what's new one can do in digital image watermarking using LSB method for the research work?
There are so many articles on this topic, so you may go for combining this method with other algorithms or optimization methods to improve it. This method actually is matured enough!
Dear A.H.M Jaffa IQBAL Barbhuiya,
In watermarking, some part of Cover image has to be chosen, to store the watermark. Generally, a watermarking scheme can be devised in either spatial or transform domain. What ever may be the domain, some sort of redundant or unnecessary portion has to be identified. The most most fundamental portion in spatial domain is LSBs of all the bytes. For instance, in the case of wavelet based transform domain techniques HH is such a portion.
In view of the fragility of this type of watermarking, I use it for the authentication of images and videos (authentication of copyright).
If you are embedding water mark in spatial domain then locate the homogeneous region or region where pixel intensity difference between neighboring pixels are less then five gray shades (in 8 bit gray image). Add the watermark in that location, it will not effect be detect by our vision system being because HVS can't resolve information by less then five gray shades.
There are so many articles on this topic, so you may go for combining this method with other algorithms or optimization methods to improve it. This method actually is matured enough!
Using data compression algorithms, and their concepts during image compression is one of the best point to start working on digital image watermarking systems. It is useful because you can understand the concepts of image much better, so you will come up with new algorithms in this area which is robust against lots of attacks and compression methods which would be big issue in this area.
Transform domain for digital watermarking and DCT data compression like JPEG algorithm could be good choice.
LSB method, as the majority of watermarking techniques, is quite simple and fast, compared to the transform domain methods. However, LSB is very weak against signal processing and geometric attacks. As a solution, LSB can be used with another transformation (DCT, DWT, DFT, etc) to design a robust watermarking scheme if the application is Copyright protection. Or, a fragile watermarking scheme to adress the authentication probleme.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images