Image analysis – Applications
Reexamination Certificate
1998-05-07
2001-02-20
Lee, Thomas D. (Department: 2724)
Image analysis
Applications
C382S235000, C380S054000, C380S200000, C283S113000
Reexamination Certificate
active
06192138
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to image data processing techniques, such as JPEG systems, MPEG systems, etc., and more particularly to an apparatus and a method for embedding/unbedding supplemental information and a record medium therefor, which are applied with an electronic watermark technology for embedding the copyright data or personal identification data into the digital image data.
BACKGROUND OF THE INVENTION
Recently, DCT (Discrete Cosine Transform)-based image compression/expansion technologies such as JPEG (Joint Photographic Coding Experts Group) systems and MPEG (Moving Picture Experts Group) systems have come into practice as fundamental technologies for realizing multimedia.
An electronic still camera has become widespread in use, which converts an imaging object into an electric signal by an optoelectric transducer device such as a CCD (charge coupled device), performs the specific digital signal processing, then records the digitized still rage into a recording medium such as a memory card. The image data are digitized for compressing the huge amount of its code, using the high efficiency encoding such as the JPEG (Joint Photographic Coding Experts Group).
Further, with the establishment of the high efficiency encoding technologies for such a still image, the digital processing of a moving image has been developed. The high efficiency encoding technology used in the moving picture is to encode the image data at the lower bit rate so as to improve the efficiency of the digital transmission or recording, and an MPEG-1 (ISO/IEC 11172) of the high efficiency encoding standard is already used in various media, such as a video CD (Compact Disc), a CD-I (Compact Disc-Interactive), etc. Further, an MPEG-2 (Moving Picture Experts Group 2) standard which is a standard corresponding to the image quality of the conventional broadcasting has also been put to practical use, and picture software items based on the MPEG-2 standard have been on sale.
An embodiment of the conventional JPEG system image compressing apparatus is shown in FIG.
1
.
An image that is input to an image compressor is divided into 8×8 pixel blocks, and then rearranged to frequency components by the DCT processing carried out in a DCT processor
51
. The blocks rearranged to frequency components (DCT coefficients) are transformed into one dimensional format from two dimensional format by the zigzag scanning or the alternate scanning carried out in a scanner
52
. A quantizer
53
divides each DCT coefficient of the blocks converted into one dimensional format by a quantization matrix table coefficient, and then obtains quantized data from the result of this division by rounding it to a nearest whole figure. A VLC processor
54
carries out the reversible compression (variable length encoding) of the quantized DCT coefficients. Compressed data of the blocks are thus obtained. By executing the above operations for all blocks of input images successively, compressed image data for one frame are obtained.
Further, in the case of the MPEG to compress moving pictures, the DCT processed blocks are quantized and the quantized blocks are transformed into one dimensional format from two dimensional format. That is, the scanning procedures of the MPEG system is reverse to those of the JPEG system.
Now, the configuration of the JPEG system image decoder is partially shown in FIG.
2
.
A VLD processor
61
carries out the variable length decoding of input image compression data by looking up the Huffman table. The variable length decoded data (quantized DCT coefficients) are reverse quantized by multiplying by coefficients of the quantization matrix table in a reverse quantizer
62
. The quantized DCT coefficients are reverse scan transformed from one dimensional format to two dimensional format blocks in a reverse scanner
63
, and then inverse DCT processed in an IDCT processor
64
, and 8×8 pixel blocks which are close to an original image are thus reproduced.
Further, in the case of the MPEG, the variable length decoded data are reverse quantized after being reverse scan transformed. Demands for and the necessity of embedding such supplemental information
72
as character information for explaining contents of an image
71
in an electronic album, as shown in
FIG. 3
, ID information for individual identification for the purpose of preventing illegal copying of digital images are increasing in recent years.
For adding character information, there are several methods available, including a method of storing image data and character information in separate files, and a method of producing one file by pasting image data and character information on separate application softwares.
However, the method of storing image data and character information in separate files results in an increase of the number of files. In addition, the method of adding image data and character information on separate application software results in an increase in the file size and necessitates an exclusive reading device and a software when a file name changes.
Further, recently, the need for protecting the copyrights of these digital still pictures or digital moving pictures has increased. One of the techniques for resolving these issues is digital watermark technology.
The digital watermark technology embeds the digital watermark such as a sentence or a logo mark which could be used to protect the copyright or a trademark, into the image data of the digital still picture such as a JPEG compression image or the digital moving picture such as an MPEG compression image. The copyright data are embedded in the JPEG image data or MPEG image data in the way that the embedded data are not displayed in accordance with the JPEG or the MPEG standard, even if the JPEG image data or the MPEG image data are decoded.
If a JPEG or an MPEG image data accompanied by a digital watermark representing a copyright has been illegally copied, a copyright owner can confirm his copyright by unbedding the digital watermark from the copy of image data in accordance with a prescribed procedure.
Now, a conventional system for embedding the digital watermark into the JPEG image will be described. Conventionally, the digital watermark is embedded into the JPEG image after a JPEG compression has been made.
That is, the image data captured by the image capturing device such as a CCD are divided into 8×8 pixel blocks and are input to the frequency converter for carrying out a DCT (Discrete Cosine Transform) which is for a quadrature transform encoding. In the DCT, which operates as a frequency converter, the 8×8 pixel blocks are arranged in different order of the frequency ingredients. The rearranged blocks (DCT coefficient) are rounded off to the nearest integer by the quantizer, after the coefficient corresponding to a matrix is divided by the quantization matrix table. Accordingly a first stage of compression processing (non-reversible compression) is carried out.
The quantized matrix data are converted into one dimensional format from two dimensional format by a scanner. The data converted into one dimensional format are processed through variable length encoding by looking up the encoding table, such as a Huffman table for instance, and further processed through the second compression processing (reversible compression) by a VLC (Variable Length enCoder) processor, so that the compressed data of the 8×8 pixel blocks are obtained.
After performing the operations mentioned above to all blocks of input image, the ultimate compressed data are obtained and stored in the external memory such as a memory card.
Then, a digital watermark presenting a copyright, a logo mark, or a trademark is embedded into the JPEG compression image data by using a special purpose machine, such as a computer equipped with software for embedding digital watermarks, before the image data are stored in the external memory.
In the conventional system described above, there was a need to embed the digital watermark, such as the copyright data, to
Chen Wenpeng
Kabushiki Kaisha Toshiba
Lee Thomas D.
Pillsbury Madison & Sutro LLP
LandOfFree
Apparatus and method for embedding/unembedding supplemental... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for embedding/unembedding supplemental..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for embedding/unembedding supplemental... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2561538