Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1998-01-14
2001-02-20
Kelley, Chris S. (Department: 2713)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06192077
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to a coding apparatus that efficiently encodes picture data to digital signals with less code amounts by motion compensation to effectively transmit, store and display pictures and a decoding apparatus for decoding the picture data.
There is a prediction method in moving picture efficient coding that displaces, prior to prediction, portions of a picture per block in accordance with the motion of the picture in inter-picture prediction.
This picture spatial displacement is called motion compensation and has widely been employed in international standard system, such as MPEG.
When the motion compensation is applied in the inter-picture prediction, picture motion information (motion vector) is also coded. Because motion compensation the same as that in coding must be done in decoding.
FIG. 1
shows a block diagram of a conventional coding apparatus with motion compensation.
In
FIG. 1
, an image (picture) signal is supplied to a subtracter
2
via input terminal
1
. The subtracter
2
subtracts an inter-picture predictive signal from the input image signal. The inter-picture predictive signal is supplied by a motion compensated predictor
6
. The subtracter
2
then generates an inter-picture predictive residue signal that is supplied to a discrete cosine transformer (DCT)
3
.
The discrete cosine transformer
3
processes the inter-picture predictive residue signal by discrete cosine transforming per 8×8 pels to generate coefficients that are supplied to a quantizer
4
. The quantizer
4
quantizes the coefficients by a predetermined quantization step width to generate fixed-length coded coefficients. The fixed-length coded coefficients are supplied to an arrangement converter
5
.
The arrangement converter
5
rearranges two-dimensional 8×8 coefficients in one-dimension in Zigzag scanning order. The coefficients rearranged in one-dimension are supplied to a variable-length encoder
10
.
The variable-length encoder
10
encodes a run length of zeros and other values of the coefficients by Huffman coding to generate a bitstream.
The bitstream of the inter-picture predictive residue is multiplexed by a multiplexer
15
with a bitstream of motion vectors and output via output terminal
16
.
The fixed-length coded coefficients are also supplied to a dequantizer
9
and then to an inverse-discrete cosine transformer (IDCT)
8
. The dequantizer
9
and the inverse-discrete cosine transformer
8
perform the reverse processing of the quantizer
4
and the discrete cosine transformer
3
, respectively, to reproduce the inter-picture predictive residue signal that is supplied to an adder
7
.
The adder
7
adds the inter-picture predictive residue signal and the inter-picture predictive signal to reproduce the image signal that is supplied to the motion compensated predictor
6
.
The motion compensated predictor
6
stores one frame picture and displaces the picture per block in accordance with the motion vectors supplied from a motion estimator
12
to generate a motion-compensated inter-picture predictive signal. The inter-picture predictive signal is supplied to the subtracter
2
and the adder
7
.
The motion vectors are obtained per block from the input image signal by the motion estimator
12
. Each block is composed of the array in the range of (8×8) and (16×16) pels. The motion vectors thus obtained are also supplied to a predictive subtracter
13
.
The predictive subtracter
13
compares vector values of an already coded block and a block to be coded per “x” and “y” components to generate residue values that are supplied to a variable-length encoder
14
.
The variable-length encoder
14
encodes the residue values by Huffman coding to generate a bitstream.
The generated bitstream is multiplexed with the bitstream of the inter-picture predictive residue signal by the multiplexer
15
as described above.
FIG. 2
shows a block diagram of a conventional decoding apparatus with motion compensation.
The coded signal output from the coding apparatus of
FIG. 1
is supplied to a demultiplexer
18
via input terminal
17
to be divided into the bitstreams of the inter-picture predictive residue signal and the motion vectors.
The bitstream of the inter-picture predictive residue signal is decoded by a variable-length decoder
19
to be the fixed-length codes.
The fixed-length codes are supplied to a reverse arrangement converter
20
that performs the reverse processing of the arrangement converter
5
of
FIG. 1
to obtain (8×8) array of coefficients.
The coefficients are supplied to a dequantizer
21
and then to an inverse-discrete cosine transformer (IDCT)
22
to reproduce the inter-picture predictive residue signal.
The inter-picture predictive residue signal is supplied to an adder
23
and added to an inter-picture predictive signal to reproduce the image signal.
The reproduced image signal is output via output terminal
24
and supplied to a motion compensated predictor
27
.
The motion compensated predictor
27
generates the inter-picture predictive signal in accordance with the motion vectors supplied from a predictive adder
26
.
The bitstream of the motion vectors divided by the demultiplexer
18
is supplied to a variable-length decoder
25
that performs the reverse processing of the variable-length encoder
14
of
FIG. 1
to obtain the residue values of the motion vectors.
The residue values are supplied to the predictive adder
26
where vector values of an already decoded block are added to the residue values to obtain motion vectors. The motion vectors are supplied to the motion compensated predictor
27
as described above.
The conventional coding apparatus described above has the following drawbacks:
Since motion vectors do not involve change, a code is generated per motion vector even if a residue value is zero. Further, when motion vectors gradually change in a zoomed scene, many codes are generated even if the change is monotonous.
Therefore, the conventional coding apparatus cannot compress the generated code amount effectively. Further, the conventional coding apparatus must be provided with motion compensation circuitry because the motion compensation is executed by a method completely different from that of inter-picture predictive residue coding.
SUMMARY OF THE INVENTION
A purpose of the invention is to provide a coding apparatus that effectively encodes moving picture data with motion compensation circuitry that can also be used to encode the inter-picture predictive residue.
Another purpose of the invention is to provide a decoding apparatus for decoding the moving picture data coded by the above coding apparatus.
The present invention provides a coding apparatus with motion compensation comprising: a unifier to unify a plurality of motion vectors corresponding to motion of portions of a picture to obtain a first motion vector group arranged in two-dimensional block; an arrangement converter to convert motion vectors in the first motion vector group into one-dimension to obtain a second motion vector group arranged in one-dimension; a vector predictor to predict a motion vector value in the second motion vector group from another motion vector in the second motion vector group to obtain a motion vector predictive residue; and a encoder to encode the motion vector predictive residue with variable-length codes including zero run length codes to obtain a bitstream per motion vector group.
Further, the present invention provides a decoding apparatus with motion compensation comprising: a decoder to decode a bitstream of motion vector groups composed of variable-length codes including zero run length codes to obtain a fixed-length motion vector predictive residue; an adder to adds a predictive value obtained from another motion vector to the fixed-length motion vector predictive residue to obtain motion vectors as a one-dimensionally arranged motion vector group; an arrangement converter to convert the one-dimensionally arranged motion vector group into two-dimension to obtain a
Jacobson Price Holman & Stern PLLC
Kelley Chris S.
Victor Company of Japan Ltd.
Wong Allen
LandOfFree
Coding and decoding apparatus with motion compensation does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Coding and decoding apparatus with motion compensation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Coding and decoding apparatus with motion compensation will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2585146