Image analysis – Image compression or coding – Predictive coding
Reexamination Certificate
1998-12-22
2002-10-29
Couso, Jose L. (Department: 2721)
Image analysis
Image compression or coding
Predictive coding
Reexamination Certificate
active
06473530
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to encoding and decoding apparatuses, and an image processing apparatus using the same.
Conventionally, image data or page description language (to be abbreviated as PDL hereinafter) data transferred from a computer to an image output device such as a printer is developed for drawing in the output device. Every time the data is developed for drawing, bitmap data is sent to the printer engine section. However, when contents to be developed for drawing are complex, the drawing development speed maybe lower than the drawing speed of the engine section. In this case, bitmap data developed for drawing is temporarily stored in a memory (this memory is called a page memory). After drawing development is complete in units of pages, and the bitmap data is stored in the memory, the bitmap data is sequentially sent to the printer engine section from the top of the page.
Assume that the size of paper on which the data is printed is A
3
, and the resolution is 600 dpi. In this case, even when the data is binary data in which the number of bits per pixel is 1, the total bitmap data amount becomes as large as 8 MB, and the large-capacity memory increases the printer cost.
To prevent this, an arrangement as shown in
FIG. 1
has been examined. Data received from a computer sequentially passes through an interface section
101
for receiving the data from the computer, a temporary buffer
102
for temporarily storing the data received from the computer, a drawing section
103
for developing the data received from the computer for drawing, a band buffer
104
for writing the data developed by the drawing section for drawing, an encoding section
105
for compression-coding the bitmap data in the band buffer, a page buffer
106
for storing the data compression-coded by the encoding section, and a decoding section
107
for decoding the encoded data in the page buffer, and is finally output to a printer engine section
108
for printing bitmap data obtained by decoding. When a plurality of band buffers
104
are used to parallelly execute development processing by the drawing section
103
and encoding processing by the encoding section
105
, the processing speed can be increased.
With this arrangement, the page memory capacity decreases from 8 MB in the arrangement without compression to about ½ to ¼. Instead, the band buffer
104
must be used, and the memory capacity increases accordingly. However, when the drawing development unit (this unit is called a band) is set to be a {fraction (1/16)} to {fraction (1/20)} page, the memory capacity can be decreased in total.
As the encoding scheme of the encoding section
105
, a compression scheme which guarantees a predetermined value as the lowest compression ratio for arbitrary bitmap data (text, graphic, or image data) is desired because of the strong requirement for cost reduction and the purpose of minimizing the page memory capacity. As such a compression scheme, JBIG encoding having a function of learning the two-dimensional features of bitmap data to be compressed can be used.
In JBIG encoding, learning is performed by updating the contents of a RAM for holding a predictive state. This learning (update of the contents of the RAM) occurs at an irregular timing, and the time required for encoding/decoding becomes long because of the write operation in the memory. Conversely, if learning (update of the contents of the RAM) need not be performed, the encoding/decoding time shortens. When encoded data is decoded by JBIG, the rate of data output from the decoding section
107
is not constant, so the output cannot be directly output to the printer engine section
108
. To solve this, a FIFO (First In First Out) memory
109
is inserted between the decoding section
107
and the printer engine section
108
. The bitmap data output from the decoding section
107
is smoothed over time and then output to the printer engine section
108
.
The present inventor has proposed the following processing method in a patent filed by the present applicant previously.
FIG. 3
shows the arrangement of this new proposal. The proposal contents are different from
FIG. 2
in the following two points.
(1) Bitmap data decoded by the decoding section
107
is written in the band buffer
104
.
(2) The bitmap data written in the band buffer
104
is output to the printer engine
108
at a predetermined timing.
The differences from
FIG. 2
are a data path
301
for processing (1) and a data path
302
for processing (2).
As the difference in function, processing of smoothing the bitmap data over time is performed not by the FIFO
109
but the band buffer
104
, unlike FIG.
2
. In
FIG. 2
, the band buffer
104
operates during drawing development and encoding. In the example of
FIG. 3
, however, the band buffer
104
also operates during decoding.
The operation timing is shown in FIG.
4
. For the descriptive convenience, bitmap data of one page is divided into six bands, and the bands are named A
1
, B
2
, A
3
, B
4
, A
5
, and B
6
from the upper side. To increase the throughput of processing, the band buffer
104
has a double buffer structure, and the two buffers are called a buffer A and a buffer B, respectively.
Drawing development processing in the band buffers is performed in the order of A
1
, B
2
, A
3
, B
4
, A
5
, and B
6
(a in FIG.
4
). A
1
, A
3
, and A
5
are developed in the buffer A, and B
2
, B
4
, and B
6
are developed in the buffer B. Since development of A
1
is ended before the start of development of B
2
, compression coding of A
1
is performed in parallel to development of B
2
(b in FIG.
4
). Subsequently, compression coding of B
2
is performed in parallel to development of A
3
, and finally, B
6
is compression-coded. When all bitmap data of one page are compression-coded, the compressed data are decoded.
Decoding is also performed in the order of A
1
, B
2
, A
3
, B
4
, A
5
, and B
6
(c in FIG.
4
), like encoding. The bitmap data A
1
of one band decoded by the decoding section
107
is written in the buffer A. Subsequently, the decoded bitmap data B
2
of one band is written in the buffer B. In parallel to the write in the buffer B, the bitmap data is read out from the buffer A and sent to the printer engine section
108
(d in FIG.
4
), and printing of one page is started (e in FIG.
4
).
Subsequently, in parallel to the write of the decoded bitmap data A
3
, B
2
is read out and transferred to the printer engine
108
, and finally, B
6
is read out and transferred to the printer engine
108
. With this processing, all bitmap data of one page are sent to the printer engine
108
, and print output is ended (e in FIG.
4
).
FIG. 5A
is a block diagram of conventional JBIG encoding and decoding apparatuses used as the encoding section
105
and decoding section
107
in
FIGS. 1
to
3
. The operation will be briefly described.
Referring to
FIG. 5A
, reference numeral
501
denotes an arithmetic operation section for performing arithmetic operation in JBIG;
502
, a learning RAM for holding a predictive state;
503
, an ST & MPS generation section for generating expectation data to be stored in the learning RAM
502
;
504
, a terminal for inputting context (CX);
505
, a terminal for inputting an mode signal to exchange address signal and data signal for RAM
502
in the memory clear mode;
511
, a counter for generating an address signal for the learning RAM
502
in the memory clear mode;
513
, a data generation section for generating zero data to be written in the learning RAM
502
in the memory clear mode;
515
, a pulse generation section for generating a write pulse to be supplied to the learning RAM
502
in the memory clear mode; and
521
,
523
, and
525
, selectors.
Before encoding or decoding, a memory clear mode signal (High) is input to the terminal
505
to clear the learning RAM
502
. When this signal goes high, the selector
521
selects the counter
511
, the selector
523
selects the data generation section
513
, and the selector
525
Canon Kabushiki Kaisha
Couso Jose L.
Fitzpatrick ,Cella, Harper & Scinto
LandOfFree
Encoding and decoding apparatuses and image processing... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Encoding and decoding apparatuses and image processing..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Encoding and decoding apparatuses and image processing... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2986430