Enhanced Steganography Framework Based On Lossless Compression And Histogram

• Main Author: Kasasbeh, Dima Suliman Mohammad
• Format: Thesis
• Language: English
• Published: 2024
• Subjects: T1-995 Technology(General)
• Online Access: http://eprints.usm.my/62336/

Steganography is an effective cybersecurity technique that facilitates covert communication by hiding information's existence within a spoof image. The trade-off between embedding capacity, imperceptibility, and reversibility has presented a new challenge in steganography. Balancing these factors is crucial and essential for the development of effective steganography. This thesis proposes an enhanced steganography framework to increase embedding capacity, maintain imperceptibility, and achieve reversibility. Three models were proposed: compression, embedding map generation, and reversible steganography. The compression model was modified to include a hybrid lossless text compression algorithm to reduce redundant secret information and boost embedding capacity. The Embedding Map Generation Model proposed two techniques to determine the best RGB channels and generate the embedding location map before the embedding process. In the reversible steganography model, a 3D embedding process is proposed. It hides compressed bitstreams in the 2D-PEH with a lower local complexity value. Additionally, it hides decompression keys along with other extraction parameters using the generated embedding location map.