Image analysis – Applications
Reexamination Certificate
1998-09-14
2001-06-12
Johns, Andrew W. (Department: 2621)
Image analysis
Applications
Reexamination Certificate
active
06246775
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a technology in which digital watermark information (hereafter referred to as the watermark) is superposed on, or embedded in hidden form in, data on image or sound data, and more particularly to a method for embedding a watermark on an information data signal in block units each consisting of a group of a small number of pieces of information data, a method of detecting the watermark from the information data signal with the watermark embedded therein, and a method in accordance with these methods.
2. Description of the Related Art
Currently, as optical recording media on which information data can be written, DVD-Rs and DVD-RAMs are being put to practical use, but in putting the DVD-Rs and DVD-RAMs to practical use, it is necessary to devise measures for preventing illegal copying of video and audio software and the like. Accordingly, attention is being focused on a digital watermark technology in which copyright information or information indicating a copy guard is expressed by a watermark which is a visually or audibly inconspicuous image pattern, and this watermark is superposed on the image data or the sound data.
The basic system of such a digital watermark technology is broadly classified into type types. One is a type in which sampled values of waveform and pixels are provided with processing, and the watermark is embedded therein. For instance, a technique in which the watermark is added to luminance values of the image corresponds to this type. As another type, there is a system in which the image data and the sound data are converted to frequency components, and the watermark is embedded in particular frequency components. The fast Fourier transform (FFT), the discrete cosine transform (DCT), and the like are used in such frequency conversion.
On the other hand, a technique is also known in which data is divided into a plurality of small pixel blocks, and watermark information is embedded (hereafter referred to as the block division method) (see FIG.
1
). In this technique, image data is divided into a plurality of small unit pixel blocks each having a size of N×N pixels, and a watermark having the same block size of this unit pixel block is added to the pixel block. According to this technique, even if a portion of image data of one frame is extracted, the watermark is left insofar as its size is not smaller than that of the unit pixel block. On the reproduction side, the embedded data is divided into similar unit pixel blocks, and the watermark is detected for each unit pixel block and is decrypted.
Here, a form is considered in which, for example, image data in which a watermark is embedded is recorded on a disk such as a DVD, and the image data is read from the disk and is reproduced.
There are cases where a disk player for reading such a disk is required to output a first image signal having an aspect ratio of 16:9 for displaying an image on a so-called wide-screen television, as well as a second image signal having an aspect ratio of 4:3 for displaying an image on ann ordinary television. If it is assumed that image data for forming a squeezed-type original image with 480 [dots] high and 720 [dots] wide has been recorded on the DVD, the disk player must perform aspect ratio conversion with respect to the image data obtained from the DVD in order to generate the first image signal. The same also applies to the second image signal, and the disk player must perform the aspect ratio conversion with respect to the image data obtained from the DVD in order to generate the second image signal.
The aspect ratio conversion into the first image signal can be basically accomplished by effecting interpolation of pixels in the horizontal direction of the screen, as shown in FIG.
2
A. Two forms such as those shown in
FIGS. 2B and 2C
are conceivable as the aspect ratio conversion into the second image signal.
As shown in
FIG. 2B
, the squeezed-type original image with 480 [dots] high and 720 [dots] wide recorded on the DVD is converted into image data by being upsampled in such a manner that the aspect ratio becomes 4:3. This is a converted image which is a so-called pan scan, and assumes a form in which left- and right-hand side end portions of the image based on the image data obtained from the DVD are cut off so that the image assumes the aspect ratio of 4:3. Meanwhile, as shown in
FIG. 2C
, the original image recorded on the DVD is converted into image data by being downsampled in such a manner that the aspect ratio becomes 4:3. This is a converted image which is a so-called letter box, and assumes a form in which predetermined images (e.g., strip-like images of a black color) are pasted on upper and lower end portions of the image based on the image data obtained from the DVD so that the image assumes the aspect ratio of 4:3.g
In the same way as the first image signal, the second image signal converted into such a pan scan or letter box image can be recorded on a recording medium such as the aforementioned DVD-R or DVD-RAM by the DVD recorder, for example. At this time, however, in the DVD recorder, the watermark embedded in the second image signal to prevent an act of infringement of the copyright is detected and decrypted. If it is detected that the relevant watermark bears, for instance, information indicating prohibition of copying, the DVD recorder prohibits its own recording operation, and does not record the second image signal even if a recordable DVD is supplied thereto. On the other hand, only when it is detected that the relevant watermark bears information indicating that copying is allowed, the DVD recorder is able to record the second image signal on the DVD supplied thereto.
In the case of
FIG. 2A
, since the DVD player effects conversion into the information of the 16:9 image by using all the information of the squeezed-type original image and by expanding the information, in the block division method, the upper leftmost unit pixel block in the original image, for example, is positioned in the converted 16:9 image at the same upper leftmost position by changing its size. Therefore, the DVD recorder is capable of properly detecting and decrypting the watermark consecutively starting with, for example, this upper leftmost unit pixel block of the 16:9 image data supplied thereto.
However, as can be seen from what is shown in
FIGS. 2B and 2C
, in the case of the images such as the pan scan and the letter box, their pixel positions and sizes have changed from those of the original images, so that the recognition of the unit pixel blocks in the DVD recorder needs to be made different from that in the case of FIG.
2
A.
To give a more detailed description, in the case of
FIG. 2B
, since the DVD player effects conversion into the information of the 4:3 image by eliminating the information of left- and right-hand side portions of the squeezed-type original image and by expanding the remaining information, in the block division method, the upper leftmost unit pixel block in the original image is already missing in the converted 4:3 image. Further, even if, for example, the third unit pixel block from left in the uppermost row is present in the converted 4:3 image, there is a high possibility that the 4:3 image starts midway in the relevant block, as illustrated in the drawing.
In such a case, if the DVD recorder starts detection of the watermark without recognizing that, for example, the third unit pixel block from the left in the 4:3 image data supplied thereto is fragmentary, the DVD recorder fails to properly detect not only that third unit pixel block but also ensuing unit pixel blocks.
In addition, in the case of
FIG. 2C
, since the DVD player effects conversion into the information of the 4:3 image by synthesizing predetermined images on upper and lower sides of the squeezed-type original image by expanding the synthesized image information, in the block division method, the leading upper leftmost unit pixel block i
Nakamura Takeshi
Oriyama Yoshiaki
Sugaya Kazumi
Johns Andrew W.
Nakhjavan Shervin
Pioneer Electronic Corporation
Sughrue Mion Zinn Macpeak & Seas, PLLC
LandOfFree
Method and appartus for superposing a digital watermark and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and appartus for superposing a digital watermark and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and appartus for superposing a digital watermark and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2503869