Digital Watermarking Algorithm Resistant to Geometric Attacks Based on Harris Corner

CHEN Qing, SUN Yan-fei, GUO Gong-xun

Packaging Engineering ›› 2018 ›› Issue (13) : 188-193.

PDF(771 KB)
PDF(771 KB)
Packaging Engineering ›› 2018 ›› Issue (13) : 188-193. DOI: 10.19554/j.cnki.1001-3563.2018.13.031

Digital Watermarking Algorithm Resistant to Geometric Attacks Based on Harris Corner

  • CHEN Qing, SUN Yan-fei, GUO Gong-xun
Author information +
History +

Abstract

The work aims to propose a new algorithm that can effectively resist the geometric attacks based on Harris extracting feature region combined with landmark region, with respect to the good anti-geometric attack ability of the second generation digital watermarking algorithm. Firstly, the carrier image was pre-processed to obtain Harris from the standard image. The feature points were extracted and screened, and they should be stable and concentrated. Secondly, the feature region to embed in the watermark was determined according to the watermark size, and then the feature region was screened and subject to wavelet decomposition. Finally, the wavelet coefficients were adjusted according to the subband coefficient mean and modulation factor. The blocked watermark information was respectively embedded into the sub-bands after the decomposition of the feature region. Experiments showed that the proposed algorithm had good invisibility, and the extracted watermarks were of high similarity. The PSNR values were all more than 65 dB and the NC values were all greater than 0.8. The watermark extraction of the proposed algorithm does not require original images but is of extremely robustness, which can effectively resist the attacks of the common signal processing.

Cite this article

Download Citations
CHEN Qing, SUN Yan-fei, GUO Gong-xun. Digital Watermarking Algorithm Resistant to Geometric Attacks Based on Harris Corner[J]. Packaging Engineering. 2018(13): 188-193 https://doi.org/10.19554/j.cnki.1001-3563.2018.13.031
PDF(771 KB)

Accesses

Citation

Detail

Sections
Recommended

/