Reversible watermarking authentication algorithm for color images based on compressed sensing

Dong Ruihong, Zuo Hangzhou, Zhang Qiuyu, Wu Dongfang

Research output: Contribution to journalArticlepeer-review

Abstract

Aiming at the shortcomings of existing reversible watermarking for image authentication, such as poor ability of tamper detection and localization, and low attention of reconstruction after tampering, a reversible watermarking authentication algorithm based on compressed sensing for color image was proposed. On the side of the sender, the original image has been divided into blocks and carried out compressed sensing to generate image hash which works as the watermark information, and was embedded by the reversible watermarking of difference histogram algorithm. On the side of the receiver, watermark was extracted for authentication. For the tampered block which was failed in authentication, original image will be restored by reconstruction of compressed sensing. Experimental results show that the algorithm combines of reversible watermarking and compressed sensing, so that the detection rate of image authentication has been improved, as well as the robustness to resist sparse noise and cutting and the ability to reconstruct the original image.

Original languageEnglish
Pages (from-to)1073-1089
Number of pages17
JournalInternational Journal on Smart Sensing and Intelligent Systems
Volume9
Issue number2
DOIs
StatePublished - Jun 2016
Externally publishedYes

Keywords

  • Compressed sensing
  • Difference histogram
  • Image authentication
  • Image reconstruction
  • Reversible watermarking

Fingerprint

Dive into the research topics of 'Reversible watermarking authentication algorithm for color images based on compressed sensing'. Together they form a unique fingerprint.

Cite this