JPH1093826A - Image companding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the compression rate by applying specific coding to 2-bit of a quantization code corresponding to a level of 4 gradation so as to preserve the gradation. SOLUTION: A CCD image sensor 15 of a read section 10 applies photoelectric conversion to an image of an original 11 and gives the result to a storage section 42. An image generating section 20 applies processing such as dither to data in the storage section 42, gives the result to a write section 22, by which a toner image formed on a photoreceptor 23 is transferred to paper. Image data outputted from a read section 10 consist of a 16-byte block of 4.4 picture elements with a depth in 8-bit. The storage section 42 applies coding processing to the block to compress it into 6-bytes consisting of a mean gradation index code in 1 byte, a gradation dispersion index code in 1 byte, and a quantization code in 2-bit each in the picture element. The quantization code is assigned to be (00), (01), (11), (10) corresponding to gradation levels 1-4 and low-order bits are excluded then the data are compressed into 4-byte per block.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image compression / expansion apparatus for compressing image data by a fixed-length block compression method, and more particularly to an image compression / expansion apparatus suitable for a digital copying machine.

[0002]

2. Description of the Related Art Generally, image data of one pixel output from a scanner of a digital copying machine is, for example, 8 bits (= 2 bits).
The image data to be processed and output to the printer is 1-bit (= binary) data. As a method of loading such multi-valued image data into a memory, for example, a method of storing multi-valued data as it is without compression as described in JP-A-63-244244 is described. However, according to this method, storing the data in the 8-bit state requires a memory having a data amount eight times as large as storing the binary data, resulting in an increase in cost. Therefore, when storing multi-valued image data in a memory, it is common to reduce the amount of data by performing multi-valued compression and storing the compressed data in order to reduce the memory cost.

As a method of compressing multi-valued image data, for example, “Journal of the Printing Society of Japan, Vol. 27, No. 3 (1990),
A fixed-length block compression method as shown in "Image compression method suitable for hard copy apparatus" is known. In this method, multi-valued image data is compressed at a fixed length in blocks of n lines × n pixels. The average tone index code in the block, the tone dispersion index code in the block as an index indicating how much the tone value of each pixel in the block varies, and the tone value of one pixel in the block The data is quantized to four levels, and is encoded into three code data consisting of a 2-bit quantization code indicating which level the gradation value of each pixel corresponds to.

According to this method, for example, when processing is performed on a block of 4 lines × 4 pixels, 16 bytes of 8-bit (256 gradations) multi-valued image data are obtained. Intra-block average gradation index code (symbol LA) Is 1 byte.-In-block gradation dispersion index code (symbol LD) is 1 byte.-Quantized value code (symbol Φ _ij ) is compressed to 2 bytes for each pixel, 4 bytes for 16 pixels, totaling 6 bytes. That is, in this case, since the original 16 bytes are compressed to 6 bytes, the compression ratio becomes 3/8, and if the amount of memory required for the original 16-byte data is 32 Mbytes, then 1
Only 2 MB is required.

Further, using this method, the quantization value code Φ _ij
Only the upper bits of the two bits are stored in the image memory,
A method of fixing the lower bits to either 0 or 1 at the time of decoding has been considered. The data amount in this case is as follows: the average tone index code LA in the block is 1 byte; the tone dispersion index code LD in the block is 1 byte. The quantization value code Φ _ij is 2 for 16 pixels of 1 bit per pixel.
Compressed to a total of 4 bytes. That is, in this case, since the original 16 bytes are compressed to 4 bytes, the compression ratio is reduced to 1/4. If the data amount of the memory required for the original 16-byte data is assumed to be 32 Mbytes, according to this method, 8 bytes are used.
Only M bytes are needed.

[0006]

However, in the above-described method of compressing and decompressing to 1/4, the lower 1 bit of the 2-bit quantized value code Φ _ij is thinned out, and the lower bit is set to 0 or 1 at the time of decompression. Since it is fixed to either,
However, there is a problem that image deterioration is more noticeable as compared with the compression method.

The present invention has been made in view of the above-mentioned conventional problems, and has as its object to provide an image compression / expansion apparatus capable of improving both the compression ratio and the image quality.

[0008]

In order to achieve the above object, a first means is to convert multi-valued image data into an average gradation index code in a block of N lines × M pixels and a pixel in a block. And a first coded data including a quantized value code obtained by quantizing the multi-valued image data of one pixel in the block into two bits, and an in-block gradation dispersion index code indicating how much the image data varies. Encoding means, the first encoded data encoded by the encoding means, the average gradation index code in the block, the gradation dispersion index code in the block, and the 2-bit quantized value code. Compression means for compressing the second encoded data including a one-bit quantized value code with the lower one bit decimated; storage means for storing the second encoded data compressed by the compression means; Read from means 1-bit data of "0" or "1" is added to the thinned-out lower bits of the quantized value code in the output second encoded data depending on whether the mode is the character mode or the photo mode. And a decompression means for decompressing the second encoded data into the first encoded data.

The second means is an average gradation index code in a block of multi-valued image data in units of N lines × M pixels, and a gradation distribution in a block indicating how much the pixel varies in the block. An index code, coding means for coding multi-valued image data of one pixel in the block into first coded data including a quantized value code obtained by quantizing the data into two bits, and coding by the coding means. In the block in which the first encoded data is replaced with the average tone index code in the block and one bit of the tone variance index code in the block is replaced with an area determination bit indicating whether the block is a text area or a photograph area. Compression means for compressing into a second coded data including a gradation dispersion index code and a 1-bit quantized value code obtained by thinning out the lower 1 bit of the 2-bit quantized value code;
Storage means for storing the second encoded data compressed by the compression means; and a second coded data read from the storage means.
The second encoded data is obtained by adding 1-bit data of “0” or “1” to the lower-order bits of the quantized value code in the encoded data of “1” according to the area determination bit. And decompression means for decompressing the first encoded data into the first encoded data.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a digital copying machine to which an embodiment of an image compression / expansion apparatus according to the present invention is applied, FIG. 2 is an explanatory diagram showing a document scanning direction of the copying machine shown in FIG. 1, and FIG. FIG. 4 is an explanatory diagram showing a synchronizing signal of the digital copying machine. FIG. 4 is a block diagram showing the storage unit of FIG. 1 in detail.
4 is an explanatory diagram showing the encoded data of the storage unit in FIG. 4, FIG. 6 is an explanatory diagram showing the encoding algorithm and the decoding algorithm of the encoding processing unit in FIG. 4, and FIG. FIG. 8 is an explanatory diagram showing a relationship between a tone index code, an in-block tone dispersion index code, and a quantized value code. FIG. 8 is a block diagram showing a memory control unit in FIG. 4 in detail, and FIG. 9 is a memory of the second embodiment. It is a block diagram which shows a control part in detail.

First, the reading process of the reading unit 10 and the image forming process of the image forming unit 20 of the digital copying machine shown in FIG. 1 will be briefly described. The reading unit 10 scans an original 11 as shown in FIG. 2 by an exposure lamp 13 movable in the sub-scanning direction along the surface of an original table 12, and reflects the reflected light by a mirror 14 so as to obtain a CCD image. The photoelectric conversion is performed by the sensor 15, and the electrical signal is converted into an electric signal corresponding to the intensity of light.

Next, after the signal is subjected to A / D conversion by an IPU (image processing unit) 16, processing such as shading correction and magnification change is performed.
Are transmitted to the image forming unit 20 or the storage unit 42 shown in detail in FIG. In order to execute such a reading process, the scanner control unit 17 performs detection of various sensors, controls a drive motor, and the like based on the control of the system control unit 44, and sets various parameters to the IPU 16.

In the image forming section 20, image data input from the reading section 10 or the storage section 42 via the selector 41 is further processed by the IPU 21 such as dither processing, and then sent to the writing section 22, where it is sent to the writing section 22. 22
Modulates the laser beam in accordance with the image data to form a latent image on the photoconductor 23. Around the photoconductor 23, a charger 24, a writing unit 22, and a developing device 2
5, a transfer charger 26, a separation charger 27, a cleaning device 28, and a charge removing charger 29 are arranged, and a latent image on the photoconductor 23 is developed by toner by a known electrophotographic process, and the toner image is transferred to a sheet.

Paper is fed from a paper feed tray 30 to a paper feed roller 31,
The sheet is discharged to a sheet discharge tray 36 via a registration roller 32, a transfer charger 26, a separation charger 27, a fixing device 34, and a sheet discharge roller 35. To execute such an image forming process, the plotter controller 37 controls the system controller 4.
Based on the control of (4), detection of various sensors and control of a drive motor and the like are performed.

The system control unit 44 detects an input state of the operator on the operation unit 45, sets various parameters for the reading unit 10 and the image forming unit 20, and executes a process execution instruction through communication. It is displayed on the operation unit 45. An instruction to the system control unit 44 is performed by an operator by performing a key input on the operation unit 45. The selector 41 selects the reading unit 10 or the storage unit 42 as a source of image data for the image forming unit 20 based on an instruction from the system control unit 44. The storage unit 42 constitutes an image compression / decompression device according to the present invention, and stores image data for one document input from the reading unit 10. The storage unit 42 is also used for a copy application such as a repeat copy and a rotation copy based on an instruction from the system control unit 44.

Next, the image synchronization signal will be described with reference to FIG. The frame gate signal (/ FGATE) is a signal indicating an image effective range in the sub-scanning direction. The image data becomes valid (active) at the low level. Also, the frame gate signal (/ FGATE) is the line synchronization signal (/ LSYN
Asserted or negated on the rising edge of C).

The line synchronization signal (/ LSYNC) is a predetermined number of clocks (= 8) at the rising edge of the pixel synchronization signal (PCLK).
And the image data in the main scanning direction becomes valid after a predetermined number of clocks (= 8) from the rising of the line synchronization signal (/ LSYNC). From the reading unit 10 to the image forming unit 2
The number of image data sent with respect to 0 is one for one cycle of the pixel synchronization signal (PCLK), and is raster-format data at a resolution of 400 dpi with the point indicated by the arrow in FIG. is there. The effective range of the image data in the sub-scanning direction is usually determined by the size of the transfer paper.

Next, the configuration of the storage section 42 will be described with reference to FIG. The encoding processing unit 4-1 is composed of a CPU and a logic circuit. The encoding processing unit 4-1 communicates with the memory control unit 4-3 to receive a command, perform an operation and a setting according to the command, and inform a state. Send status information to. The data line used for input / output of image data between the IPUs 16 and 21 is 8 bits. The encoding processing unit 4-1 converts the serially input image data into 4 bits.
A line memory (FIFO) is provided in the input / output unit to convert the data into a block of lines × 4 pixels. When an image input command is received, input image data is taken into an internal line memory in accordance with an input image synchronization signal. These 4 lines x
The data is read from the line memory in units of blocks of 4 pixels and encoded, and encoded data in units of blocks is output to the memory control unit 4-3 together with the memory access signal as needed. Also,
When an image output command is received, the frame memory 4
In -2, the encoded data stored in block units is read out and decoded by the memory control unit 4-3 using the memory access signal, and the image data is output to the selector 41 in synchronization with the output image synchronization signal. . In this embodiment, a block of 4 lines × 4 pixels is illustrated.
M pixels may be used.

FIG. 5 shows the pixels "00" to "33" of one block consisting of four lines "00" to "30" × 4 pixels "00" to "03". Figure 6
The original 16-byte image data is converted into a total of 6-byte encoded data "1" as follows based on the encoding algorithm shown in (1).

The average gradation index code LA in the block is 1 byte. The gradation dispersion index code LD in the block is 1 byte. The quantization value code Φ _ij is 4 for 16 pixels of 2 bits per pixel.
Byte Here, as shown in FIG. 7, the average tone index code LA in the block is the average value of the tone values of the pixels in the block, and the tone dispersion index code LD in the block is the variation of the pixels in the block. Represents Two bits are assigned to each pixel in the block for the quantization value code Φ _{ij, and} the gradation value of the pixel is set to 4 levels (2 bits) level “4” → data “10” level “3” → data “ 11 ”level“ 2 ”→ data“ 01 ”level“ 1 ”→ data“ 00 ”. Here, level “4” is not data “11” but data “10”, and level “3” is data “1”.
The data is “11” instead of “0”. Note that the encoded data “1” and “2” shown in FIG. 5 indicate the data structure (shown) of the first embodiment, and the encoded data “3” and “4” are the second
3 shows a data structure (illustration) of the embodiment.

The frame memory 4-2 is composed of a semiconductor memory device such as a DRAM having a capacity of 8 Mbytes, and writing and reading are controlled by a memory controller 4-3. The memory control unit 4-3 is configured by a CPU and a logic circuit, receives a command by communicating with the system control unit 44, performs an operation and a setting according to the command, and notifies the state of the storage unit 42. For this purpose, status information is transmitted to the system control unit 44. The operation commands from the system control unit 44 are an image input command, an image output command, and the like, and these two commands are also transmitted to the encoding processing unit 4-1. Encoding processing unit 4-
1 and the memory control unit 4-3 are connected via a data bus 4-4 of 48 (= 8 × 6) bits and a control line, and between the memory control unit 4-3 and the frame memory 4-2. , 32
It is connected to a (= 8 × 4) -bit data bus 4-5 via a control line.

Next, referring to FIG. 8, the memory control unit 4-3
Will be described in detail. Memory access control unit 431
Converts a frame memory control signal into a frame memory 4 based on an input / output memory access signal from the encoding processor 4-1.
-2. The input / output image address counter 432 is a counter that counts up based on an input / output memory access signal from the encoding processing unit 4-1. The input / output image address counter 432 has 21 bits (2M space) indicating an area where code data is stored in block units. The memory address is stored in the frame memory
Output to

An MPU (micro processing unit) 433 includes a CPU, an SCI (serial communication I / F),
An input / output image address counter 43 having an input / output port
A parameter indicating the write or read start address to the frame memory 4-2 for the second 2 is set on the CPU bus by 8 bits. The bidirectional buffer 434a with the encoding processing unit 4-1 and the bidirectional buffer 4 with the frame memory 4-2
34b are three-state, and the data direction is M
This is set by the PU 433.

The bidirectional buffer 434a and the bidirectional buffer 434b are connected via 16-bit data buses A and B, respectively, and the bidirectional buffer 434a and the data selector 435 are connected via a 16-bit data bus C. Connected through. The data buses A to C indicate the direction from the bidirectional buffer 434a to the bidirectional buffer 434b (when an image is input), and the data buses A 'to C' are from the bidirectional buffer 434b and the data selector 435 to the bidirectional buffer 434a. (At the time of image output).
The data bus C is not used at the time of image input, and at the time of image output, the data selector 435 outputs a signal “0” (in the character mode) or “0” to all of the 16-bit data buses C ′ based on a selection signal applied from the MPU 433. 1 "(in photo mode) and output.

In such a configuration, first, before reading a document, the system control unit 44 sends information indicating “character mode” or “photo mode” as the image processing mode to the MPU 433. Next, when the original is read, the input image data is compressed by the encoding processing unit 4-1 to obtain a 1-byte average gradation index code LA in a block,
6-byte coded data “1” including a 1-byte gradation dispersion index code LD in a block and a 4-byte quantized value code Φ _ij is transmitted to the bidirectional buffer 43 via the data bus 4-4.
4a.

As shown in FIG. 5, the encoded data "1" is compressed into 4-byte encoded data "2" in which the lower one bit of the two bits of the quantized value code Φ _ij is deleted. The 4-byte encoded data "2" is transferred from the bidirectional buffer 434a to the data buses A and B and the bidirectional buffer 4
34b, the frame memory 4- via the data bus 4-5.
2

On the other hand, at the time of image output, the 4-byte encoded data "2" read from the frame memory 4-2 is supplied to the data bus 4-5, the bidirectional buffer 434b,
Bidirectional buffer 434a via data buses A ', B'
At the same time, and 16
A signal “0” (in the character mode) or “1” (in the photo mode) is transferred to all of the bit data buses C ′.

The coded data "2" read from the frame memory 4-2 in the bidirectional buffer 434a.
6-byte encoded data "1" in which a signal "0" (in character mode) or "1" (in photo mode) is inserted into the lower one bit of the two bits of the quantized value code Φ _ij of The encoded data "1" is generated and transferred to the encoding processing unit 4-1 via the data bus 4-4. As a result, the encoded data "1" is output on paper after being decoded.

Next, a second embodiment will be described with reference to FIG. Here, in the image data input from the IPU 16 shown in FIG. 1, an area determination signal indicating whether the area is a "character area" or a "photograph area" is set for each pixel. Therefore, the encoding processing unit 4-1 generates an area determination bit indicating whether the area is a character area or a photograph area for each block of 4 lines × 4 pixels based on the area determination signal in the input image data from the IPU 16. . In this case, for a block in which pixels of a character area and a picture area are mixed, an area determination bit indicating a picture area in which image deterioration is not conspicuous.

Then, as shown in FIG. 5, the encoding processing section 4-1 provides the memory control section 4-3 with the 1-byte average gradation index code LA in the block and the 1-byte gradation in the block. 1-byte data obtained by replacing the area determination bit with the dispersion index code LD, and a 4-byte quantized value code Φ _ij
6 bytes of encoded data "3"
-4 to the bidirectional buffer 434a. Next, as in the first embodiment, 4-byte encoded data “4” from which the lower 1 bit of the quantized value code Φ _ij is deleted is transferred from the bidirectional buffer 434a to the data buses A and B and the bidirectional buffer 434b. Are transferred to the frame memory 4-2 via the data bus 4-5.

On the other hand, at the time of image output, 4-byte encoded data "4" read from the frame memory 4-2 is supplied to the data bus 4-5, the bidirectional buffer 434b,
Bidirectional buffer 434a via data buses A ', B'
At the same time, the data selector 435 determines the area determination bit on the data bus A ′, and outputs a signal “0” (in the character mode) or “1” to all of the 16-bit data bus C ′ according to the determination result. (In the photo mode).

In the bidirectional buffer 434a, the encoded data "4" read from the frame memory 4-2 is read.
6-byte coded data "3" in which a signal "0" (in character mode) or "1" (in photo mode) is inserted into the lower one bit of the two bits of the quantized value code Φ _ij of The encoded data "1" is generated and transferred to the encoding processing unit 4-1 via the data bus 4-4. As a result, the encoded data "1" is output on paper after being decoded.

Here, with reference to FIG. 7, when “0” is inserted into the lower one bit of the two bits of the quantized value code Φ _ij in the character mode at the time of decoding, the level becomes “1” or “4”. Therefore, the data value in the block has a sharp difference and the edge is emphasized. On the other hand, when "1" is inserted into the lower one bit of the two bits of the quantization value code Φ _{ij in} the photo mode, the level becomes "2" or "3". Data with good key reproducibility. Therefore, even if the lower one bit of the 2-bit quantized value code Φ _ij is thinned out and stored in the frame memory 4-2, both the compression ratio and the image quality can be improved.

[0034]

As described above, according to the first aspect of the present invention, the lower one bit of the two-bit quantized value code is thinned out and stored, and when reading, the character mode or the photograph mode is used. Since the data “0” or “1” is added to the lower 1 bit of the quantization code, both the compression rate and the image quality can be improved.

According to the second aspect of the present invention, one bit of the in-block gradation dispersion index code is replaced with an area determination bit indicating whether the block is a character area or a photograph area, and a two-bit quantized value code is used. The lower one bit is thinned out and stored, and at the time of reading, data “0” or “1” is added to the lower one bit of the quantized value code according to the area determination bit. Can be improved.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a digital copying machine to which an embodiment of an image compression / decompression device according to the present invention is applied.

FIG. 2 is an explanatory diagram showing a document scanning direction of the copying machine of FIG. 1;

FIG. 3 is an explanatory diagram showing a synchronization signal of the digital copying machine shown in FIG. 1;

FIG. 4 is a block diagram illustrating a storage unit of FIG. 1 in detail;

FIG. 5 is an explanatory diagram showing encoded data in a storage unit in FIG. 4;

FIG. 6 is an explanatory diagram showing an encoding algorithm and a decoding algorithm of the encoding processing unit in FIG. 4;

FIG. 7 is an explanatory diagram showing a relationship among 8-bit image data, an average gradation index code in a block, a gradation dispersion index code in a block, and a quantization value code.

FIG. 8 is a block diagram illustrating a memory control unit of FIG. 4 in detail;

FIG. 9 is a block diagram illustrating a memory control unit according to a second embodiment in detail;

[Explanation of symbols]

 4-1 Encoding processing unit 4-2 Frame memory 4-3 Memory control unit 433 MPU 434a, 434b Bidirectional buffer 435 Data selector

Claims (2)

[Claims] 1. An in-block average tone index code for multi-valued image data in block units of N lines × M pixels, an in-block tone dispersion index code indicating how much each pixel in a block varies, A first value including a quantized value code obtained by quantizing multi-valued image data of one pixel in a block into 2 bits;
Encoding means for encoding the encoded data of the first, the first encoded data encoded by the encoding means the average tone index code in the block, the tone dispersion index code in the block, Compression means for compressing the second encoded data including the 1-bit quantized value code obtained by thinning out the lower 1 bit of the 2-bit quantized value code; and the second encoded data compressed by the compressing means. Storage means for storing, and for the lower bits decimated in the quantized value code in the second coded data read from the storage means, "0" according to the character mode or the photograph mode, Or 1 of "1"
An image compression / expansion device comprising: expansion means for expanding the second encoded data into the first encoded data by adding bit data. 2. An in-block average tone index code for multi-valued image data in block units of N lines × M pixels, an in-block tone dispersion index code indicating how much each pixel in a block varies, A first value including a quantized value code obtained by quantizing multi-valued image data of one pixel in a block into 2 bits;
Encoding means for encoding the coded data of the first, second, and third bits of the first encoded data encoded by the encoding means; Is replaced by an area determination bit indicating whether the block is a text area or a photograph area, and a 1-bit quantized value code obtained by thinning out the lower 1 bit of the 2-bit quantized value code Compression means for compressing into second encoded data including: a storage means for storing the second encoded data compressed by the compression means; and a second encoded data read from the storage means. Is added to the lower bits decimated by the quantization value code of 1 in accordance with the area determination bit to add 1-bit data of “0” or “1”, thereby converting the second encoded data into the first code. Data Image decompression device and a decompression means for decompressing.