Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1999-08-11
2003-09-30
Kelley, Chris (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240250
Reexamination Certificate
active
06628719
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an MPEG video decoder for decoding a bit-stream which has been encoded according to a Moving Picture Expert Group (hereinafter, referred to as an MPEG) standard, and an MPEG video decoding method.
The MPEG standard has been heretofore employed as an international standard for a bit rate reduction of picture data. A moving picture encoding technology and a moving picture decoding technology, which are in conformity with the MPEG standard, are inevitable for recent multi-media environments. Many moving picture encoding apparatus and moving picture decoding apparatus which adopt the MPEG standard have been developed.
A bit rate reduction encoding by the MPEG standard includes a constant bit rate encoding (hereinafter, referred to as a CBR encoding) in which a data rate after encoding is substantially constant and a variable bit rate encoding (hereinafter, referred to as a VBR encoding) in which the data rate after encoding is not constant.
The CBR encoding is used when a transmission line is in a STM mode (Synchronous Transfer Mode). The VBR encoding is used when the transmission line in an ATM mode (Asynchronous Transfer Mode), as well as when the transmission line is storage media such as a DVD (Digital Video Disc).
Descriptions for a picture type, a bit-stream, a MPEG video encoder and a MPEG video decoder, which are used in the MPEG standard, will be made below.
(1) Picture Type
In the MPEG standard, used are three types of pictures for a high efficiency encoding, which are called an intra coded picture (hereinafter, referred to as an I picture), a predictive coded picture (hereinafter, referred to as a P picture) and a bidirectional predictive coded picture (hereinafter, referred to as a B picture), respectively.
The I picture does not use information relating to other pictures, and encoded by only information relating to itself. The P picture is encoded by using the I picture or the P picture in the past as a reference picture. The B picture is encoded using the I picture or the P picture in the past and future.
Although the I picture exhibits a low compression rate, the I picture can be encoded independently from other pictures, so that the I picture is utilized as an access point in random accessing. For decoding the I picture, information relating to other pictures is unnecessary. Although the P picture exhibits a compression rate higher than that of the I picture, the P picture requires information relating to the I picture in the past for its decoding. Moreover, although the B picture exhibits the highest compression rate in all of the pictures, the B picture needs the information relating to the I or P picture for its decoding.
(2) Bit-stream
FIG. 1
is a schematic view showing a structure of the bit-stream adopting the MPEG standard. The bit-stream consists of a sequence header
31
, a GOP header (a group of picture header)
32
, a picture header
33
, picture data
34
and a sequence end code
35
. The bit-stream includes a sequence extension and extension and user data in addition to these components. Illustrations and descriptions for them are omitted here.
The sequence header
31
always exists at the initial portion of the bit-stream. In the sequence header
31
, included are a horizontal size value, a vertical size value and a parameter showing an aspect ratio information.
The GOP header
32
is added to the bit-stream when many pictures are managed for each GOP. The GOP consists of a plural types of pictures, and the first picture of the GOP is always the I picture. The GOP is inevitable in a MPEG
1
(Moving Picture Experts Group Phase
1
), and optional in a MPEG
2
(Moving Picture Experts Group Phase
2
).
The picture head
33
indicates the beginning of decoded data for one picture. A parameter (a temporal reference) indicating the order of the pictures and a picture type are included in the picture head
33
.
The picture data
34
is encoded one for one picture. The picture data
34
is followed by the GOP header
32
, the subsequent picture head
33
or the sequence end code
35
. The sequence end code
35
indicates the end of the bit-stream.
(3) Structure of MPEG Video Encoder
FIG. 2
is a block diagram showing a MPEG video encoder.
The MPEG video encoder consists of a picture rearrangement section
41
, a motion estimation section
42
, an adder
43
, a discrete cosine transform (hereinafter, referred to as a DCT) section
44
, a quantization section
45
, a variable length code section
46
, a multiplexer section
47
, a buffer
48
, an inverse quantization section
49
, an inverse discrete cosine transform (hereinafter, referred to as an IDCT) section
50
, an adder
51
, a picture storage section
52
and a motion prediction section
53
.
Since the B picture decoded by referring to the pictures in the past and future exists in the MPEG, it is necessary to process the picture in the future prior to the picture in the past. The picture rearrangement section
41
rearranges the pictures in the processing order.
The motion estimation section
42
receives the picture from the picture rearrangement section
41
, and outputs various kinds of parameters required in decoding, such as a picture type, a presentation time stamp (PTS), a quantization step size, a motion vector and an encoded mode. These parameters are supplied to the motion prediction section
53
, and added also to the bit-stream in the multiplexer section
47
.
The adder
43
calculates the differential between the picture outputted from the motion estimation section
42
and the reference picture outputted from the motion prediction section
53
. Since the reference picture is not outputted from the motion prediction section
53
at the time when the I picture is processed, the picture outputted from the motion estimation section
42
is inputted to the DCT section
44
via the adder
43
. Moreover, when either the P picture or the B picture is processed, the adder
43
calculates the differential between the picture outputted from the motion estimation section
42
and the reference picture outputted from the motion prediction section
53
, and outputs the differential to the DCT section
44
.
The DCT section
44
performs a discrete cosine transform for the data supplied from the adder
43
, divides the data into each of frequency components, and removes high frequency components, thus reducing the quantity of the data. The quantization section
45
removes information with less importance by the quantization that is a non-reversible step. The variable length code section
46
zigzag-scans the quantized data so as to perform a variable encoding for the quantized data, thus further reducing the quantity of data.
On the other hand, the inverse qunatization section
49
and the IDCT section
50
performs an inverse discrete cosine transform and an inverse quantization for the data that has been subjected to the discrete cosine transform by the DCT section
44
and the quantization by the quantization section
45
. The adder
51
adds the data outputted from the IDCT section to the reference picture outputted from the motion prediction section
53
so as to restore the picture, thus storing the restored picture in the picture storing section
52
. The motion prediction section
53
performs a motion prediction based on the motion vector outputted from the motion estimation section
42
, the output from the adder
51
, and the picture stored in the picture storing section
52
. Furthermore, the motion prediction section
53
creates a reference picture from either the output of the adder
51
or the picture stored in the picture storing section
52
, and outputs the reference picture to the adder
43
.
The multiplexer section
47
multiplexes the data outputted from the variable length code section
46
and the parameters such as the encoded mode and the motion vector outputted from the motion estimation section
42
, thus creating the bit-stream. The buffer
48
temporarily stores the bit-stream created.
(4) Structure of MPEG Vide
Inagaki Hirohiko
Kono Tadayoshi
Ota Mitsuhiko
Kelley Chris
Parsons Charles
Staas & Halsey , LLP
LandOfFree
MPEG video decoder and MPEG video decoding method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with MPEG video decoder and MPEG video decoding method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and MPEG video decoder and MPEG video decoding method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3071448