الفهرس 
CHAPTER (1) INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 268 |  |  |
| | | from | 268 | | |
Abstract
This thesis proposed image steganography algorithms and presented their Field Programmable Gate Array (FPGA) implementation using Xilinx System Generator (XSG), while the H/W co-simulation is done using Spartan 3E FPGA Kit. Different cases of Least Significant Bit (LSB), Random Pixel Embedding (RPE) and developed Generalized Exploiting Modification Direction (developed GEMD) spatial domain image steganography algorithms for embedding secret messages in grayscale and color images are proposed and different cases of hybrid Integer Wavelet Transform and Least Significant Bit (IWT-LSB) image steganography algorithms for embedding data in grayscale and color images are proposed. The proposed image steganography algorithms achieve high payload capacity, security, robustness and imperceptibility, low complexity and perfect recovery of the secret message at destination. The results reveal that, the proposed LSB achieve high stego-image quality but it is not secure, the proposed developed GEMD achieve high embedding rate and high payload capacity with acceptable stego-image quality and the proposed IWT-LSB is secure but it has a lower payload capacity.
This thesis also proposed hybrid watermarking techniques for Three Dimension High Efficiency Video Coding (3D-HEVC) streams and presented their FPGA implementation using XSG. The proposed hybrid watermarking techniques for 3D-HEVC streams increase the immunity of the watermarked streams to attacks. In this thesis, two efficient watermarking techniques are proposed, the proposed homomorphic transform based on Singular Value Decomposition (homomorphic+SVD) in Discrete Wavelet Transform (DWT) domain and the proposed three-level Discrete Stationary Wavelet Transform (three-level DSWT) in Discrete Cosine Transform (DCT) domain. These techniques achieve high degree of capacity, security and robustness without affecting on the 3DV perceptual quality in the presence of attacks. The results reveal that the proposed homomorphic+SVD watermarking in DWT domain is better than the proposed three-level DSWT watermarking in DCT domain.
IV
Furthermore, this thesis proposed different image fusion techniques and presented the FPGA implementation of the proposed optimized DWT based fusion technique using XSG. The results of the proposed algorithms are appreciated and recommended than the results of the traditional fusion algorithms. The proposed fusion algorithm improves noticeably of the fused image quality with shorter processing times. Therefore, the proposed optimized fusion algorithm is more recommended and can be applied in real time medical applications.