Image self-embedding based on EZW coding and fountain codes

Chun Liao,Zhenxing Qian,Jing Ni,Shuozhong Wang
DOI: https://doi.org/10.3969/j.issn.0255-8297.2014.01.011
2014-01-01
Abstract:We first utilize EZW to encode the wavelet coefficients of the image, and then the fountain encoding to generate the reference data. By extracting forensic data and reference data which is self-embedded into the image, tampered area is detected and reference data is also reconstructed. Comparing to block-based DCT, we overcome the blocking artifact of the recovered image. Besides this, we propose a method which makes the data embedding and image quality dynamic by adjusting the parameters. A balance between the recovered area and recovered quality is also achievable.
What problem does this paper attempt to address?