The Research On Digital Image Watermarking Algorithm Based On HVS And DWT Transform

The Digital Watermarking Technology as information hiding as the trend of future development deserved more and more attention by people. It can use human's visual redundancy to verify the ownership of the protection of digital products by embedding the watermark in the original image information. This method can hidden watermark information in the original image at the premise of not undermine the original image, and has the merit of hidden information invisible, at the same time it also has the merit of good robustness, even if the image is transmitted, copied, or illegally tampered, it also can make judgments and determine the question of attribution of ownership by the watermark that extracted. But now, there are many problems in watermarking technology, such as resilience against is not strong in geometric attacks and so on, this is the value and significance in this article's study.In this article, we introduce the basic theory of digital watermarking technology at first, discuss the digital watermarking technology's classical algorithm, proposed a digital watermarking technique based on wavelet analysis. Combine the digital watermarking technique based on wavelet analysis and human visual system, through this method make the embedded watermark not only has the good robustness but also not has the error in the visual.The algorithm in this article mainly studied the invisible watermark. Now the Embedding and extraction algorithm mainly based on HVS digital watermarking algorithm in wavelet, this algorithm before in the embedded watermark, Arnold has taken on the watermark scrambling treatment, this method is to ensure the saboteurs can not deserve the true face of the watermark information even if the watermark is extracted by vandals, so it can improve the security. The scrambled watermark with wavelet transform,s multi-resolution decomposition characteristic and human vision embedded by addition criteria, it can deserve the watermarked image after wavelet the watermarked image which is the embedded watermarked image. The disadvantage of this method is in the watermark extraction process must require the original image, For the shortcomings of the above algorithm, this article present a new algorithm that not require the original image, that is HVS-based wavelet domain adaptive image watermarking algorithm. The merit of this algorithm is not only combine wavelet transform and human visual system but also in the process of watermark embedding it can add watermark strength adaptively, at the same time it can ensure the contradiction of invisible and robustness, make the watermark information has the good robustness, also invisibility has been met. This algorithm not require the original image when extract the watermark, achieved bland watermark detection.At last the article by Matlab simulation experiment, from experiment results, it can deserve the conclusion that HVS-based wavelet domain adaptive image watermarking algorithm can accept all kinds of attacks, and it has good robustness, and the merits of embedded watermark invisibility.