Anti-counterfeiting new technology: digital watermarking (below)

Looking at the previous research work on digital watermarking in print anti-counterfeiting, there is not much data available. The existing digital watermarking techniques based on halftone images generally have the following types:

A digital watermarking technique based on halftone images embeds invisible digital information into a halftone image, and obtains embedded watermark information by scanning the halftone image and implementing a certain extraction algorithm. One of the techniques is to use two different dithering matrices to generate sequential dithered halftone images. Due to the different statistical characteristics of the two different dithering matrices, the presence or absence of the watermark can be detected.

The second technique is to use vector quantization to embed the watermark into the least significant bit (LSB) of the error diffusion halftone image.

3 The third technique is to conceal the watermark data in a random position in a sequenced jitter or error diffusion image, and to change the pixel values ​​of these random positions by self-tightening or paired nesting techniques to achieve information hiding. 3 is shown).

4 The fourth technique is to hide the watermark data in the error diffusion image and use the error diffusion method to reduce the image distortion caused by the hidden data so that the image can obtain a good visual effect. In the above method, although the hidden data itself is also an image, these algorithms do not provide a method of directly observing these hidden images on the halftone image.

5 Another technique is to embed hidden visual images in two or more halftone images. When these two images are superimposed, the hidden images can be directly revealed. One of these techniques is to use stochastic or conjugated screens to embed hidden information in halftone images. Because of the nature of the conjugation, when the two halftone images are superimposed, the hidden information will be presented (the effect diagram is shown in Figure 4).

6 Another technique is to consider noise balance error diffusion and halftone digital watermarking techniques that alternately use different error diffusion kernels.

7 There is also a human eye vision system model based on halftone technology, and a direct binary search (DBS) method is proposed. DBS is used for the design of halftone images to achieve optimal visual quality and watermark detection.

3, results discussion

We used Matlab6,5 programming simulation to achieve the error diffusion screening of 256 grayscale images (Fig.3a) (using the Jarvis-Nike error diffusion kernel) to obtain the halftone image (Fig. 3b), and use Different digital watermark embedding algorithms have obtained halftone images containing watermark information respectively: as shown in Fig. 3 - c and Fig. 4 - b.

The watermark embedding technique used in Fig. 3-c hides the watermark data in a random position (determined by the key) in the error diffusion image, and conceals the information by changing the pixel values ​​of these random positions. The resulting watermarked halftone image is compared with the original Iena halftone image shown in Fig. 3-b, and it can be seen that it is equivalent to adding some noise to the original image when the embedded watermark information is not large, within the contrast limit of the HVS. , it will not change its visual effects. The watermark embedding technique used in Fig. 4b is to embed the watermark information into the original halftone image based on the features of the original halftone image and the watermark image to obtain the halftone image containing the watermark information (Fig. 4-b). When the image is overlaid with the original halftone image, the watermark image can be reproduced.

Under the experimental conditions, after the obtained images are printed, scanned, and preprocessed, corresponding watermark extraction algorithms can be used to extract the corresponding watermark images. In the experiment, in addition to using the binary image in the figure, we also used the text, Wuhan University emblem, and other satisfactory results.

Mediabridge technology pioneered by Digimarc in the United States in 1995 was integrated into Phot0Sh0P and CorelDraw image processing software as a plug-in.

Subsequently, it introduced a new way to open up access to the Internet - by hiding invisible digital watermarks in magazine ads, product packaging, catalogs, and even various types of tickets. Users place these traditional media in front of the webcamera and can directly take the user to the web site associated with the printed image content and display the relevant information of the product on the computer. The VerlData family of software from Signum, UK, uses an invisible watermark to block trademark piracy and the forgery, piracy and unauthorized changes of important documents. The SafePaper software designed by AlpVision, Switzerland for print documents can be embedded in Microsoft Word software. It hides the watermark information (such as trademark patent name amount, etc.) in the printing paper to prove the authenticity of the document. Japan’s IBM Tokyo Research Laboratory proposed using data hiding as a solution to identify sources of digitized photographs, confirming the integrity of digital photographs, judging whether photographs have been tampered with, and locating tampering.

Researchers in related departments of our country are stepping up their research work on the digital watermark printing anti-counterfeiting technology. Digital watermark anti-counterfeiting printing technology researched by Dalian University of Technology can realize the addition of digital watermarks to the printed images, and can complete the digital printing of images through scanners and special software. Automatic detection of watermarks. Shanghai Aspire Digital Technology Co., Ltd. has also launched an image digital watermarking software ASureImageMark2 with its own intellectual property rights. 0: The digital watermarking technology of printing and printing researched by Chengdu Yufei Information Engineering Co., Ltd. has been commercialized and applied in April 2004. On the 1st, it was funded by the National Science and Technology SME Innovation Fund. It indicates that China's digital watermarking technology has entered a substantive stage, and the gap between its level of research and development and application and the international advanced level has gradually narrowed.

The most new technology - the application of digital watermarking in the printing field is a new direction. We believe that digital watermarking technology will play an increasingly important role in the field of security printing and will bring considerable economic benefits. And social benefits. However, we should also see that the current digital watermarking products used in print anti-counterfeiting are not yet technically mature and there is still a lot of research work to be done.