الفهرس | Only 14 pages are availabe for public view |
Abstract The marvelous development in data transmission technology has put lots of potential on the process of securing data than before. Lots of methods have last decades for data been developed in the protection such as steganography and cryptography. Steganography refers to the science and art of hiding information inside a carrier, where nobody except the intended recipient, knows about the existence of hidden information. On the other hand, cryptography can be identified as the conversion of data into a secret code for transmission over a public network. The hidden information can existed be in a form of (text, audio, image or video). These approaches used in concealing secret texts are seeking to find a kind of harmony between the secret text and image pixel values. The main objective of this work is to improve and propose a new hybrid technique for data security through the integration between cryptography and steganography algorithms. This system will be used to embed an encrypted secret message into a cover image to get high imperceptibility and durability with minimal deterioration in the received stego image. The proposed system was developed based on the Least Significant Bit (LSB) approach as a common steganography technique and 2D Discrete Wavelet Transform (DWT). This is in addition to improve the currently methods used in hiding secret messages, through making an integration between steganography techniques and cryptographic algorithms (Advanced Encryption Standard (AES) and Rivest Shamir Adleman (RSA)). In this hybrid approach the secret message is encrypted first before being hidden and that is an image part of the image stego. In this work, a comparison is carried out between the original image file (cover image) and the stego image coming after performing the proposed techniques. Also, the hidden text is analyzed before being transmitted and after being received by the intended recipient. This is to make sure that less distortion happens to the original cover file after embedding the secret text and were also calculated to evaluate the obtained results from the proposed approaches based on statistical values. The proposed techniques in this work were performed on different message sizes varied between 1 to 256 bytes. These messages were hidden in some grayscale and color (RGB) cover images obtained from http://sipi.usc.edu/database/database.php 2011. Various parameters were used for evaluating, the proposed techniques the Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE), and Mean Absolute Error (MAE), and Bit Error Ratio (BER), and Signal to Noise Ratio (SNR), and Structural Similarity (SSIM), and Structural Content (SC), and correlation. Experimental results presented at the end of the thesis confirm a relative improvement in the criteria used to measure the efficiency of the system. The PSNR in some extensions came to the limits of 88%. MSE in some extensions has become very low and close to zero. MAE parameter shows very small values with both color and grayscale images. This indicates no significant difference between the cover and stego image. The BER retrieval accuracy examining of the text after the adjustment has become close to 100%. The last three parameters (SSIM, SC and Correlation) take a value of one or very close to one with all the studied images (color and grayscale). This indicates a very high correlation between both the cover and stego image. |