CSNTSteg: color spacing normalization text steganography model to improve capacity and invisibility of hidden data

The rapid growth of online communication has increased the demand for secure communication. Most government entities, healthcare providers, the legal sector, financial and banking, and other industries are vulnerable to information security issues. Text steganography is one way to protect secure communication by hiding secret messages in the cover text. Hiding a large amount of secret information without raising the attacker’s suspicion is the main challenge in steganography. This paper proposes the Color and Spacing Normalization stego (CSNTSteg) model to resolve the low capacity and invisibility problems in text steganography. CSNTSteg consists of two stages: the pre-embedding stage, which achieves high capacity by utilizing RGB coding and character spacing. It is designed to increase the number of bits per location and usable characters. Besides, it applies the Huffman coding technique to compress the secret message to add more capacity enhancement. The second stage is color and space normalization, which accomplishes high invisibility by normalizing the RGB coding and character spacing of the cover and stego text. CSNTSteg overcomes the color difference issue between the cover and stego texts regardless of the color of the cover text. To assess the quality of CSNTSteg, the experimental results are compared with existing works. CSNTSteg shows superior capacity over the existing studies with a percentage of 98.85%. CSNTSteg also achieves high invisibility by reducing the color difference with a percentage of 4.7% and 5.07% for black and colored cover text, respectively. Furthermore, CSNTSteg improves robustness by 94.22% by reducing the distortion in stego text. Overall, the CSNTSteg model embeds a high capacity of secret data while maintaining invisibility and security, offering a new perspective on text steganography to protect against visual and statistical attack issues.

Actions (login required)

View Item