Robust recursive watermarking technique in discrete wavelet transform
Presently, data sharing and information searching is easier to perform on the internet and has resulted in the digital contents becoming widely available and easily accessible. However, many users abuse these contents through piracy and forgery practices, thus there is a need for copyright protectio...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/37963/1/NurulBadriahAbuBakarMFSKSM2013.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Presently, data sharing and information searching is easier to perform on the internet and has resulted in the digital contents becoming widely available and easily accessible. However, many users abuse these contents through piracy and forgery practices, thus there is a need for copyright protection which can be achieved with digital watermarking. A robust digital watermark should be able to withstand intentional and unintentional attacks but the various available techniques for watermarking have yet to attain the best defence performance against these attacks. This study proposed an alternative watermarking technique referred as Recursive Watermarking Technique (RWT) on digital image content where multiple watermarks are embedded in the host image. In this technique, multisegmentation was carried out. Embedding and extracting of watermark was performed in the Discrete Wavelet Transform (DWT) domain, after the image segmentation process. Besides that, reconstruction image stage was carried out to get the most robust watermark. These multiple watermarking processes in RWT have the capability to minimize the effect of the attacks. The robustness of RWT against attacks was tested against motion blur, Gaussian noise (1%, 5% and 10%), salt and pepper noise (0.02), cropped image, JPEG compression, intensity adjustment, sharpen and mosaic tile attacks. The results showed that RWT has a higher NC value which is equal to 1. At the same time, Gaussian blur, salt and pepper noise (0.05 and 0.1), and histogram equalization attacks gained an NC value of 0.99. These results imply that RWT is able to withstand the attacks successfully and performs better than the other known techniques. |
|---|