Image analysis – Image compression or coding – Transform coding
Reexamination Certificate
2001-03-15
2004-03-30
Tran, Phuoc (Department: 2621)
Image analysis
Image compression or coding
Transform coding
C382S233000
Reexamination Certificate
active
06714686
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing device for use in, e.g., a digital still camera, a facsimile machine, a digital copy machine, a videophone, a video CD player, a DVD player, etc. Specifically, the present invention relates to an image processing device which achieves an increase in the processing speed of pipeline image processing performed between discrete cosine transform processing/inverse discrete cosine transform processing and Huffman-encoding processing/Huffman-decoding processing for compression/expansion of color images; accurate encoding/decoding; and a decrease in circuitry size.
In the description hereinbelow, a discrete cosine transform is referred to as “DCT”, and an inverse discrete cosine transform is referred to as an “inverse DCT”.
2. Description of the Related Art
There are standards for high efficiency image encoding processing (data compression processing for image information) provided by a plurality of organizations, whereby compatibility of compressed data among various applicable fields is secured. For example, in the field of electronic communications, there is the H. 261 standard (which is a CCITT recommendation). Furthermore, in the field of recording technology, the International Standards Organization (ISO) standardize an MPEG as dynamic image encoding processing and a JPEG as still image encoding processing.
In encoding processing based on such standards, as shown in
FIG. 10
, encoding processing for dynamic images is achieved by a combination of a plurality of unit processing algorithms.
In the standardized dynamic image encoding processing as shown in
FIG. 10
, captured image information is converted into an electric signal by a charge coupled device (CCD)
1
, and the electric signal is digitally converted by an A/D converter
2
. The digitally converted image data is supplied to a motion estimation processing section
3
. The motion estimation processing section
3
compares the image data from the A/D converter
2
with data output from a frame memory
12
in which preceding frame images are accumulatively stored to determine an optimum one of the preceding frame images in the frame memory
12
. Output data from the motion estimation processing section
3
is subjected to a frame differential process
4
, and supplied to a DCT processing section
5
.
The DCT processing section
5
divides an image corresponding to the supplied image data into blocks each having a particular size, and DCT processes the image data on a block-by-block basis. In a quantization processing section
6
, the DCT-processed image data is quantized according to a quantization factor into image data components. The image data components are compressed by Huffman-encoding processing in a Huffman-encoding processing section
7
. The compressed image data components are recorded in a recording section
8
.
On the other hand, in an inverse quantization processing section
9
, the image data components obtained in the quantization processing section
6
are inversively quantized according to the above quantization factor into image data. The inversively quantized image data is output to an inverse DCT processing section
10
. The inverse DCT processing section
10
inversively DCT-processes the inversively quantized image data on the block-by-block basis. The inversively DCT-processed image data is subjected to addition processing with the optimum one of the proceeding frame images which has been determined by the motion estimation processing section
3
, and is accumulated in a frame memory
12
. Frame image data in the frame memory
12
is fed back to the motion estimation processing section
3
for subsequent motion estimation processing.
Furthermore, standardized encoding/decoding processing for still images is achieved by a combination of a plurality of unit processing algorithms as shown in FIG.
11
.
In the standardized encoding/decoding processing for still images shown in
FIG. 11
, captured image information is converted by a CCD 1 into an electric signal, and the electric signal is digitally converted by an A/D converter
2
. The image data digitally converted by the A/D converter
2
is output to a DCT processing section
5
. The DCT processing section
5
divides an image of the image data into a plurality of blocks each having a particular size, and DCT-processes the image data on block-by-block basis.
In a quantization processing section
6
, the image data which has been DCT-processed in the DCT processing section
5
is quantized according to a quantization factor into image data components. The image data components are compressed by Huffman-encoding processing in a Huffman-encoding processing section
7
. The compressed image data components obtained in the Huffman-encoding processing section
7
are recorded in a recording section
8
.
On the other hand, in the case where the compressed image data components recorded in the recording section
8
are decompressed to reproduce the original captured image, the compressed image data components are read out from the recording section
8
and decoded in a Huffman-decoding processing section
13
, whereby the compressed data components are converted into image data components. These image data components are inversively quantized by an inverse quantization processing section
9
, whereby the image data components are converted into image data. This image data is output to an inverse DCT processing section
10
. The inverse DCT processing section
10
inversively DCT-processes the image data on the block-by-block basis, thereby reproducing an image which is substantially the same as the original captured image.
In the encoding process for dynamic images shown in
FIG. 10
, in almost all the processing performed in the motion estimation processing section
3
and the downstream sections thereof, an entire image is divided into a mesh-like matrix formed by a plurality of image regions, and each type of processing is performed on each image region. Also in the encoding/decoding process for still images shown in
FIG. 11
, in almost all the processing performed in the DCT processing section
5
and the downstream sections thereof, an entire image is divided into a mesh-like matrix formed by a plurality of image regions, and each type of processing is performed on each image region. There are two types of image region-based processing; an image is processed on the units of a region formed by 8×8 pixels (a block) (e.g., processing in the DCT processing section
5
); or an image is processed on the units of a region formed by 16×16 pixels (a macroblock) (e.g., a processing in the motion estimation processing section
3
). Hereinbelow, processing on the units of an 8×8 pixel region (on the block-by-block basis) is described in detail.
In the encoding process, image data is sequentially input to the DCT processing section
5
on the block-by-block basis (i.e., on the units of an 8×8 pixel region). In the DCT processing section
5
, DCT processing is performed on the image data (two-dimensional image information) on the block-by-block basis (i.e., on the units of an 8×8 pixel region). This two-dimensional DCT processing is performed by executing one-dimensional DCT processing in a horizontal (row) direction and one-dimensional DCT processing in a vertical (column) direction. The two-dimensional DCT processing on the units of an 8×8 pixel region is represented by following arithmetic expression (1):
F
(
U,V
)=(¼)
C
(
U
)
C
(
V
)&Sgr;&Sgr;
f
(
i,j
)cos[(2
i
+1)
U&pgr;/
16]cos[(2
j
+1)
V&pgr;/
16] (1)
where f(i,j) is pixel data, and the initial value for i and j is 0. In expression (1),
when U=0, C(U)=1/2;
when U≠0, C(U)=1;
when V=0, C(V)=1/2; and
when V≠0, C(V)=1.
Data Fuv which has been obtained after the two-dimensional DCT processing is output to the quantization processing section
6
. The quantization proce
Birch & Stewart Kolasch & Birch, LLP
Sharp Kabushiki Kaisha
Tran Phuoc
LandOfFree
Image processing device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3291459