Using quad smoothness to efficiently control capacity-distortion of reversible data hiding

  • Authors:
  • Chi-Nan Lin;Daniel J. Buehrer;Chin-Chen Chang;Tzu-Chuen Lu

  • Affiliations:
  • Department of Computer Science and Information Engineering, National Chung Cheng University, 168 University Rd., Minhsiung Township, Chiayi 62102, Taiwan, ROC and Department of Management Informat ...;Department of Computer Science and Information Engineering, National Chung Cheng University, 168 University Rd., Minhsiung Township, Chiayi 62102, Taiwan, ROC;Department of Computer Science and Information Engineering, National Chung Cheng University, 168 University Rd., Minhsiung Township, Chiayi 62102, Taiwan, ROC and Department of Information Enginee ...;Department of Information Management, Chaoyang University of Technology, Taichung County 41349, Taiwan, ROC

  • Venue:
  • Journal of Systems and Software
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

One of the main uses of data hiding is to protect secret messages being transmitted on the Internet. Reversible data hiding can fully recover the original host image after extracting the secret message. It is especially suitable for applications where, after extracting the secret message, the quality of the recovered host image cannot be compromised, such as for medical or military image data. Many difference-expansion-based (DE-based) reversible data hiding methods have made use of a threshold value to control the stego-image's quality. Usually repeated trial and error is required to find a relatively good threshold with acceptable capacity-distortion behavior. This paper introduces a scheme that does not require a threshold value, such as is used in Alattar's quad-based reversible data hiding. It applies a prediction of quad of quads smoothness to determine the embedding sequence. The proposed scheme is shown to perform better than other DE-based schemes. Results showed that it has the ability of maintaining embedding quality at all capacity levels, especially when the embedding capacity is at low to medium levels.