Blind Digital Image Watermarking Robust Against Histogram Equalization
G. S. Kalra, R. Talwar and H. Sadawarti
DOI : 10.3844/jcssp.2012.1272.1280
Journal of Computer Science
Volume 8, Issue 8
Problem statement: Piracy in the presence of internet and computers proves to be a biggest damage to the industry. Easy editing and copying of images yields a great damage to the owner as original images can be distributed through internet very easily. To reduce the piracy and duplicity of the digital multimedia files, digital watermarking technique is dominating over the other available techniques. There are certain methods or attacks which are used to damage the watermark. One of the major attacks is histogram equalization and reducing the number of histogram equalized levels. Thus, there is a need to develop a method so that the watermark can be protected after histogram equalization. Approach: A blind digital watermarking algorithm is presented which embed the watermark in frequency domain. Firstly, DWT is applied on the original image and then DCT on the 4×4 blocks to target the particular frequencies of the image for embedding the watermark which does not have more effect after histogram equalization. Also, to enhance the security of the watermark dual encryption technique is deployed. Results: Algorithm applied to four images which are Lena, Cameraman, Baboon and Peppers. The evaluation of the algorithm is calculated in terms of peak signal to noise ratio and non correlation. The results prove that the algorithm is robust to histogram equalization attack up to 2 grey levels. Conclusion/Recommendations: The developed algorithm proved its performance against histogram equalization but the algorithm can also be checked for the other attacks which can be addition of white noise, Gaussian noise, filtering.
© 2012 G. S. Kalra, R. Talwar and H. Sadawarti. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.