Open AccessResearch on reversible image steganography of encrypted image based on image interpolation and difference histogram shift
Author(s) -
Ye Hanmin,
Su Keqin,
Cheng Xiaohui,
Huang Shiming
Publication year - 2022
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/ipr2.12461
Subject(s) - encryption , information hiding , scrambling , computer vision , interpolation (computer graphics) , artificial intelligence , image scaling , pixel , histogram , mathematics , image quality , algorithm , ciphertext , computer science , image (mathematics) , image processing , operating system
In view of the mutual restriction between the image encryption and information hiding process in the ciphertext domain reversible algorithm, the existing ciphertext domain image reversible image steganography algorithm can hardly take into account the performance of security and hiding capacity. This paper proposes a reversible image steganography algorithm for encrypted images based on a new interpolation image and difference histogram shift. The pixel position of the carrier image is changed through the double scrambling operation of random image block recombination and Arnold transform. The chaotic sequence generated by logistic mapping is used as the diffusion sequence of the encryption algorithm, and DNA encoding and calculation are performed respectively to obtain the final encrypted image. According to the newly proposed interpolation algorithm, the encrypted image is enlarged, determine the pixel value of the target interpolated image according to the pixel value of the encrypted image, calculate the difference between the three adjacent pixel values of the target interpolated image, and use the difference histogram translation algorithm to convert the secret Information is embedded in an interpolated image, and only one bit of the pixel value can be modified in a single embedding process, and the embedding method is simple and effective. Experimental results show that under the premise of ensuring information security and image quality, the embedding capacity has also been greatly improved, reaching 1,572,864 bits. When the embedding rate is 0.67 bpp, the PSNR can also reach 58 dB and the image quality is maintained better.