Image analysis – Image compression or coding – Adaptive coding
Reexamination Certificate
1996-04-03
2002-07-02
Johnson, Timothy M. (Department: 2623)
Image analysis
Image compression or coding
Adaptive coding
C382S103000, C382S282000
Reexamination Certificate
active
06415057
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a method and apparatus for picture compression which may be used with advantage for compressing a picture.
2. Description of the Related Art
FIG. 1
shows a conventional arrangement of a picture compression apparatus employed with advantage for encoding a picture for compression.
In the picture compression apparatus, shown in
FIG. 1
, picture data digitized with the following numbers of pixels are supplied to an input terminal
1
, as shown in FIG.
2
: luminance components (Y): 352(H)×240(V)×30 frames chroma components (Cb): 176(H)×120(V)×30 frames chroma components (Cr): 176(H)×120(V)×30 frames
The input picture data, supplied to the input terminal
1
, are sent to a motion detector
20
and a block divider
11
via a frame memory
10
designed to store the input picture data transiently and to interchange the data sequence appropriately.
The block divider
11
divides each frame supplied from the frame memory
10
into blocks of 8×8 pixels for luminance components Y and chroma components Cr, Cb, as shown in FIG.
3
. The four blocks Y
0
, Y
1
, Y
2
and Y
3
of luminance components Y, one block of chroma components Cb and one block Cr of chroma components, totalling six blocks, are termed a macro-block MB.
The macro-block based data from the block divider
11
is sent to a subtractor
12
.
The subtractor
12
finds a difference between data from the block divider
11
and inter-frame prediction picture data as later explained and transmits a difference output to a fixed contact b of a changeover switch
13
as frame data to be encoded by inter-frame predictive coding as later explained. To the opposite side fixed contact a of the changeover switch
13
are supplied data from the block divider
11
as frame data to be intra-coded as later explained.
The block-based data from the changeover switch
13
are discrete-cosine-transformed by a DCT circuit
14
to produce DCT coefficients which are then supplied to a quantizer
15
. The quantizer
15
quantizes the DCT output at a pre-set quantization step width to produce quantized DCT coefficients (quantized coefficients) which are then supplied to a zigzag scan circuit
16
.
The zigzag scan circuit
16
re-arrays the quantized coefficients by zig-zag scan as shown in FIG.
4
. The resulting output is supplied to a variable length encoding circuit
17
. The variable length encoding (VLC) circuit
17
variable length encodes the output data of the zigzag scan circuit
16
and sends the VLC output to an output buffer
18
. The variable length encoding circuit
17
also sends the information specifying the code amount from the variable length,encoding circuit
17
to a quantization step controller
19
. The quantization step controller
19
controls the quantization step width of the quantizer
15
based upon the information specifying the code amount from the variable length encoding circuit
17
. Output data of the output buffer
18
is outputted at an output terminal
2
as compressed encoded data.
An output of the quantizer
15
is dequantized by a dequantizer
27
and further inverse discrete-cosine-transformed by an inverse DCT circuit
26
. An output of the inverse DCT circuit
26
is supplied to an adder
25
.
To the adder
25
is supplied inter-frame prediction picture data from a motion compensator
21
via a changeover switch
24
turned on for a frame for inter-frame predictive coding. The inter-frame prediction picture data is added by the adder
25
to output data of the inverse DCT circuit
26
. Output data of the adder
25
is transiently stored in a frame memory
22
and thence supplied to the motion compensator
21
.
The motion compensator
21
performs motion compensation based upon the motion vector detected by the motion detector
20
and outputs the resulting inter-frame prediction picture data.
An illustrative operation of the conventional picture compression apparatus shown in
FIG. 1
is explained in detail. For explanation sake, the following appellations of the respective frames are used.
First, the frames arrayed in the display sequence are termed I
0
, B
1
, B
2
, P
3
, B
4
, B
5
, P
6
, B
7
, B
8
, I
9
, B
10
, B
11
, B
12
, . . . Of these frames, I, P and B specify the methods of compression, as later explained, and the numerical figures affixed to I, P and B simply specify the display sequence.
Of the Moving Picture Expert Group (MPEG), a work group for international standardization of the color moving picture encoding system, the MPEG
1
provides the following for compressing the above pictures.
First, the picture I
0
is compressed.
Next, in compressing the picture P
3
, difference data between P
3
and I
0
is compressed in place of the picture P
3
itself.
Next, in compressing the picture B
1
, difference data between B
1
and I
0
, difference data between B
1
and P
3
or difference data between B
1
and a mean value of I
0
and P
3
, whichever is smallest in information amount, is compressed in place of the picture B
1
itself.
Next, in compressing the picture B
2
, difference data between B
2
and I
0
, difference data between B
2
and P
3
or difference data between B
2
and a mean value of I
0
and P
3
, whichever is smallest in information amount, is compressed in place of the picture B
2
itself.
Next, in compressing the picture P
6
, difference data between P
6
and P
3
is compressed in place of the picture P
6
itself.
The following is the representation of the above processing
counterpart picture(s) for
picture being processed
difference taking
(1)
I0
—
(2)
P3
I0
(3)
B1
I0 or P3
(4)
B2
I0 or P3
(5)
P6
P3
(6)
B4
P3 or P6
(7)
B5
P3 or P6
(8)
P9
P6
(9)
B7
P6 or P9
(10)
B8
P6 or P9
(11)
I9
—
(12)
P12
I0
(13)
B10
I9 or P12
(14)
B11
I9 or P12
Thus the encoding sequence. is partially interchanged in sequence from the display sequence and becomes:
I
0
, P
3
, B
1
, B
2
, P
6
, B
4
, B
5
, P
9
, B
7
, B
8
,
19
, P
12
, BI
0
, B
11
, The compressed data, that is encoded data, is arrayed in this sequence.
The reason the pictures are arrayed in this manner is now explained in connection with the operation of the arrangement shown in FIG.
1
.
In encoding the first picture
10
, data of a first picture to be encoded, supplied from the frame memory
10
, is blocked by the block divider
11
. The block divider
11
outputs block-based data in the sequence of Y
0
, Y
1
, Y
2
, Y
3
, Cb and Cr and transmits the data to the DCT circuit
14
via the changeover switch
13
the movable contact of which is set to the fixed contact a. The DCT circuit
14
executes two-dimensional DCT on the respective blocks for transforming the data from time axis to the frequency axis.
The DCT coefficients from the DCT circuit
4
are routed to the quantizer
15
so as to be quantized at a pre-set quantization step width. The quantized coefficients are then re-arrayed in a zig-zag order by the zing-zag scan circuit
16
, as shown in FIG.
4
. If the quantized coefficients are re-arrayed in the zig-zag order, the coefficients are arrayed in the order of increasing frequency so that the values of the coefficients become smaller in a direction proceeding towards the trailing end of the coefficient array. Therefore, if the coefficients are quantized with a given value S, the results of quantization tend to become zero towards the trailing end so that high-frequency components are cut off.
The quantized components are then sent to the variable length encoding circuit
17
where they are encoded by Huffman coding. The resulting compressed bitstream is transiently stored in the output buffer
18
from which it is transmitted at a constant bit rate. The output buffer
18
is a buffer memory for outputting an irregularly produced bitstream at a constant bit rate.
The above-described encoding for the picture by itself is termed intra-frame coding. The encoded picture is termed an I-picture.
If a decoder receives the bitstream for the I-picture, the above procedure is followed in the reverse order to complete th
Mitsuhashi Satoshi
Suzuki Kazuhiro
Johnson Timothy M.
Kessler Gordon
Sony Corporation
LandOfFree
Method and apparatus for selective control of degree of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for selective control of degree of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for selective control of degree of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2896559