An Enhanced Steganography Network for Concealing and Protecting Secret Image Data

Area of Science:

• Computer Science
• Information Security
• Digital Image Processing

Background:

• Internet technology facilitates data sharing but introduces significant security risks for sensitive information.
• Protecting private or confidential image data requires robust security measures against unauthorized access.

Purpose of the Study:

• To design an enhanced steganography network for secure embedding and recovery of secret image data.
• To improve the visual quality, imperceptibility, and security of hidden images.

Main Methods:

• Developed a steganography network with concealing and revealing components using a down-up structure to reduce computational complexity.
• Implemented long skip concatenation to preserve information and non-activated feature fusion (NAFF) for enhanced image synthesis.
• Integrated an attention mechanism for salient target reconstruction and a hybrid loss function (pixel and structure domains) for improved visual quality and security.

Main Results:

• The proposed network effectively synthesizes high-quality hidden images with fine texture details.
• Experimental results confirm high levels of imperceptibility and security for the embedded secret data.
• The enhanced module and hybrid loss function significantly boost visual quality and data protection.

Conclusions:

• The designed steganography network offers a secure solution for protecting sensitive image data.
• The novel network architecture and loss function contribute to superior imperceptibility and security.
• This method enhances the visual quality and structural integrity of hidden images, ensuring data confidentiality.