Optimal Image Based Information Hiding with One-dimensional Chaotic Systems and Dynamic Programming
Yinglei Song1, Jia Song2, and Junfeng Qu3
1School of Electronics and Information Science, Jiangsu University of Science and Technology, China
2Department of Electronic and Information Technology, Suzhou Vocational University, China
3Department of Computer Science and Information Technology, Clayton State University, USA
Abstract: Information hiding is a technology aimed at the secure hiding of important information into digital documents or media. In this paper, a new approach is proposed for the secure hiding of information into gray scale images. The hiding is performed in two stages. In the first stage, the binary bits in the sequence of information are shuffled and encoded with a set of integer keys and a system of one-dimensional logistic mappings. In the second stage, the resulting sequence is embedded into the gray values of selected pixels in the given image. A dynamic programming method is utilized to select the pixels that minimize the difference between a cover image and the corresponding stego image. Experiments show that this approach outperforms other information hiding methods by 13.1% in Peak Signal to Noise Ratio (PSNR) on average and reduces the difference between a stego image and its cover image to 0 in some cases.
Keywords: Encryption and hiding, minimized hiding effects, improved security, convenient recovery.
Received June 3, 2019; accepted February 15, 2021