A Robust Coverless Image Steganography Based on DCT-SVD

Abstract

Coverless image information hiding plays a pivotal role in secure communications. In this paper, a novel coverless image steganography algorithm based on Discrete Cosine Transform and Singular Value Decomposition (DCT-SVD) with a focus on enhancing its robustness and resistance to steganalysis has been proposed. To begin with, DCT-SVD is implemented to the each image of the image database. Notably, the generation of hash sequence is shared between the sender and receiver to ensure correct information extraction. Then corresponding hash sequence of each image is calculated, and an inverted index is created, which links the hash sequences to image paths. After that, the transmitted secret message is seperated into a number of information segments according to the length of hash sequence. Then each information segment is matched with the DCT-SVD hash sequence, and the relevant image is identified. The receiver, upon receiving these images, employs the same method to generate corresponding hash sequences and subsequently reconstructs the secret message. The experimental results and analysis indicate that, compared to existing coverless image steganography algorithms, the algorithm proposed in this paper demonstrates enhanced robustness, particularly in its resistance to geometric attacks, and in common image processing scenarios.