer wavelet coefficients smaller than a certain predefined threshold. Experimental results show high visual quality of stego-images at high payloads. The compression ratio of coding algorithm is affected by about 3% on average because of hiding process. Both of the secret message and the original cover image are reconstructed without any loss.
Final results on many gray-scale and color images show that the first method (EZHS) gives high quality of stego-images than the second one (EZTE). Through our objective test using PSNR at the same payload, the first method gains additional quality of about 8 dB on average. On the other hand, EZTE method is more flexible to increase the capacity of payload more than EZHS method. Moreover, EZTE shows high visual quality using human visual system at huge payloads of secret data. Finally, the compression ratio of coding EZW algorithm is affected by about 1% in EZTE more than that of EZHS.