الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 137 |  |  |
| | | from | 137 | | |
Abstract
With the tremendous development in data transmission technology, the process of securing data has become more important than ever before. There are a lot of methods in data protection such as encryption and data hiding. Encryption methods are used to change the content and shape of the sent information between the transmission parties, while the process of data hiding is used to prevent any external party from attention to the presence of data sent between the sender and the receiver.
For Image files, the process of data hiding inside a cover object may be done in two ways: either in the spatial domain by changing the values of the color cells (pixels) in order to contain values of the hidden message, or in the frequency domain through modifying the values of color frequencies after applying any kind transformations on the image. Data hiding system becomes more efficient and successful if it can hide as much of the data within the cover object without causing any visible effect, and without the discovery of any external party that data is hidden, but the intended entity who possesses the key to decode the process of concealment and extract data.
In our thesis, we have made a comprehensive study on methods of concealment of data and addressing the advantages and disadvantages, before we provide our innovative technique avoiding some of those defects and raise the rates of the results used to evaluate the efficiency of the system.
Our system depends on the idea of merging encryption and data hiding in order to provide a high level of protection. We also used the video files as a cover to carry hidden data, by extracting some of the images (frames) from the video file and use our proposed Algorithm to adjust the pixel values in an invisible manner using a data hiding key then merging them back into the video. All that is required of the other party is using the same key and apply the reverse processes to extract the data concealed. Our system provides an extra level of protection by conducting a noise detection process. This is done by examining a unique value of the image before and after data transmission to make sure that the image was not exposed to any process of modification or intended noise across the network.
Experimental results presented at the end of the thesis confirm a relative improvement in the criteria used to measure the efficiency of the system. Storage capacity in our system arrived to about 66% and the Peak signal to Noise Ratio (PSNR) in some extensions came to the limits of 83%. Also Mean Square Error (MSE) in some extensions has become very low and close to zero. The retrieval accuracy examining the pixels of the image after the adjustment has become close to 100%.