Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1998-07-06
2001-09-18
Rao, Andy (Department: 2713)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240110, C375S240050
Reexamination Certificate
active
06292512
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is directed to a scalable video coding system which codes video data using both frame-prediction and fine-granular scalable images. The invention has particular utility in connection with variable-bandwidth networks and computer systems that are able to accommodate different bit rates, and hence different quality images.
2. Description of the Related Art
Scalable video coding in general refers to coding techniques which are able to provide different levels, or amounts, of data per frame of video. Currently, such techniques are used by lead video coding standards, such as MPEG-2 and MPEG-4 (i.e., “Motion Picture Experts Group” coding), in order to provide flexibility when outputting coded video data.
In the scalable coding techniques currently employed by MPEG-2 and MPEG-4, an encoder codes frames of video data and divides the coded frames into a base layer (“BL”) and an enhancement layer (“EL”). Typically, the base layer comprises a minimum amount of data required to decode the coded video data. The enhancement layer, on the other hand, comprises additional information which enhances (e.g., improves the quality of) the base layer when it is decoded. In operation, the encoder transmits all frames from the base layer to a receiving device, which can be a personal computer or the like. However, the encoder only transmits frames from the enhancement layer in cases where the receiving device has sufficient processing power to handle those additional frames and/or the medium over which the frames are transmitted has sufficient bandwidth.
FIGS. 1 and 2
show “scalability structures” which are currently used in MPEG-2 and MPEG-4 for the base layer and the enhancement layer. More specifically,
FIG. 1
shows a scalability structure
1
which employs frame-prediction in base layer
2
to generate predicative (or “P”) frames from an intra (or “I”) frame or from a preceding P frame. As shown in the figure, frame-prediction is also used in the enhancement layer to generate P frames based on frames in the base layer.
FIG. 2
shows another scalability structure
3
which is currently used in MPEG-2 and MPEG-4. In the scalability structure shown in
FIG. 2
, frame-prediction is again employed to determine P frames in the base layer. Unlike scalability structure
1
, however, scalability structure
3
also uses frame-prediction in the enhancement layer to generate bi-directional (or “B”) frames which, in this case, are interpolated from preceding frames in the enhancement layer and contemporaneous frames in the base layer. In general, MPEG-2 and MPEG-4 encoders use frame prediction in the manner set forth above to increase data compression and thus increase coding efficiency.
Another well-known scalable video coding technique is called fine-granular scalability coding. Fine-granular scalability coding codes the same image (e.g., a frame of video) using progressively more data each time coding takes place. For example, as shown in
FIG. 3
, image
4
is initially encoded using data sufficient to produce image
5
. Thereafter, additional data is coded which is sufficient to produce enhanced images
6
,
7
and
8
in succession.
Fine-granular scalability coding has several advantages over the frame-prediction techniques described above. Specifically, because fine-granular scalability coding can provide a wider range of enhanced images than frame-prediction techniques, fine-granular scalability coding is generally preferred in environments, such as the Internet, which have a wide range of available bandwidth. For similar reasons, fine-granular scalability coding is also generally preferred when dealing with receiving devices that have varying processing capabilities and/or bandwidth. That is, because fine-granular scalability coding produces a wide range of enhanced images, it is possible to match the appropriate image relatively closely to an amount of available bandwidth. As a result, in theory, it is possible to obtain the most amount of data for an image for a given amount of available bandwidth. On the down-side, fine-granular scalability coding does not permit the use of frame-prediction. As a result, it requires more data than the frame-prediction techniques described above and, consequently, degrades coding efficiency.
Thus, there exists a need for a scalable video coding technique which incorporates the efficiency of frame-prediction coding and the accuracy of fine-granular scalability coding.
SUMMARY OF THE INVENTION
The present invention addresses the foregoing need by coding a portion (e.g., a base layer) of input video data using a frame-prediction coding technique and then coding another portion (e.g., residual images in an enhancement layer) of the video data using fine-granular scalability coding. By coding a base layer using a frame-prediction coding technique, the present invention reduces the amount of bits required to code the video data and thus maintains coding efficiency. By coding the residual images using fine-granular scalability coding, the present invention is able to provide a wide range of residual images, one or more of which can be selected for transmission based, e.g., on an available bandwidth of a receiving device.
Thus, according to one aspect, the present invention is a system (i.e., a method, an apparatus, and computer-executable process steps) for coding video data comprised of one or more frames. The system codes a portion (e.g., a base layer) of the video data using a frame-prediction coding technique, and then generates residual images based on the video data and the coded video data. Thereafter, the system codes the residual images using a fine-granular scalability coding technique, and outputs the coded video data and at least one of the coded residual images to a receiver, such as a variable-bandwidth network or a networked device thereon.
In preferred embodiments of the invention, the system determines a bandwidth of the receiver, and then selects which of the coded residual images to output based on the bandwidth of the receiver. By doing this, the invention is able to output a coded residual image which is most appropriate for the available bandwidth.
In other preferred embodiments, the system codes the portion of the video data at a plurality of different bit rates so as to produce multiple versions of the coded video data, and generates a plurality of residual images for each version of the coded video data. In these embodiments, the system codes the residual images using a fine-granular scalability coding technique, determines variations in a bandwidth of the receiver over time, and then selects which one of the multiple versions and the coded residual images to output based on the variations in the bandwidth of the receiver.
By way of example, for a receiver bandwidth increasing from B
1
to B
2
, where B
1
<B
2
, the system selects a first version of the coded video data and successively selects coded residual images corresponding to each frame of the first version of the coded video data, which are coded at successively higher bit rates. For a receiver bandwidth increasing from B
2
to B
3
, where B
2
<B
3
, the system selects a second version of the coded video data and successively selects coded residual images corresponding to each frame of the second version of the coded video data, which are coded at successively higher bit rates. Conversely, for a receiver bandwidth decreasing from B
3
to B
2
, where B
3
>B
2
, the system selects a first version of the coded video data and successively selects coded residual images corresponding to each frame of the first version of the coded video data, which are coded at successively lower bit rates. Likewise, for a receiver bandwidth decreasing from B
2
to B
2
, where B
2
>B
1
, the system selects a second version of the coded video data and successively selects coded residual images corresponding to each frame of the second version of the coded video data, which are coded at successively lower bit rates.
As is
Chen Yingwei
Cohen Robert A.
Radha Hayder
Gross Russell
Rao Andy
U.S. Philips Corporation
LandOfFree
Scalable video coding system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Scalable video coding system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Scalable video coding system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2538764