Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1999-04-01
2001-11-20
Le, Vu (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240120
Reexamination Certificate
active
06320908
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to moving pictures decompression device and method for decompressing compressed moving picture data, and in particular, the invention relates to such moving pictures decompression device and method capable of effectively accessing to memory in case when the data is to be processed by software.
DESCRIPTION OF THE RELATED ART
Regarding a system for compressing and decompressing data in case when high-quality moving picture data is to be stored or used in communication, there is an internationally standardized system called MPEG (moving picture coding experts group). In order to execute a real-time decompression of moving picture data, which is compressed on the basis of the MPEG standard, an enormous amount of computation would be required. Therefore, it has been typical that dedicated designed hardware for this purpose is used. However, as microprocessors have become highly efficient these days, such data compression or decompression is expected to become possible only with an application of software to the ordinary microprocessor without using special hardware.
Algorithm for data compression and decompression in the MPEG, for example, is described in “Communications of the ACM”, (Vol. 34, No. 4, April 1991). In rough definition, the decompression algorithm of the MPEG contains a decoding process of variable length codes, an inverse quantization process, an inverse discrete cosine transformation process, and a motion compensation process. Among those processes, the motion compensation process is a kind of process which uses temporal redundancy of the moving picture, and thus restores a picture by cutting off a part of a picture from a picture frame being previously decompressed. The position where the part of picture is cut off from is determined by a parameter called a motion vector. The picture which is to be cut off and restored is a square range called macro block which contains 16 by 16 pixels. A value for each pixel is defined as 8 bit.
As to a conventional moving pictures decompression device, the one shown in
FIG. 1
is known. The moving pictures decompression device comprises a CPU
301
, an internal memory
302
, a display device
303
, an external storage
304
, and a cache memory
305
. A bit stream compressed by the MPEG is to be stored in the external storage
304
. Then the bit stream is read out from the external storage
304
, and temporarily stored in the internal memory
302
. The bit stream stored in the internal memory
302
is decompressed by an operation of decoding software which operates on the CPU
301
, and as a result, picture data is stored in the internal memory
302
. The picture data being stored in the internal memory
302
is transformed into a display format by the software on the CPU
301
to be transmitted to the display device
303
.
FIG. 2A
, FIG.
2
B and
FIG. 2C
are a set of diagrams showing an order of picture frames when conducting compression and decompression by the MPEG. For encoding a picture of a certain frame, the MPEG uses three kinds of pictures, i.e. I-picture, P-picture, and B-picture. I-picture can be defined as a picture which executes compression without the execution of a motion compensation but only with a use of information within the frame. P-picture can be defined as a picture which executes a motion compensation process by forecasting on the basis of picture information of a frame in the past. B-picture can be defined as a picture which executes a motion compensation process by forecasting on the basis of picture information of both frames before and beyond the frame.
FIG. 2A
shows a typical distribution of picture types at a time of compressing a moving picture by the MPEG. Frame #
1
is I-picture, which is compressed without referring to other frames. Frame #
2
is B-picture, which is compressed by a bidirectional motion forecasting process using frame #
1
and frame #
3
. Frame #
3
is P-picture, which is compressed by a motion forecasting process using frame #
1
. Frame #
4
is B-picture, which is compressed by a bidirectional motion forecasting process using frame #
3
and frame #
5
. Frame #
5
is P-picture, which is compressed by a motion forecasting process using frame
43
. Frame #
6
is B-picture, which is compressed by a bidirectional motion forecasting process using frame #
5
and frame #
7
. Frame #
7
is P-picture, which is compressed by a motion forecasting process using frame #
5
. The arrows of broken lines indicate reference directions.
FIG. 2B
shows processing order of the frames when conducting compression and decompression on the basis of the picture types as indicated above. In a compression process, first, frame #
1
is compressed, and then frame #
3
is compressed by referring to a picture of frame #
1
. Next frame #
2
is compressed by referring to pictures of frame #
1
and frame #
3
. Next, frame #
5
is compressed by referring to the picture of frame #
3
. Next, frame #
4
is compressed by referring to pictures of frame #
3
and frame #
5
. Next, frame #
7
is compressed by referring to the picture of frame #
5
. Then frame #
6
is compressed by referring to pictures of frame #
5
and frame #
7
. On the MPEG bit stream where the compressed pictures are written, information of each frame is written in the order of being compressed. Therefore, the order of the frame information would be different from the arrangement order of the original moving picture frames.
In decompressing the MPEG bit stream which is compressed in the above manner, first, frame #
1
is decompressed. Then frame #
3
is decompressed by referring to a decompressed picture of frame #
1
. Then frame #
2
is decompressed by referring to decompressed pictures of frame #
1
and #
3
. Then, frame #
5
is decompressed by referring to the decompressed picture of frame #
3
. Then frame #
4
is decompressed by referring to decompressed pictures of frame #
3
and frame #
5
. Then frame #
7
is decompressed by referring to the decompressed picture of frame #
5
. Then frame #
6
is decompressed by referring to decompressed pictures of frames #
5
and #
7
.
FIG. 2C
shows display order of the decompressed pictures. In conducting decompression and display operations using such moving pictures decompression device, it becomes necessary to display the pictures by rearranging the decompressed frames. Therefore, the decompression and display processing will have to be done in the following order. After decompressing frame #
1
, nothing will be displayed. Then after frame #
2
is decompressed, frame #
2
is displayed. Then after frame #
5
is decompressed, frame #
3
is displayed. The further procedure is self evident in the figure.
An operation of decoding software will be explained with reference to a flow chart in FIG.
3
. In a decompression process over a single frame, a start of frame decoding is confirmed (step S
401
). Variable length code decoding (step S
402
), inverse quantization (step S
403
), inverse DCT (step S
404
), and motion compensation (step S
405
) are repeated with respect to each macro block. When the last macro block of the frame is being processed (step S
406
), the decompression process of the frame is over (step S
407
). After that, picture data of the frame is transmitted to the display device
303
in the above-mentioned order (step S
408
).
In the CPU
301
of the moving pictures decompression device shown in
FIG. 1
, the cache memory
305
is built in. Compared to the internal memory
302
, the cache memory
305
has a smaller capacity and is capable of faster access. The CPU
301
controls the cache memory
305
such that it stores recently accessed data contents, and discharge the ones which have not been accessed for a long time since last being accessed. The CPU
301
then controls so that the discharged data contents are written back in
Foley & Lardner
Le Vu
NEC Corporation
LandOfFree
Moving pictures decompression device and method with reduced... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Moving pictures decompression device and method with reduced..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Moving pictures decompression device and method with reduced... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2572720