Advancements and challenges in coverless image steganography: A survey

Abstract

Coverless image steganography has emerged as a significant research direction in the field of steganography in recent years. Unlike traditional image steganography, it does not require modifying the cover image to achieve information hiding. This review aims to systematically summarize the research progress and challenges in coverless image steganography. Firstly, the paper introduces the basic principles and classification methods of coverless image steganography, including embedding methods based on low-level image features and those combining advanced semantic features from deep learning. Secondly, it discusses key research achievements in this field, such as novel embedding algorithms, efficient extraction methods, and robustness enhancement techniques against various attacks. Additionally, the review highlights major challenges faced by current coverless image steganography, including difficulties in secret information extraction, capacity limitations, and practicality issues, and explores potential solutions and future research directions. Through comprehensive analysis of existing literature, the review aims to provide researchers with a holistic perspective, fostering further development and application of coverless image steganography. The paper includes 124 key contributions, offering a comprehensive overview of coverless image steganography, covering its fundamental principles, research progress, challenges, and solutions.