Enhanced Coverless Image Steganography Using Bit-Plane Segmentation, DWT and ORB for Secure Data Transmission

Le résumé fourni par la source

Coverless Image Steganography (CIS) is an advanced technique for secure data transmission that avoids direct modification of cover images, thereby reducing vulnerability to steganalysis. This work introduces an enhanced CIS method that combines Bit-Plane Segmentation, Discrete Wavelet Transform (DWT), and Oriented FAST and Rotated BRIEF (ORB) feature extraction to improve security, robustness, and imperceptibility. The process begins with bit-plane segmentation, where the most significant bit plane (MSB) of an image are extracted, focusing on the higher-order bits that retain crucial structural information. Next, DWT is applied to decompose the image into frequency subbands, with the LL subband selected for further processing due to its high energy concentration and resilience against compression and noise. ORB feature extraction is then performed on the MSB planes, detecting keypoints and generating binary descriptors that uniquely represent the image content. These ORB-based descriptors are mapped to secret information, ensuring secure data embedding without altering the original image. By integrating these techniques, the proposed approach improves imperceptibility, maintains robustness against image transformations, and strengthens security by using featurebased data retrieval instead of traditional embedding methods, making it a promising solution for covert communication.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.