Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-05-09
2001-11-20
Lee, Young (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06320909
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a picture decoding/display unit which decodes encoded picture data and outputs decoded pixel data for display on a display unit, and more particularly, it relates to a picture decoding and display unit which decodes predictively encoded moving picture data and outputs the decoded data for display. More specifically, the present invention relates to a structure for reducing the storage capacity of a storage element storing B pictures in decoding along the MPEG (moving picture experts group) standard.
2. Description of the Background Art
In transmission of an enormous quantity of picture data, the data quantity is reduced by decoding the picture data in high efficiency. Techniques of such high efficiency encoding include an inter-frame (or inter-field) predictive encoding system. This inter-frame (or inter-field) predictive encoding, which utilizes temporal correlation between pictures, is generally employed for encoding moving picture data.
The inter-frame (or inter-field) predictive encoding includes the following procedure: A screen picture (hereinafter simply referred to as “picture”) which is temporally approximate to a current screen picture (frame or field) to be encoded is employed as a reference picture for predicting values of the pixel data of the current picture. Differences (prediction errors) between the predicted values and the pixel data of the current picture are obtained, and then the prediction errors are encoded. The predicted values are formed by data of reference picture pixels (generally pixels of a motion-compensated reference picture in moving picture encoding) corresponding to the current picture pixels to be encoded. When the pictures have high temporal (time) correlation, the difference (prediction error) is small and hence the quantity of data to be transmitted can be reduced to implement effective compression of information.
Systems of such predictive encoding include the MPEG standard which is directed to moving pictures. Systems of encoding and decoding picture data along the MPEG standard, which are explained in Nikkei Electronics, Mar. 14, 1994, pp. 82 to 116, for example, are now briefly described.
Pictures include I, P and B pictures, and the predictive system as employed depends on the type of the pictures.
FIG. 27
illustrates exemplary temporal (timewise) arrangement of pictures. This figure representatively shows nine pictures G
1
to G
9
. The picture G
3
is an I picture, which is subjected to intra-frame or intra-field encoding so that its pixel data themselves are encoded. The pictures G
6
and G
9
are P pictures, which are subjected to inter-frame or inter-field predictive encoding through past reproduced pictures. The pictures G
1
, G
2
, G
4
, G
5
, G
7
and G
8
are B pictures, which are predictively encoded through either past or future reproduced pictures or both of these pictures. These B pictures are predictively encoded through combinations of the I and P pictures or of the P pictures. The B pictures are not employed as reference pictures in predictive encoding and decoding. Referring to
FIG. 27
, the B pictures G
1
and G
2
are predictively encoded through the future I picture G
3
which is temporally subsequent thereto. The B pictures G
4
and G
5
are predictively encoded through both of the I picture G
3
which is a past reproduced picture and the P picture G
6
which is a future reproduced picture temporally subsequent thereto. The B pictures G
7
and G
8
are predictively encoded through the P pictures G
6
and G
9
which are past and future reproduced pictures respectively.
FIG. 28
illustrates the structure of a single picture. Referring to
FIG. 28
, the single picture (field or frame)
1020
is divided into a plurality of segments called macro blocks. For the purpose of simplification,
FIG. 28
illustrates such an exemplary structure that the picture
1020
is divided into 32 macro blocks MB#
1
to MB#
32
. Processing of a moving picture is generally executed in units of the segments called macro blocks, in both encoding and decoding. Each of the macro blocks MB#
1
to MB#
32
generally includes 256 pixels arranged in 16 rows and 16 columns, as shown in FIG.
28
. Therefore, the picture
1020
shown in
FIG. 28
is formed by 128 by 64 pixels. The MPEG standard defines such conditions that the picture is structured by not more than 720 pixels per line (scanning line), not more than 576 lines per frame, and not more than 30 frames per second. However, DCT (discrete cosine transformation) processing and IDCT (inverse discrete cosine transformation) processing are performed in units of blocks of 8 by 8 pixels.
FIG. 29
schematically illustrates the structure (syntax) of a bit stream (a plural bit width) of picture data along the MPEG standard. Referring to
FIG. 29
, the bit stream is divided into a plurality of layers including a sequence layer, a GOP (group of pictures) layer, a picture layer, a slice layer, a macro block layer and a block layer in the order from the uppermost layer.
The block layer is formed by a block
1100
including a region
1100
a
including data of DCT coefficients and a region
1100
b
storing an end of block (EOB) data indicating the end of the block. The region
1100
a
stores DCT coefficient data of pixels of 8 rows and 8 columns serving as a unit of the DCT processing. When the last AC coefficient of the region
1100
a
is a nonzero coefficient, the end of block EOB of the region
1100
b
may not be employed in the block
1100
. The DCT processing is performed to reduce spatial redundancy (high correlationship between adjacent pixels) in the picture thereby reducing the picture data quantity in encoding. DCT coefficients can be made localized in a low frequency coefficient region by the DCT processing. It is possible to enlarge values of low frequency components and substantially zeronize values of high frequency components in both of horizontal and vertical directions in space frequencies by quantizing the DCT coefficients. Thus, the picture data quantity is reduced.
The macro block layer is formed by a block
1110
including a prescribed number of (six) blocks
1100
and a macro block header
1115
storing attributes of the data of the macro blocks, motion vectors and the like.
The slice layer includes a slice
1120
which is formed by one or a plurality of macro blocks which are concatenated in picture scanning order. A slice header
1125
storing information indicating the vertical position of the slice on the screen and information such as a start code having a prescribed pattern indicating starting of this slice is provided at the head of the slice
1120
.
The picture layer includes a picture
1130
which is formed by a plurality of slices
1120
. A picture header
1135
storing information indicating the type (I, P or B picture) of the picture and a start code indicating starting of the picture is arranged at the head of the picture
1130
.
The GOP layer includes a GOP
1140
including a plurality of pictures
1130
. The pictures
1130
included in the GOP
1140
include at least one I picture and zero or a plurality of P or B pictures. A GOP header
1145
storing a start code indicating starting of the GOP and information such as a flag indicating that this GOP requires no reference from picture data of a GOP which is precedent thereto is arranged at the head of the GOP
1140
.
The sequence layer includes a sequence
1150
which is formed by one or a plurality of GOPs
1140
each including one or a plurality of pictures
1130
. A sequence header
1155
storing information such as the format of the screen is arranged at the head of the sequence
1150
. This sequence header
1155
can be arranged at the head of every GOP
1140
which is included in the sequence
1150
, in order to allow reproduction of pictures from an intermediate portion of the sequence. The sequence header
1155
stores information such as a start code having a prescribed pattern indicating starting of the se
Takabatake Akihiko
Uramoto Shin-ichi
Lee Young
McDermott & Will & Emery
Mitsubishi Denki & Kabushiki Kaisha
LandOfFree
Picture decoding and display unit including a memory having... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Picture decoding and display unit including a memory having..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Picture decoding and display unit including a memory having... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2586794