Details

Autor(en) / Beteiligte

• Lan, Chi-Feng
• Wang, Chung-Ming
• Lin, Woei

Titel

A Novel Adaptive Image Data Hiding and Encryption Scheme Using Constructive Image Abstraction

Ist Teil von

• Applied sciences, 2023-05, Vol.13 (10), p.6208

Ort / Verlag

Basel: MDPI AG

Erscheinungsjahr

2023

Quelle

EZB Electronic Journals Library

Beschreibungen/Notizen

• Image abstraction simplifies complex images, highlights specific features, and preserves different levels of structures to achieve a desired style. This paper presents a constructive and adjustable data hiding algorithm to convey various secret messages and resist modern steganalytic attacks. Our scheme produces an abstracted stego image, while synthesizing an original image during the image abstraction process. Our algorithm is flexible, applicable to two types of images: high-dynamic-range images and ordinary color images, aka low-dynamic-range images. Additionally, we introduce a novel image encryption scheme suitable for the above two types of images, which incorporates a two-dimensional logistic tent modular map and a bit-level random permutation technique, thereby further protecting the content of the stego image and the carried secret messages. Compared with the current state-of-the-art methods, our algorithm provides a 14% to 33% larger embedding rate, while lowering the distortion of the abstracted stego image. A comprehensive security analysis confirmed that our algorithm provides high security to resist statistical, differential, brute force, chosen-plaintext, and chosen key attacks.