JPH04373261A - Image data compressing and elongating device - Google Patents

Abstract

PURPOSE:To permit an image reading part and an image recording part to continuously operate even if a buffer memory consisting of memory elements which are fewer than page memories is used. CONSTITUTION:This system is provided the image reading part 1 and/or the image recording part 2 which continuously operates, a compression/elongating part 5, the buffer memory 4 which is used to temporarily store image data and a code storing part 6 storing code data obtained by compressing with the compression/elongating part 5. The capacitance of the buffer memory 4 is decided based on the capacitance of the code storing part 6, the worst compressiblity in a encoding system to adopt and the ratio of a data transferring speed between the compression/elongating part 5 and the buffer memory 4 to that between the buffer memory 4 and the image recording part 2 or the image reading part 1.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image data compression/expansion device used in facsimile machines and the like.

0002

[Prior Art] An image data compression/expansion device is a device that compresses image data obtained by reading an image through encoding to create code data, or conversely expands code data through decoding to create image data. Yes, and used in facsimile machines, etc.

[First conventional example: intermittent operation method using buffer memory] FIG.
This is a facsimile machine using a conventional example. In FIG. 8, 1 is an image reading section, 2 is an image recording section, 3 is a buffer memory control circuit, 4 is a buffer memory, 5 is a compression/expansion section, 6 is a code storage section, 7 is an operation panel, and 8 is a CPU (central 9 is ROM (read only memory), 10 is RAM (random access memory),
11 is a modem, 12 is a line control unit, 13 is a communication line, 1
4 is a bus, 19 is an image data input control circuit, 20 is an image data output control circuit, and 22 is an image data compression/expansion device.

The image data input control circuit 19 controls the writing of image data from the image reading section 1 into the buffer memory 4, and the image data output control circuit 20 controls the writing of image data from the buffer memory 4 to the image recording section 2. Controls when reading image data.

The main operations performed by the image data compression/expansion device 22 are (1) image reading operation, (2) printing operation, and (2) copying operation. FIG. 11 shows the paths through which data flows when performing these operations.

FIG. 11A shows a path through which data flows during an image reading operation. An image reading unit 1 reads an image of a document to generate image data, which is temporarily stored in a buffer memory 4. Next, the image data is taken out from the buffer memory 4, encoded by the compression/expansion section 5, and converted into code data. The time required for encoding varies depending on the complexity of the image. For example, if the white parts are continuous, it will take only a short time, but the more the white and black are mixed together, the longer the time will take. The longest time is required for images called "1-bit ladder", in which white and black appear alternately for each bit. The encoded code data is stored in the code storage section 6. The code storage unit 6 is, for example, a DRAM (dynamic memory).
random access memory).

[0007] When transmitting the code data stored in the code storage unit 6 to another facsimile device, the code storage unit 6 → modem 11 → line control unit 12 → communication line 1
3, and is recorded in the code storage unit 6 in the other facsimile machine.

FIG. 11B shows a path through which data flows during a printing operation. The printing operation is performed when code data is transmitted from another facsimile machine. The code data in the code storage section 6 is sent to the compression/expansion section 5, decoded, and converted into image data. After the image data is temporarily stored in the buffer memory 4, it is sent to the image recording section 2 and printed.

FIG. 11C shows a path through which data flows during a copy operation. Image data generated by reading a document in an image reading section 1 is temporarily stored in a buffer memory 4, and then sent to an image recording section 2 for printing.

The operation panel 7 is a panel through which an operator issues various operation commands. ROM9 has CPU8
Stores programs for actions to be performed. RAM10
provides a work area necessary for the CPU 8 to operate.

Incidentally, the capacity of the buffer memory 4 is normally extremely small compared to the amount of image data for one page of a document. FIG. 7 is a diagram illustrating a buffer memory and a page memory. 7(A) shows one page of the original 17, FIG. 7(B) shows the page memory 16 which can store all the image data of the original 17, and FIG. 7(C) shows the buffer memory 4. buffer memory 4
can store only part of the original document 17, for example, only the image data of the solid line part, and cannot store the image data of the dotted line part. This caused
The image reading section 1 and the image recording section 2 must operate intermittently as described below.

FIG. 10 is a diagram illustrating the reason why the first conventional image data compression/expansion apparatus must use an image reading section and an image recording section that operate intermittently. In FIG. 10, 4-1 is a used part, 4-3 is an overlap part, 18 is printing paper, 18-1 and 18-3 are printing parts, and 18-2 is a non-printing part. FIG. 10(a) shows the printing state on the printing paper 18 during the printing operation, and FIG. 10(b) shows the usage state of the buffer memory 4 during the reading operation.

The buffer memory 4 is used as a ring buffer in which the address value is incremented until the last address is reached, and then the first address is specified. Therefore, if the buffer memory 4 is read until it is full during an image reading operation and reading is continued, new image data will be written overlapping the previously read image data (Fig. 10 (Lower)). ) overlap part 4-3). In this case, the previous image data will be deleted before it is used.

[0014] Since this would be undesirable, the reading operation of the image reading section 1 is temporarily stopped. After the image data in the buffer memory 4 has been encoded by the compression/expansion section 5, the reading operation of the image reading section 1 is resumed from the position where it was stopped earlier. Therefore, as the image reading section 1,
An image scanner that can read images intermittently must be used.

Furthermore, during the printing operation, the image data stored in the buffer memory 4 from the code storage unit 6 via the compression/expansion unit 5 is only a portion of one page. The image data in the buffer memory 4 is transferred to the image recording unit 2 at a constant speed.
sent to. However, since the decoding speed in the compression/expansion section 5 varies depending on the complexity of the image, the data is not sent from the compression/expansion section 5 to the buffer memory 4 at a constant speed. Therefore, the buffer memory 4 may become empty. If the image recording unit 2 continues to operate in this state, a blank area will be created on the printing paper (the non-printing area 18-
2), the image will be cut off.

Since this would be undesirable, when the buffer memory 4 becomes empty, the printing operation of the image recording section 2 is temporarily stopped and resumed after the next image data arrives in the buffer memory 4. Therefore, as the image recording section 2, a thermal printer, a dot printer, etc., which can perform printing intermittently, is used.

[Second Conventional Example: Continuous Operation Method Using Page Memory] In the first conventional example described above, the image reading section 1
Also, since the image recording section 2 operates intermittently, there is a drawback that reading time and printing time are long. Therefore, there are devices that can operate continuously rather than intermittently.

FIG. 9 shows a facsimile machine using a second conventional image data compression/expansion device. The code is shown in Figure 8.
It corresponds to that of Further, 15 is a page memory control unit, and 16 is a page memory. The capacity of the page memory 16 is A3 size with a dot density of 400 dpi (dpi).
If the target is binary data of 200 MB per inch, approximately 4 MB is required. The operations of the same components as in FIG. 8 are the same as in the first conventional example, so their explanations will be omitted. However, as the image reading section 1, an image scanner that can operate continuously is used, and as the image recording section 2, a printer that operates continuously, such as a laser beam printer, is used. These perform reading operations and printing operations at a constant speed determined by each model.

As shown in FIG. 7(b), the page memory 16 has a capacity capable of storing image data of one page of the original, so it is possible to read one page of the original,
When printing, there is no need to stop reading or printing in the middle. Therefore, the overall time required for reading and printing becomes shorter and faster.

[0020] Conventional documents related to image data compression/expansion devices include, for example, Japanese Patent Laid-Open No. 140975/1983.
JP-A-64-36461, JP-A-62-126430, JP-A-63-26706
0, Japanese Patent Application Laid-Open No. 64-44678, etc.

[0021]

However, the page memory used in the second conventional example described above is expensive because it requires a large number of memory elements, which raises the problem of increasing the cost of the image data compression/decompression device. There was a point. The present invention aims to solve these problems.

That is, the present invention provides image data compression that allows one or both of the image reading section and the image recording section to operate continuously even when using a buffer memory composed of fewer memory elements than a page memory. The purpose of the present invention is to provide a decompression device.

[0023]

[Means for Solving the Problems] In order to solve the above-mentioned problems, the present invention provides an image data compression/decompression device that includes an image reading section that operates continuously and/or an image recording section that operates continuously. It includes a compression/expansion section, a buffer memory used for temporarily storing image data, and a code storage section for storing code data compressed by the compression/expansion section, and the capacity of the buffer memory is determined by the code storage section. capacity, the worst compression rate of the encoding method used, and the ratio of the data transfer speed between the compression/expansion section and the buffer memory to the data transfer speed between the buffer memory and the image recording section or image reading section. shall be established.

Further, the image data decompression device includes an image recording section that operates continuously, and a buffer memory having a capacity that allows image data for one page of a document to be sent out without interruption during the recording operation of the image recording section. , an expansion unit that supplies image data to the buffer memory, and a system that prevents the buffer memory from becoming empty even when decompressing code data that takes the longest time to decompress when the expansion unit performs an expansion operation. The configuration includes a buffer memory control circuit that controls the buffer memory so that it becomes full.

Furthermore, the image data compression device includes an image reading section that operates continuously, and a buffer memory having a capacity that allows image data for one page of a document to be sent out without erasing it during the reading operation of the image reading section. , a compression unit that receives image data from the buffer memory, and a compression unit that prevents the buffer memory from becoming full even when compressing code data that takes the longest time to compress when the compression unit performs a compression operation. and a buffer memory control circuit that controls the buffer memory to become empty.

[0026]

[Operation] In the present invention, the capacity of the code storage section is reduced to C.
, where K is the worst compression rate of the adopted encoding method, and P is the ratio of the data transfer speed between the compression/expansion section and the buffer memory to the data transfer speed between the buffer memory and the image recording section or the image reading section. , the capacity Y of the buffer memory is
It is determined that the following relationship is satisfied.

[0027] By doing this, at the time of compression,
Image data that is sent one after another to the buffer memory at a constant speed from a continuously operating image reading section can be processed without the buffer memory becoming full. Furthermore, during decompression, image data must be sent from the buffer memory at a constant speed to the continuously operating image recording section, but this can be done without the buffer memory becoming empty.

[0028] By adjusting the values of C and P, the capacity Y can be made smaller than the capacity when used as a page memory, so that costs can be reduced.

[0029]

Embodiments Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 shows a facsimile machine using the image data compression/expansion apparatus of the present invention. The symbols correspond to those in FIG. As the image reading section 1 and the image recording section 2, those that operate continuously are used.

FIG. 2 is an enlarged view of the main parts of the image data compression/expansion apparatus of the present invention. The symbols correspond to those in Figure 1, 3-
1 is a memory usage counter, 3-2 is a memory usage determination circuit, 3-3 is a frame width register, 3-4 to 3-6 are address registers, 3-7 to 3-9 are address counters, DREQ1 to 3 is the data request signal, ACK1
3 are acknowledge signals, and LE and LE 1 to 3 are line end signals.

FIG. 3 is a diagram illustrating the structure of the buffer memory 4. FIG. 3A shows that the address of the buffer memory 4 is set in units of words. In this example, one word is made up of 16 bits from 0 to 15. Buffer memory 4 is used as a ring buffer. That is, when it is desired to advance the address to the next address after the final address Z, the address is specified so as to return to the initial address 0.

FIG. 3(B) shows the capacity of the buffer memory 4 when the length of one line (this is called the "frame width") is W.
It is represented by a surface made of words. The starting address of line 1 on this surface is W, and the starting address of line 2 is 2W. In the case of a document with a large frame width W, the number of lines that can be stored is small. Conversely, in the case of a document with a small frame width W, the number of lines that can be stored is large.

FIG. 4 is a diagram illustrating the process of using the buffer memory during decompression. 4-1 is a used portion, 4-2 is an empty portion, LR is a read line (line where reading is being performed), and LW is a writing line (line where writing is being performed). Figure 4 (a) shows the state before use.
Neither the read line LR nor the write line LW has entered the first line yet.

FIG. 4B shows a state in which image data writing is in progress. The write line LW advances in the direction of the arrow and has reached near the last line of the buffer memory 4. Reading has not yet started and read line LR has not yet entered the first line. The dotted part is the used part 4-1, and the blank part is the vacant part 4.
-2. Since the buffer memory 4 is used as a ring buffer, once the last line is written, the next line is returned to the first line.

FIG. 4C shows a state where writing and reading are being performed. The read line LR starts from the first line and progresses to near the last line. The write line LW returns from the last line to the first line and writes over the image data that has been read out and is no longer used. Therefore, the memory used part 4-1 and empty part 4-2 are as shown in the figure.

The memory usage counter 3-1 in FIG. 2 is a counter that counts the number of lines of image data that have not yet been read out, that is, the number of lines in the usage section 4-1 in FIG. When the line end signal LE indicating that writing of one line is completed is input to the buffer memory control circuit 3, the count is incremented by 1, and the line end signal LE indicating that reading of one line is completed is input. For example, the count is decremented by 1.

[0037] The memory usage amount determination circuit 3-2
-1 is completely gone (Empty),
Or, it is a circuit that determines whether the buffer memory 4 is filled with the used portion 4-1 (Full). Specifically, the determination is made by comparing the value of the memory usage counter 3-1 (that is, the number of lines in the used section 4-1 with dots in FIG. 4) and the number of lines 0 or the maximum number of lines LMAX. . When the number of lines matches 0, an empty signal Em is output, and the maximum number of lines LMA that can be stored in the buffer memory 4
When it matches X, a full signal Fu is output.

A frame width value (the number of words constituting one line of an image) is set in the frame width register 3-3. In this example, the value is W. For example, the number of bits per line when reading a document with the width of an A3 size paper using the 400 dpi image reading unit 1 is 4864 bits. This frame width is 304 words, assuming that one word is 16 bits.

When the compression/expansion section 5 performs the compression operation,
It is desirable that the buffer memory 4 be in a near-empty state so that there is sufficient room to accept image data sent one after another from the image reading section 1. On the other hand, when carrying out the decompression operation, it is desirable that the buffer memory 4 be in a state where there is plenty of image data (nearly a full state) so that the image data sent to the image recording section 2 is not interrupted.

Before explaining what conditions the capacity of the buffer memory 4, etc. should satisfy in order to realize the above state, let us explain the operation of the image data compression/expansion device. I will explain the operation, printing operation, and copying operation separately.

[Image reading operation] This operation is an operation in which image data read by the image reading section 1 is temporarily stored in the buffer memory 4, and then converted into code data by the compression/expansion section 5 and stored in the code storage section 6. It is.

(1) Image reading unit 1→writing operation to buffer memory 4 First, the compression/expansion unit 5 is activated to start the compression operation. However, since the image data has not yet been read into the buffer memory 4, an empty signal Em is output from the memory usage determination circuit 3-2. When there is no image data to be compressed, the compression operation is in a standby state. When the image reading section 1 is activated, image data is input to the image data input control circuit 19 one bit at a time in series according to a synchronizing signal.

Image data is input from the image data input control circuit 19 to the buffer memory 4 in units of words (
For example, if one word is 16 bits, the data is processed in parallel (as a group of 16 bits). The input is performed when an acknowledge signal ACK1 issued in response to the data request signal DREQ1 is received.

The address register 3-4 and address counter 3-7 are used when inputting data from the image data input control circuit 19 to the buffer memory 4. Address register 3-4 indicates the start address of the line currently being written. For example, during the period in which the word belonging to line 1 is being written in FIG. 3(b), the address W is maintained. Address counter 3-7 indicates the address of the word being written. For example, when writing the second word from the beginning of line 1 in Figure 3 (b), it is W+1, and every time one word is written,
be done. Note that when writing the first word of the first line, the contents (0) of the address register 3-4 are set in the address counter 3-7.

Each time a word corresponding to one line is sent,
A line end signal LE1 is issued. This line end signal LE1 causes the value of the address register 3-4 to be the sum of the frame width W. For example, when line 2 ends, W is added to 2W to make 3W (2W+W=3W). Writing one line of image data to the buffer memory 4 increases the memory usage in the buffer memory 4 by one line, so the line end signal LE1 is input to the memory usage counter 3-1.
Add 1 to the count.

(2) Buffer memory 4 → code storage section 6
When even one line is written to the buffer memory 4, the empty signal Em disappears. When the empty signal Em disappears, the compression/expansion section 5 is caused to start the compression operation. Acknowledge signal AC in response to data request signal DREQ3
When K3 is issued, the image data in the buffer memory 4 is read out. Reading is performed from the beginning of the first line. The read image data is compressed and
The code is stored in the code storage section 6.

The address register 3-6 is the address register 3-6 of the compression/expansion section 5.
Indicates the start address of the line being read. Address counter 3-9 indicates the address of the word being read to compression/expansion section 5. Therefore,
When reading the word at the head of the first line, the value (0) of the address register 3-6 is set in the address counter 3-9.

Each time one word is read, the value of the address counter 3-9 is incremented by 1. Every time the reading of one line is completed (that is, every time the line end signal LE3 is output), the frame width W is added to the value of the address register 3-6.
is added. Furthermore, when one line is read from the buffer memory 4, the image data of that line becomes obsolete, and the memory usage amount decreases by one line. Therefore,
When the line end signal LE3 is received, the value of the memory usage counter 3-1 is decremented by 1.

[Printing operation] This operation is carried out by the code storage section 6.
This is an operation in which the code data stored in the image data is converted into image data by the compression/expansion section 5, temporarily stored in the buffer memory 4, and sent to the image recording section 2 for printing.

(1) Code storage section 6 → buffer memory 4
Regarding the writing operation, the compression/expansion section 5 is activated, the code data in the code storage section 6 is converted into image data, and the image data is written into the buffer memory 4. At the beginning of the line to be written, the value of the address register 3-6 is set in the address counter 3-9. The address to write the image data is indicated by the address counter 3-9. Every time you write one word,
The value of address counter 3-9 is incremented by 1. When writing of one line is completed and a line end signal LE3 is issued, the frame width W is added to the value of the address register 3-6, and the value of the memory usage counter 3-1 is incremented by 1.

(2) Regarding the operation of reading data from the buffer memory 4 to the image recording unit 2 When the amount of memory used in the buffer memory 4 increases and the full signal Fu is output, the image recording unit 2 is activated. Then, when the acknowledge signal ACK2 is issued in response to the data request signal DREQ2 from the image data output control circuit 20, the image data is read from the buffer memory 4. This readout is performed in word units, but the image data is sent out from the image data output control circuit 20 to the image recording section 2 in a series according to a synchronization signal.

When reading image data from the buffer memory 4 to the image data output control circuit 20, the value of the address register 3-5 is set in the address counter 3-8 at the beginning of the first line. address counter 3
The value of -8 is incremented by 1 every time one word is read. When the reading of one line is completed, a line end signal LE2 is output, the frame width W is added to the address register 3-5, and the value of the memory usage counter 3-1 is decremented by 1.

[Copy operation] This operation is performed by the image reading section 1.
This is an operation in which the image data read in is temporarily stored in the buffer memory 4 and printed by the image recording section 2. The image data read by the image reading unit 1 is stored in the buffer memory 4.
The operation of writing to the image data is the same as the operation described in the [Image reading operation] section, and the operation of reading the image data from the buffer memory 4 to the image recording section 2 is the same as the operation described in the [Printing operation] section. be.

In order to perform the above operations continuously for one page of the original without any interruption, the buffer memory 4 must not be empty during the decompression operation, and must not be empty during the compression operation. Sometimes it should not be full. The present invention focuses on this relationship and calculates the minimum capacity of the buffer memory 4. Next, an example of the calculation will be explained.

[Example of calculation of capacity of buffer memory 4] Figure 5
1 is a diagram for explaining calculation of the memory capacity required for the buffer memory 4, and the explanation will be given by taking the case of expansion as an example. 17 is a document, 17-1 is a white portion, 17-2 is a worst compressed image portion, 6-1 is a white code portion, 6-2 is a worst compressed image code portion, and 21 is a 1-bit ladder. Figure 5 (
b) shows the path through which data flows during printing operations. FIG. 5B shows the states of the code storage section 6, buffer memory 4, and document 17 when the decompression operation is performed under the worst conditions.

The worst compressed image portion 17-2 of the original 17 is:
When read and compressed by the image reading unit 1, this is the image part that has the lowest compression ratio. Such an image portion is a 1-bit ladder 21, which is an image in which white and black are arranged alternately for each bit. Therefore, the worst compressed image section 17-2 is completely filled with the 1-bit ladder 21. MH method (Modif) used in facsimile machines
When the 1-bit ladder 21 is compressed using the Huffman method), the compression ratio is 1/4.5. In other words, 1
When bit image data is compressed, it becomes 4.5 bits of code data. In this case,
Instead of getting shorter, it gets longer.

The white code section 6-1 is a section that stores code data obtained by compressing the image data of the white section 17-1, and the worst compressed image code section 6-2 is a section that stores code data obtained by compressing the image data of the white section 17-1. This is the part that stores code data that is compressed image data.

The capacity and data transfer speed of each part in FIG. 5(a) are set as follows. ■Capacity of buffer memory 4... Y bytes ■Capacity of code storage unit 6... C bytes ■Transfer rate from buffer memory 4 to image recording unit 2 (constant)...X bytes/s (s... seconds) ■Code Transfer speed from storage unit 6 to compression/expansion unit 5... n
X bytes/s (However, n...a positive constant. A number indicating how many times the number is compared to the above X) ■Decompression operation capacity of the compression/decompression unit 5...
However, P...A coefficient indicating how many times it is compared to the above X)

When the code data in the code storage section 6 is converted into image data by the compression/expansion section 5 and sent to the image recording section 2 via the buffer memory 4, even if data for one page is transferred continuously, It is required to prevent the formation of non-printed areas 18-2 as shown in FIG. 10(A) on the printing paper. To this end, even if there is code data that takes the longest time to decompress in the compression/decompression unit 5,
It is necessary to be able to send up to the last data to the image recording section 2 without the buffer memory 4 becoming empty. Therefore, assuming the worst condition that is most likely to cause a non-printed portion, the following is an estimate of the capacity of the buffer memory 4 that will not cause a non-printed portion even under such conditions. Calculated as follows.

First, the worst condition will be explained with reference to FIG. 5(b). The first condition that defines the worst condition is that, as shown in the diagram with dots, if it is decompressed and converted to image data, the amount will be just enough to fill the buffer memory 4, and it will be in a compressed state. The condition is that code data corresponding to an image that can be stored in the code storage unit 6 using the smallest area exists at the beginning of the code storage unit 6. Such an image is a white image. Therefore, at the beginning of the code storage section 6, there is a white code section 6.
Assume that -1 exists.

The second condition that defines the worst condition is that code data (with the worst (minimum) compression ratio) that takes the longest time to decompress is stored after the white code section 6-1. The condition is that the worst compressed image code section 6-2 exists.

Transfer from the buffer memory 4 to the image recording section 2 starts when the buffer memory 4 is filled with image data (when the full signal Fu is output).

The width of the original 17 is the same as that of A3 size paper, and the dot density is 400 dpi (dot pe
When the image data is read by the image reading unit 1 of the R inch, one line of image data has 4864 bits. Therefore,
The following number of lines can be stored in the buffer memory 4 having a capacity of Y bits.

Next, the memory capacity of the white code section 6-1 corresponding to the white line image data filled in the buffer memory 4 is determined. To express one white line of image data consisting of 4864 bits with code data, use MH.
Since the method requires 44 bits, the memory amount (bytes) of the white code section 6-1 is as follows.

Then, the memory capacity of the worst compressed image code section 6-2 can be determined by the following equation. Since the value of the second term of this equation is smaller than 1K byte when Y=1M byte, in order to simplify the calculation, the memory capacity of the worst compressed image code section 6-2 ≈C...(1 ).

When the worst (minimum) compression rate is K, the amount (bytes) of image data generated by expanding the code data of the worst compressed image code section 6-2 is: Image data amount corresponding to = K
C...(2).

The total amount (bytes) of image data generated by expanding the code data of the white code section 6-1 and the worst compressed image code section 6-2 is Y+KC. After the image recording unit 2 is activated, reading is performed continuously from the buffer memory 4 at a constant speed of X bytes/s, so the time T1 required to finish reading the total amount continuously is as follows. Become.

On the other hand, the code data of the worst compressed image code section 6-2 is transferred to the compression/expansion section 5 using the white code section 6-1.
As soon as the transfer to the compression/expansion section 5 is completed, the transfer is performed immediately after that. The compression/expansion section 5 is the worst compressed image code section 6.
The decompression of the code data of -2 is started at the same time when the image recording section 2 starts reading from the buffer memory 4. Since the transfer speed is PX bytes/s, the time required to complete the transfer of the image data obtained by decompressing the worst-case compressed image code section 6-2 to the buffer memory 4 is T2.
becomes as follows.

In order to prevent non-printing areas from occurring on the printing paper being printed in the image recording unit 2, it is necessary to finish decompressing the code data in the worst-compressed image code unit 6-2 before time T1 elapses. It is necessary to keep it. That is, T1 ≧T2
...The relationship (5) must hold. (5)
By substituting equations (3) and (4) into the equations and organizing them, the capacity Y of the buffer memory 4 is calculated as follows.

That is, when the values of K, C, and P are determined, Y
If you set the values to satisfy equation (6), even if you use continuous operation as the image reading section 1 and image recording section 2, part of the read image data will disappear or the printed image will be interrupted. Things will go away.

[0071] If C=1M byte, P=0.8, K
= 1/4.5, then Y≧1M/18 bytes. If, as in the conventional example shown in FIG. However, compared to this, the capacity of the buffer memory 4 of the present invention of 1M/18 bytes is extremely small.

FIG. 6 is a diagram showing a print path when a reduction circuit is added. The reduction circuit 23 is a circuit for reducing the paper size of the document to be read by the image reading section 1 to match the size of the printing paper when the paper size is larger than the size of the printing paper. For example, reduction is performed by thinning out part of the data.

Since thinning is performed, the speed at which image data is supplied to the buffer memory 4 is lower than when the reduction circuit 23 is not inserted. If the supply speed is low, the buffer memory 4 is likely to become empty, so if the values of C, n, P, etc. described above are the same, the buffer memory 4 is It is necessary to increase the capacity somewhat.

In the above example, both the image reading section 1 and the image recording section 2 are operated continuously, but either one of them may be operated continuously. Further, it is possible to apply the present invention to an image data expansion device that performs only an expansion operation, and it is also possible to apply it to an image data compression device that performs only a compression operation.

[0075]

As described above, according to the image data compression/expansion device of the present invention, even if the image reading section and the image recording section are operated continuously, the capacity is much smaller than that of the page memory. Since it is only necessary to provide a buffer memory of 1, it is possible to reduce costs.

[Brief explanation of the drawing]

[Figure 1] Facsimile machine using the image data compression/expansion device of the present invention

[Figure 2] Enlarged view of main parts of image data compression/decompression device

[Figure 3] Diagram explaining the structure of buffer memory

[Figure 4]
Diagram explaining the process of using buffer memory

[Figure 5]
Diagram to explain calculation of required buffer memory capacity

[Figure 6] Diagram showing the print path when a reduction circuit is added

[Figure 7] Diagram explaining buffer memory and page memory

[Figure 8] A facsimile machine using the first conventional example of an image data compression/expansion device

[Figure 9] Facsimile machine using the second conventional example of image data compression/expansion device

FIG. 10 is a first conventional image data compression/expansion device,
Diagram explaining the reason for using an image reading section and an image recording section that operate intermittently

[Fig. 11] Diagram showing the paths through which data flows during various operations in the image data compression/expansion device.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1... Image reading unit, 2... Image recording unit, 3... Buffer memory control circuit, 3-1... Memory usage amount counter, 3-2... Memory usage amount determination circuit, 3-3... Frame width register, 3
-4 to 3-6...address register, 3-7 to 3-9...address counter, 4...buffer memory, 4-1...used section, 4-2...vacant section, 4-3...overlap, 5...compression Extension section, 6... Code storage section, 6-1... White code section, 6-
2... Worst compressed image code section, 7... Operation panel, 8... CP
U, 9...ROM, 10...RAM, 11...modem, 12...
Line control unit, 13...Communication line, 14...Bus, 15...Page memory control unit, 16...Page memory, 17...Document, 1
7-1... White area, 17-2... Worst compressed image area, 18... Printing paper, 18-1, 18-3... Printing area, 18-2... Non-printing area, 19... Image data input control circuit, 20... Image data output control circuit, 21... 1-bit ladder, 22... Image data compression/expansion device, 23... Reduction circuit

Claims (3)

[Claims] [Claim 1] An image reading section and/or an image reading section that operates continuously.
or an image recording section that operates continuously, a compression/expansion section,
It comprises a buffer memory used for temporarily storing image data, and a code storage section that stores code data compressed by the compression/expansion section, and the capacity of the buffer memory is determined by the capacity of the code storage section. It is characterized by being determined based on the ratio of the data transfer speed between the compression/expansion section and the buffer memory to the worst compression rate in the encoding method used, and the data transfer speed between the buffer memory and the image recording section or the image reading section. Image data compression/decompression device. 2. An image recording section that operates continuously; a buffer memory having a capacity that allows image data for one page of a document to be sent out without interruption during the recording operation of the image recording section; An expansion unit that supplies data, and when the expansion unit performs an expansion operation, the buffer memory is kept full so that the buffer memory does not become empty even when decompressing code data that takes the longest time to decompress. An image data decompression device having a buffer memory control circuit that controls the image data in the following direction. 3. An image reading section that operates continuously; a buffer memory having a capacity to send image data for one page of a document without erasing it during the reading operation of the image reading section; A compression section receives data supply, and when the compression section performs a compression operation, the buffer memory is emptied so that the buffer memory does not become full even when compressing code data that takes the longest time to compress. An image data compression device having a buffer memory control circuit that controls the image data in the direction of the image data compression device.