A Study of Image Authentication and Recovery
| Autor: | Chun-Wei Yang, 楊竣崴 |
|---|---|
| Rok vydání: | 2009 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 97 Today’s watermarking techniques can be divided into two types. The first type is robust watermarking technique which can resist malicious attacks better. The second type is fragile watermarking technique, also called image authentication, can verify whether the image is tampered in the process of image transfer or not. There are some applications requires accurate images such as medical or military images which users need to employ original images. Based on those specific requirements, we proposed a scheme of image recovery. The first proposed scheme improves the robust digital watermarking technique based on random sequence into reversible watermarking technique. Therefore, this method will use the robust watermarking technique of the well-known random sequence as embedding approach. Sobel edge detection technique is employed to extract the pixel value of edges from the watermarked image. And the final watermarked image is produced by replace the original image corresponding to the pixel value of the edges for the purpose of embedding watermark. Because the robust watermarking technique can tolerate the image which is destroyed to protect the copyright, there is no watermarking examination problem although this approach causes some loss of watermark information according to the experiment result. Moreover, it can not only examine whether the image has embedded watermarks, but also restore the original image. Majority of researches related to the image authentication are mainly focusing on improving the robustness and the ability of tamper detection on image authentication. The practical application of restrictions on reversible watermarking technique, which can’t tolerate any image processing attack, neither can tell whether the image is tampered or not. Nevertheless, image recovering is also an important issue. We proposed an image tamper detection and recovery method, which can recovery image effectively, especially for images with large tampered regions. First, a secret key is used to obtain a random sequence and determine where to embed watermarks on the basis of this random sequence. As a result, it is hard to find out the corresponding location of watermarks from embedded images. Later, through vector quantization, we create an index table as a basis for recovery and embed it in the original image. Finally, Wong’s watermarking scheme is integrated to perform tamper detection and achieve image authentication and recovery. Summary of our method, Wong’s watermarking scheme is employed to perform tamper detection. Wong’s watermarking scheme can be used to accurately locate tampered regions. If an image has been tampered, the index table can be used to recover the tampered regions. Although the image recovery technique unable to restore the image completely, receiver can identify the tampered locations and recovery immediately, as it will save additional bandwidth to retransmit the image. According to the experiment results, the first method we propose only needs to pay 5.61 k bytes overheads in average for restoring original images. Besides, the second method has the higher probability of image recovery and our scheme provides not only a better quality of recovered images but also better results at edge regions. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|