Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-03-09
2003-05-27
Diep, Nhon (Department: 2713)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06570921
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a data-compression assembly having several data-compression channels. The data-compression channels may be, for example, encoders which encode different programs in accordance with a standard of the Moving Pictures Experts Group (MPEG). The encoded programs may be multiplexed so as to form a so-called bouquet. The bouquet may then be transmitted to a receiver which selects a particular program from the bouquet and decodes the selected program.
BACKGROUND ART
The International Application published under number WO 96/20568 describes a video signal encoding apparatus for encoding two or more video signals. The apparatus comprises an encoder stage for each video signal. The encoder stage encodes the video signal as a specified number of data bits per encoded video frame. The encoder stage outputs an encoded video signal at a bit rate determined by the specified number of data bits. A target setting sets specified numbers of data bits for respective encoder stages such that the sum of the respective encoder stage output bit rates remain substantially constant. The respective encoded video signals are written into a buffer memory and are subsequently read from the buffer memory at respectively specified output rates.
The target setting thus provides a joint-bit rate control: a constant overall output bit-rate is maintained despite variations in the individual channel bit rates. The target setting may derive a ratio relating the respective information contents of the video image frames received by the respective encoder stages. The specified number of bits would then be set by dividing a total number of bits available among the channels according to the ratio.
SUMMARY OF THE INVENTION
Let it be assumed that a data compression assembly has several data-compression channels. Let it further be assumed that a data-compression channel comprises a data compressor and a buffer memory space. The data compressor compresses input data so as to obtain compressed data. The buffer-memory space temporarily stores the compressed data and provides the compressed data in the form of an output data stream. The background-art encoding apparatus is an example of such a data-compression assembly.
It is an object of the invention that the data-compression assembly yields a relatively good quality of compressed data. To that end, the invention takes the following aspects into consideration.
Compressed-data quality will generally decrease as the extent to which the data is compressed increases, and vice versa. For example, in MPEG video encoding, the extent to which a picture is compressed depends on a quantization parameter. An encoded picture will comprise relatively few bits if the quantization parameter has a high value. However, the encoded picture will then have a relatively poor resolution. Conversely, an encoded picture have a relatively good resolution if the quantization parameter has a low value. However, the encoded picture will then comprise relatively many bits.
The extent of compression may be controlled, for example, by means of a target. This is done in the background art. A target is a desired number of bits for an encoded picture. The quantization parameter is controlled such that the target is substantially achieved. When encoding a picture, the quantization parameter may need to have a relatively high value in order to achieve the target. In that case, one could say that the picture is difficult to compress. Conversely, the target may be achieved with the quantization parameter having a relatively low value. In that case, one could say that the picture is easy to compress. Thus, supposing that there is a fixed target, a picture which is easy to compress will have a relatively good quality once it has been encoded, whereas a picture which is difficult to compress will have a relatively poor quality.
The background art seems to apply the following principle. There is a total target for each successive set of concurrent pictures which are to be encoded in the different encoding channels. The total target is such that the sum of the respective encoder stage output bit rates remains substantially constant. Thus, the total target should be the same for each set of concurrent pictures. For each set, it is established which pictures are relatively easy to compress and which pictures are relatively difficult to compress. The total target is partitioned among the pictures in the following manner. A picture which is rather difficult to compress is given a relatively high individual target. Conversely, a picture which is relatively easy to compress is given a relatively low individual target. Thus, in a manner of speaking, there is a fixed budget in terms of number of bits to be spend in encoding a set of concurrent pictures. The budget is partitioned such that concurrent compressed pictures have a substantially uniform quality.
The principle applied in the background art has the following inconvenience. It may happen that, at a certain instant, there are relatively many pictures which are difficult to compress whereas, at an other instant, there are relatively few pictures of that kind. Since the total target is constant, pictures which are difficult to compress will have a relatively low individual target at the former instant whereas these pictures will have a relatively high individual target at the latter instant. Consequently, the quality of the compressed pictures will vary in time.
According to the invention, indications of compressed-data quality versus extent of compression are established for respective data-compression channels. A total output bit rate is partitioned over respective output data streams in dependence on these indications. The extent of compression in respective data-compression channels is controlled individually on the basis of output data stream bit rate and amount of compressed data stored in the buffer memory space.
Since, in the invention, the extent of compression in respective data-compression channels is controlled individually, a data compression channel has its own budget in terms of number of bits it can spend for compressing data over a certain future period in time. The budget is determined by the output data stream bit rate and the amount of coded data stored in the buffer memory space. When a data-compression channel receives data which is difficult to compress, the output data stream bit rate of that data-compression channel can be increased. As a result, the budget will increase. How many bits will be spend for encoding a particular data portion is decided by the data-compression channel individually.
Thus, in the invention, the respective encoding channels have flexible budgets and the decision how to spend a budget is locally made. In contrast, in the background art, there is a fixed budget and the decision how to spend the budget is centrally made. Thus, the invention provides greater flexibility in data-compression control. Consequently, the invention allows the respective data-compression channels to better maintain the quality of compressed data at a substantially constant level.
The invention and additional features, which may be optionally used to implement the invention to advantage, are apparent from and will be elucidated with reference to the drawings described hereinafter.
REFERENCES:
patent: 5216503 (1993-06-01), Paik et al.
patent: 5500676 (1996-03-01), Tanaka et al.
patent: 5835149 (1998-11-01), Astle
patent: 5838686 (1998-11-01), Ozkan
patent: 5986712 (1999-11-01), Peterson et al.
patent: 6055270 (2000-04-01), Ozkan et al.
patent: 2001/0055336 (2001-12-01), Krause et al.
patent: 0 643 971 (1995-03-01), None
patent: 0 685 971 (1995-12-01), None
Limin Wang and Andre Vincent; Joint Rate Control For Multi-Program Video Coding; Aug. 1996, IEEE, vol. 42, pp. 300-305.
Bailleul Nicolas
Fert Etienne
Gautier Pierre
Grandchamp Myriam
Diep Nhon
Gross Russell
Koninklijke Philips Electronics , N.V.
LandOfFree
Data compression assembly having several data-compression... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Data compression assembly having several data-compression..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Data compression assembly having several data-compression... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3046380