Cryptography – Video cryptography – Video electric signal modification
Reexamination Certificate
1999-08-11
2003-12-30
Barron, Gilberto (Department: 2132)
Cryptography
Video cryptography
Video electric signal modification
C713S176000, C348S394100, C382S236000, C382S238000
Reexamination Certificate
active
06671376
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a video scramble/descramble apparatus corresponding to motion predictive/orthogonal transform coding of videos.
Various encryption techniques have been studied and developed to prevent unauthorized duplication and unauthorized access for the purpose of protecting the copyrights of products containing audio or video information.
For example, in a DVD (Digital Versatile Disc) using MPEG2 video coding, reconstruction regions are limited by region codes, and coded data is encrypted by a CSS (Contents Scrambling System).
As a scramble technique for a baseband video signal, techniques called line rotation which randomly sets one cut point per line and replacing the right and left line sections of the cut point and line permutation for randomly replacing scan lines are known. Line rotation is used to limit access in cooperation with a billing system as a scramble technique for pay-per-view programs for satellite broadcast and CATV (cable television).
For the purpose of preventing unauthorized duplication by consumer analog video tape recorders, a copy protection technique (Microvision Corporation) is known. In this technique, AGC signals or color stripe signals in the vertical blanking interval are manipulated to disable normal recording of copy-protected tape contents by the VTRS, although such copy-protected tape contents can be normally displayed on a TV.
Furthermore, a technique called “digital watermarking” corresponding to digital contents including audio or video information is known. Digital watermarking embeds data, which cannot be visually or aurally perceived, in a baseband signal or coded data of audio or picture data, or the like. Information to be hidden by digital watermarking includes, for example, copyright information, copy generation management information, playback control information, scramble key information, and the like.
The aforementioned techniques have both merits and demerits. For example, management using region codes unconditionally allows playback in designated regions, and data encryption by a CSS or the like does not inhibit playback using an authorized player. Hence, the region code or CSS can prevent coded data itself from being duplicated, but cannot prevent unauthorized duplication of a decoded video signal. On the other hand, the duplication protection system for analog VTRs depends on models of VTRs, and cannot always assure the duplication protection effect. In addition, since only sync signals are manipulated, resistance against unauthorized attacks is not always high. Furthermore, hiding of copyright information by, e.g., digital watermarking does not always technically limit prevention of unauthorized duplication of a video signal.
More specifically, in order to prevent unauthorized duplication of a video signal, more robust copyright protection method for the video signal itself must be used. However, when a conventional video scramble technique such as line rotation or the like is used, if the scrambled video signal is coded by MPEG2 which is used in a DVD or digital broadcast, the coding efficiency lowers compared to coding of a non-scrambled picture, thus deteriorating the picture quality of the reconstructed picture. This is because the conventional video scramble makes an original video picture hard to discern by lowering temporal spatial correlation of the picture by random manipulation of the picture, and is contradictory to motion predictive/orthogonal transform coding such as MPEG2 or the like, that improves coding efficiency using the temporal spatial correlation of a picture.
This point will be described in more detail below.
MPEG2 coding uses correlation of a video signal in the space domain (intraframe correlation) and correlation in the time domain (interframe correlation), and compresses the data size by removing redundancy in both these domains. Motion prediction in units of blocks anticipates an effect of reducing video signal power using interframe correlation. To reduce the data size by the DCT (discrete cosine transform) and variable-length coding in consideration of correlation between neighboring pixels in a frame and also quantization with weights depending on frequency in consideration of the nature of human vision, or to variable-length code only the difference between DC components of neighboring blocks anticipate reduction of video signal power using intraframe correlation.
Furthermore, upon coding motion vector information in units of macroblocks, the difference between the motion vectors of neighboring macroblocks is variable-length coded as a motion vector to be coded in consideration of motion similarity between frames in association with neighboring macroblocks. In this manner, the information size to be transmitted can be reduced.
However, in the conventional video scramble technique, correlation is lowered or video contents are made hard to recognize by random manipulations for the video signal. When a video signal that has undergone processes such as conventional line rotation, line permutation, or the like is coded by MPEG2, interline correlation in a frame considerably lowers, and a reduction of signal power can no longer be expected in a combination of DCT and variable-length coding.
When vertical motion components exist in the time domain, even when an original video picture has predictive efficiency in motion prediction in units of macroblocks, the similarity between a reference picture and picture to be coded lowers as a result of scrambling, and the predictive efficiency considerably drops. More specifically, the correlation of a video signal expected in MPEG2 coding considerably lowers, and it consequently becomes hard to reduce video signal power. In order to achieve coding at a predetermined bit rate, the number of coded bits must be reduced by coarse quantization, resulting in drop of image quality of the decoded picture.
As described above, as robust copyright protection method for a video signal, a scramble process for a video signal itself is effective. However, when the conventional video scramble technique is combined with coding such as MPEG2 that uses temporal spatial correlation, the coding efficiently suffers, resulting in deterioration of image quality of the reconstructed picture.
BRIEF SUMMARY OF THE INVENTION
It is an object of the present invention to provide a video scramble apparatus and video descramble apparatus, which are free from deterioration of image quality even in coding as a combination of motion predictive and orthogonal transform like MPEG2 coding.
It is another object of the present invention to provide a video scramble apparatus which can implement video scramble that can minimize coding efficiency drop and can maintain high image quality by selecting a frame which is not used as a reference picture in interpicture predictive coding, i.e., interfield or interframe predictive coding and scrambling the selected frame using one or both of pixel replacing in units of m slices in a predetermined vertical range or pixel replacing in units of n consecutive macroblocks within a predetermined horizontal range.
According to the first aspect of the present invention, there is provided a video scramble apparatus comprising a scramble unit which scrambles a video signal, and a coding unit which performs interpicture predictive coding of the video signal scrambled by the scramble unit, wherein the scramble unit selects a picture, which is not used as a reference picture for interpicture prediction in the coding unit, from the video picture signal, and replaces slices as sets of macroblocks located on identical scan lines in the video picture signal of the selected picture in units of m slices which are consecutive in a vertical direction in the picture.
In MPEG2 coding, a picture which is not used as a reference picture means all B-pictures (bi-directional predictive coded pictures), I-pictures (interframe coded pictures) which are not referred to from other frames, and P-pictures (forward predictive coded pictures) wh
Koto Shin-ichiro
Masuda Tadaaki
Barron Gilberto
Stulberger Cas
LandOfFree
Video scramble/descramble apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Video scramble/descramble apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Video scramble/descramble apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3182420