JP2003274173A - Image processor and image forming device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain quantized data which has a smooth image in quantizing multivalued gradation image data into low gradation data with an error diffusing method. <P>SOLUTION: This image processor ACP provided with a threshold generating means for generating a quantizing threshold that periodically vibrates and gradation processing means 4c, 1485, 1487 and 1488 for converting multi- gradation image data into quantized data of low gradation by the error diffusing method, is provided with threshold generator 2a to 2c, 4a and 4b for generating a quantizing threshold such that an image to be recorded grows along a screen angle according to the quantized data within a particular period on the image space as the density level of the multi-gradation image data goes up. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus which performs gradation processing for converting multi-value gradation image data into low gradation quantized data by an error diffusion method, and an image forming apparatus using the same.

[0002]

2. Description of the Related Art An example of this type of image processing apparatus and image forming apparatus is disclosed in Japanese Patent Laid-Open No. 2001-352448. This device generates a quantization threshold value using a dither threshold value matrix arranged spirally from the inner side to the lower side in the order of decreasing threshold value, and uses the quantization threshold value for input multi-value gradation data. Then, it is converted into low gradation quantized data by the error diffusion method. The laser printer forms an image on the transfer paper based on the quantized data.

[0003]

In the above example, the threshold value for multi-valued error diffusion is a dot concentration type in which output dots (recording dots) grow spirally within a specific period in the image space. Although it depends on the image forming characteristics of the image, it tends to be "rough" in the highlight part of low density recording, and "smooth" required for images output by color copiers.
Images are difficult to obtain easily.

It is an object of the present invention to obtain quantized data that easily expresses a smooth image when quantizing multi-value gradation image data into low gradation data by an error diffusion method.

[0005]

[Means for Solving the Problems] (1) Threshold generating means for generating a quantized threshold that periodically oscillates, and a floor for converting multi-tone image data into low tone quantized data by an error diffusion method. In the image processing device including the tone processing means (4c, 1485, 1487, 1488), the multi-tone image data is recorded in accordance with the quantized data within a specific cycle on the image space as the density level of the image data increases. An image processing apparatus (ACP), comprising: threshold generation means (2a to 2c, 4a, 4b) for generating a quantization threshold so that an image grows along a screen angle.

For ease of understanding, in parentheses, symbols for corresponding elements or corresponding matters in the embodiments shown in the drawings and described later are added for reference. The same applies to the following.

According to this, the quantized data capable of easily reproducing the smooth image can be obtained even by the tone processing for converting the multi-tone image data into the quantized data of the low tone by the error diffusion method.

[0008]

BEST MODE FOR CARRYING OUT THE INVENTION (2) The conversion to the quantized data is performed from the upper left to the lower right in the image space ((d) of FIG. 13), and the quantization threshold is set to the specific cycle. The image processing apparatus (ACP) is switched so as to increase from left to right in the inside (FIGS. 13A and 14);

Depending on the value of the error weight matrix and the processing direction, the direction of the error diffusion quantization processing is affected, and depending on the quantization threshold value of the error diffusion processing, the image formation along the screen angle is smooth. In some cases, a rough image may be obtained. In this embodiment, the quantization threshold of the error diffusion process is increased from the left to the right along the processing direction, so that the lines along the screen angle are formed neatly and the occurrence of "roughness" of the image is suppressed. And a smooth image is obtained.

(3) The conversion into the quantized data is performed from the upper right to the lower left in the image space (see FIG. 24).
(A)), the quantization threshold is switched so as to increase substantially from right to left within the specific period (FIG. 24).
(B), FIG. 25); The image processing apparatus (ACP) according to (1) above.

Depending on the value of the error weight matrix and the processing direction, the direction of the error diffusion quantization processing is affected, and depending on the value of the quantization threshold value of the error diffusion processing, the image formation along the screen angle may occur. The image may not be smooth and may be rough.

In this embodiment, the quantization threshold value of the error diffusion processing is increased from the right to the left along the processing direction, so that the lines along the screen angle are formed neatly and the "roughness" of the image is reduced. Generation is suppressed, and a smooth image is obtained.

(4) In the low density level area of the multi-tone image data, the quantization threshold value is switched so that output dots are preferentially grown in the main scanning direction (D in FIG. 14).
T1, PM of FIG. 15, DT16 of FIG. 22, P of FIG.
M4, DT19 in FIG. 25); The image processing apparatus (ACP) according to any one of (1) to (3) above.

When the quantization threshold value control of the error diffusion process is grown by giving priority to the diagonal direction, the pixels along the screen angle are not smoothly formed, and the image may be rough. For example, in the case of a laser printer formed by using laser light and toner, when an electrostatic latent image is formed on one isolated pixel with energy in a low image density level region, toner is formed by the laser light. It may not adhere to the latent image, and as a result, the image may be rough.

In the present embodiment, the quantization threshold is switched so that it grows preferentially in the main scanning direction. For example, by forming two pixels continuously in the main scanning direction, the toner adheres more stably and the roughness of the image is reduced.

(5) In the low density level area of the multi-tone image data, the quantization threshold value is switched so that the output dots are preferentially grown in the sub-scanning direction (D in FIG. 26).
T22, PM7 in FIG. 27); The image processing apparatus (ACP) according to any one of (1) to (3) above.

When the quantization threshold value control of the error diffusion processing is performed with priority given to the oblique direction, the pixels are not smoothly formed along the screen angle, and the image may be rough. Further, in the above (4), the quantization threshold of the error diffusion process is switched so that the output dots are preferentially grown in the main scanning direction. By performing such image formation, it is called banding or jitter. There is a possibility that another problem will occur. Due to the occurrence of banding or jitter, for example, a light-density striped pattern that is regularly or irregularly arranged in the sub-scanning direction is generated on an image even when the image density should be originally uniform.

In this embodiment, the quantization threshold is switched so that the output dots are preferentially grown in the sub-scanning direction, so that the image defect due to banding or jitter can be reduced.

(5a) In the low density level area of the multi-tone image data, 2 in the main scanning direction or the sub-scanning direction.
The image processing device according to any one of (1) to (5) above, wherein the quantization thresholds are switched so that the quantization thresholds of pixels or more are substantially the same value.

According to this, it is possible to stably form an image in which a low-density highlight has less roughness.

(6) The image processing device (ACP) according to any one of the above (1) to (5a); and the gradation processing means (4c, 1485, 1487) of the image processing device (ACP). , 1488) to form an image on a transfer sheet based on the quantized data output by the printer (100);

According to this, in the image formation using the printer, the same advantageous effects as the above (1) to (5a) can be obtained.

Other objects and features of the present invention will become apparent from the following description of embodiments with reference to the drawings.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the appearance of a multi-function full color digital copying machine according to the first embodiment of the present invention. This full-color copying machine is roughly composed of an automatic document feeder (ADF) 30, an operation board 20, a color scanner 10, a color printer 100, and a finisher 34 with a tray on which a stapler and image-formed sheets can be stacked. , Double-sided drive unit 33, paper feed bank 35, and large-capacity paper feed tray 3
It is composed of 6 units. A LAN (Local Area Network) connected to a personal computer PC is connected to the image data processing device ACP (FIG. 3) in the machine, and a telephone line PN (facsimile communication line) is connected to the facsimile control unit FCU (FIG. 3). The switch PBX connected to the switch is connected. The printed paper of the color printer 100 is discharged onto the paper discharge tray 108 or the finisher 34.

FIG. 2 shows the mechanism of the color printer 100. The color printer 100 of this embodiment is a laser printer. In this laser printer 100, four sets of toner image forming units for forming images of magenta (M), cyan (C), yellow (Y), and black (black: K) are arranged in a moving direction of the transfer paper. They are arranged in this order along (from the lower right to the upper left in the figure). That is, 4
It is a continuous drum type full-color image forming apparatus.

The magenta (M), cyan (C), yellow (Y), and black (K) toner image forming units are respectively provided on the photoconductor drums 111M, 111C, and 11D.
A photoconductor unit 110M having 1Y and 111K,
110C, 110Y and 110K, and developing unit 1
20M, 120C, 120Y and 120K. Further, the arrangement of each toner image forming unit is such that the rotation axes of the photoconductor drums 111M, 111C, 111Y and 111K in each photoconductor unit are parallel to the horizontal x-axis.
In addition, the transfer paper moving direction (on the y and z planes, 4
It is set so as to have an arrangement of a predetermined pitch on the left upward line (which forms 5 °). As the photoconductor drum of each photoconductor unit, a photoconductor drum having an organic photoconductor (OPC) layer on the surface and having a diameter of 30 mm was used.

In addition to the toner image forming unit, the laser printer 100 carries an optical writing unit 102 by laser scanning, paper feed cassettes 103 and 104, a pair of registration rollers 105, and a transfer paper to carry each toner image. Transfer transport belt 160 that transports the transfer position of the forming unit
A transfer belt unit 106 having the above, a fixing unit 107 of a belt fixing system, a paper discharge tray 108, a double-sided drive (surface reversal) unit 33, and the like. The laser printer 100 also includes a manual feed tray, a toner supply container, a waste toner bottle, and the like, which are not shown.

The optical writing unit 102 includes a light source, a polygon mirror, an f-θ lens, a reflection mirror, etc., and each photoconductor drum 111M, 111C, 11 based on image data.
The surface of 1Y and 111K is irradiated with laser light while being swung in the x direction. Further, the alternate long and short dash line in FIG. 2 indicates the transfer paper conveyance path. Paper feed cassette 103, 10
The transfer paper fed from No. 4 is transported by the transport rollers while being guided by a transport guide (not shown), and is sent to the registration roller pair 105. The transfer paper sent to the transfer / transport belt 160 at a predetermined timing by the registration roller pair 105 is carried by the transfer / transport belt 160, and is transported so as to pass through the transfer position of each toner image forming unit.

Photosensitive drum 111 of each toner image forming section
The toner images formed on M, 111C, 111Y, and 111K are transferred onto the transfer paper carried and conveyed by the transfer / conveyance belt 160, and the transfer paper on which the color toner images are superposed, that is, the color image is formed, is a fixing unit. Sent to 107. That is, the transfer is a direct transfer method in which the toner image is directly transferred onto the transfer paper. When passing through the fixing unit 107, the toner image is fixed on the transfer paper. The transfer paper on which the toner image is fixed is discharged or fed to the paper discharge tray 108, the finisher 36, or the double-sided drive unit 33.

The outline of the yellow Y toner image forming unit will be described below. The other toner image forming units have the same structure as that of the yellow Y. As described above, the yellow Y toner image forming unit corresponds to the photoconductor unit 110Y.
And a developing unit 120Y. The photoconductor unit 110Y includes, in addition to the photoconductor drum 111Y, a brush roller for applying a lubricant to the surface of the photoconductor drum, a swingable blade for cleaning the surface of the photoconductor drum, and a discharge lamp for irradiating the surface of the photoconductor drum with light. , A non-contact type charging roller for uniformly charging the surface of the photosensitive drum, and the like.

In the photoconductor unit 110Y, the optical writing unit 102 modulates the surface of the photoconductor drum 111Y uniformly charged by the charging roller to which the AC voltage is applied, based on the print data, and deflected by the polygon mirror. When the laser light L is emitted while being scanned, an electrostatic latent image is formed on the surface of the photoconductor drum 111Y. The electrostatic latent image on the photoconductor drum 11IY is the developing unit 20Y.
Is developed into a yellow Y toner image. At the transfer position on the transfer / transport belt 160 where the transfer paper passes, the toner image on the photoconductor drum 11IY is transferred to the transfer paper. After the toner image is transferred, the surface of the photoconductor drum 111Y is coated with a predetermined amount of lubricant by a brush roller, cleaned by a blade, and discharged by light emitted from a discharge lamp to remove static electricity. Prepared for forming a latent image.

The developing unit 120Y contains a two-component developer containing a magnetic carrier and a negatively charged toner. Then, a developing roller provided so as to be partially exposed from the opening of the developing case 120Y on the side of the photosensitive drum,
Conveyor screw, doctor blade, toner concentration sensor,
Equipped with a powder pump. The developer contained in the developing case is triboelectrically charged by being stirred and conveyed by the conveying screw. Then, a part of the developer is carried on the surface of the developing roller. The doctor blade uniformly regulates the layer thickness of the developer on the surface of the developing roller, the toner in the developer on the surface of the developing roller is transferred to the photoconductor drum, and the toner image corresponding to the electrostatic latent image is transferred to the photoconductor. Appears on the drum 111Y. The toner density of the developer in the developing case is detected by the toner density sensor. When the density is insufficient, the powder pump is driven to replenish the toner.

The transfer / conveying belt 160 of the transfer belt unit 106 is provided on the photosensitive drum 111 of each toner image forming portion.
It is wound around four grounded tension rollers so as to pass through the respective transfer positions in contact with and opposed to M, 111C, 111Y and 111K. Of these tension rollers,
An electrostatic attraction roller to which a predetermined voltage is applied from a power source is arranged so as to face the inlet roller on the upstream side in the transfer paper moving direction indicated by the two-dot chain line arrow. The transfer paper passing between these two rollers is electrostatically adsorbed on the transfer / transport belt 160. Further, the outlet roller on the downstream side in the transfer paper moving direction is a drive roller that frictionally drives the transfer conveyance belt, and is connected to a drive source (not shown). Further, a bias roller to which a predetermined cleaning voltage is applied from a power source is arranged so as to come into contact with the outer peripheral surface of the transfer / transport belt 160. The bias roller removes foreign matter such as toner adhering to the transfer / transport belt 160.

Further, the photosensitive drums 111M, 111C,
A transfer bias applying member is provided so as to come into contact with the back surface of the transfer / conveying belt 160 that forms a contact facing portion that contacts and faces 111Y and 111K. These transfer bias applying members are fixed brushes made of Myra, and a transfer bias is applied from each transfer bias power source. By the transfer bias applied by the transfer bias applying member, transfer charge is applied to the transfer / transport belt 160, and a transfer electric field having a predetermined strength is formed between the transfer / transport belt 160 and the surface of the photosensitive drum at each transfer position. .

FIG. 3 shows the system configuration of the image processing system of the copying machine shown in FIG. In this system, a color original scanner 12 including a reading unit 11 and an image data output I / F (Interface) 12 is used as an image data interface control C of an image data processing device ACP.
It is connected to a DIC (hereinafter simply referred to as CDIC).
The image data processing device ACP also includes a color printer 1
00 is connected. The color printer 100 includes an image data processing device IPP (Imag (Imag) of the image data processing device ACP.
e Processing Processor; simply described as IPP below)
Then, the write I / F 134 receives the recorded image data and the image forming unit 135 prints it out. The image forming unit 135 is the one shown in FIG.

Image data processing apparatus A which is an image processing apparatus
CP (hereinafter simply referred to as ACP) is a parallel bus P
b, image memory access control IMAC (hereinafter, simply I
MAC), a memory module M that is an image memory
EM (hereinafter simply referred to as MEM), system controller 1, RAM 4, non-volatile memory 5, operation board 20,
It has font ROM6, CDIC, IPP, etc. The facsimile control unit FCU is connected to the parallel bus Pb.
(Hereinafter simply referred to as FCU) is connected.

Reading unit 11 for optically reading a document
Is a photoelectric converter for converting the light reflected by the lamp onto the original by a CCD to generate R, G, B image data, and outputs the output I / F 12
Is converted into RGB image data, shading corrected, and sent to the CDIC.

The CDIC outputs I / O for image data.
Data transfer between the F12, the parallel bus Pb, and the IPP, and communication between the process controller 131 and the system controller 1 that controls the overall ACP are performed. Also,
The RAM 132 is used as a work area of the process controller 131, and the non-volatile memory 133 stores a boot program of the process controller 131 and the like.

The IPP is a programmable arithmetic processing means for performing image processing. Output I / F 12 to CDIC
The image data input to is sent to IPP via CDIC.
The signal deterioration (signal deterioration of the scanner system) due to the quantization into the optical system and the digital signal is corrected by the IPP, and is output (transmitted) to the CDIC again.

The image memory access control IMAC (hereinafter simply referred to as IMAC) controls writing / reading of image data to / from the MEM. The system controller 1 controls the operation of each component connected to the parallel bus Pb. RAM 4 is the system controller 1
Used as a work area of the non-volatile memory 5 and stores a boot program of the system controller 1 and the like.

The operation board 20 inputs the processing to be performed by the ACP. For example, the type of processing (copying, facsimile transmission, image reading, printing, etc.) and the number of processings are input. As a result, the image data control information can be input. The contents of the FCU will be described later.

The processing of the image data read by the reading unit 11 includes a job of storing the read image data in the MEM for reuse and a job of not storing the read image data in the MEM. Each case will be described. As an example of accumulating the read image data in the MEM, a plurality of sheets may be copied for one document. In this case, the reading unit 11 is operated only once, and the image data read by the reading unit 11 is read. The image data stored in the MEM is read out a plurality of times.

As an example of not using MEM, there is a case where only one original is copied, and in this case, the read image data can be reproduced as it is.
Access to the MEM is not required. ME
When M is not used, the data transferred from the IPP to the CDIC is returned from the CDIC to the IPP again. I
In the PP, the RGB signal from the CCD in the output I / F 12 is color-converted into a YMCK signal, and image quality processing such as printer γ conversion, gradation conversion, and gradation processing such as dither processing or error diffusion processing is performed.

The image data after the image quality processing is transferred from the IPP to the writing I / F 134. The write I / F 134 is
Laser control is performed on the tone-processed signal by pulse width and power modulation. After that, the image data is sent to the image forming unit 135, and the image forming unit 135 forms a reproduced image on the transfer paper.

Next, the flow of image data when the image data is accumulated in the MEM and is subjected to additional processing at the time of image reading, such as rotation in the image direction and image composition, will be described. The image data transferred from the IPP to the CDIC is C
It is sent from the DIC to the IMAC via the parallel bus Pb.

Based on the control of the system controller 1, the IMAC controls access to image data and MEM, L
The print data of a personal computer PC (not shown) connected to the AN (hereinafter simply referred to as PC) is expanded, and the image data is compressed / expanded for effective use of the MEM.

The image data sent to the IMAC is stored in the MEM after data compression, and the stored image data is read out as needed. The read image data is
The image data is decompressed, restored to the original image data, and returned from the IMAC to the CDIC via the parallel bus Pb. After transfer from CDIC to IPP, image quality processing is performed and write I / F 13
4 and the reproduced image is formed on the transfer paper in the image forming unit 135.

In the flow of image data, the function of the digital multi-function peripheral is realized by the bus control by the parallel bus Pb and the CDIC.

Facsimile transmission is performed by performing image processing on the read image data by the IPP and transferring it to the FCU via the CDIC and the parallel bus Pb. The FCU converts data into a communication network and sends it to the public line PN as facsimile data. The facsimile reception is performed by converting the line data from the public line PN into image data by the FCU and transferring the image data to the IPP via the parallel bus Pb and the CDIC. In this case, no special image quality processing is performed, the writing I / F 134 outputs the image, and the image forming unit 135 forms a reproduced image on the transfer paper.

In a situation where a plurality of jobs, for example, a copy function, a facsimile transmission / reception function, and a printer output function operate in parallel, the reading unit 11 and the image forming unit 13
The allocation of the usage right of the bus 5 and the parallel bus Pb to the job is controlled by the system controller 1 and the process controller 131. Process controller 1
Reference numeral 31 controls the flow of image data, the system controller 1 controls the entire system, and manages the activation of each resource. Further, the function selection of the digital multi-function peripheral is performed on the operation board 20, and the selection contents of the operation board 20 set the processing contents such as the copy function and the facsimile function.

The system controller 1 and the process controller 131 communicate with each other via the parallel buses Pb, CDIC and serial bus Sb. Specifically, communication between the system controller 1 and the process controller 131 is performed by converting data between the parallel bus Pb and the serial bus Sb in the CDIC and a data format conversion for the interface.

Various bus interfaces such as the parallel bus I / F 7, serial bus I / F 9, local bus I / F 3 and network I / F 8 are I / F.
It is connected to the MAC. Controller unit 1
Connects with related units via multiple types of buses in order to maintain independence within the entire ACP.

The system controller 1 controls other functional units via the parallel bus Pb. The parallel bus Pb is used for transferring image data.
The system controller 1 issues an operation control command for accumulating image data in the MEM to the IMAC. This operation control command is IMAC, parallel bus I /
It is sent via F7 and the parallel bus Pb.

In response to this operation control command, the image data is transferred from the CDIC to the parallel bus Pb and parallel bus I.
/ F 7 to IMAC. Then, the image data is stored in the MEM under the control of IMAC.

On the other hand, the ACP system controller 1
Is called as a printer function from a PC,
Functions as a printer, controller, network control, and serial bus control. If via the network, the IMAC receives the print output request data via the network I / F 8.

For general-purpose serial bus connection, IMA
C receives the print output request data via the serial bus I / F 9. The general-purpose serial bus I / F 9 supports multiple types of standards, for example, USB (Universal
It corresponds to the interface of the standard such as Serial Bus), 1284 or 1394.

The print output request data is expanded into image data by the system controller 1. The deployment destination is an area in MEM. The font data necessary for development can be obtained by referring to the font ROM 6 via the local bus I / F 3 and the local bus Rb.
The local bus Rb connects the controller 1 to the nonvolatile memory 5 and the RAM 4.

Regarding the serial bus Sb, in addition to the external serial port 2 for connection with a PC, there is also an interface for transfer with the operation board 20 which is the operation unit of the ACP. This is not print data, IM
It communicates with the system controller 1 via AC, receives processing procedures, and displays the system status.

Data transmission / reception between the system controller 1, the MEM and various buses is performed via the IMAC. Jobs using MEM are centrally managed in the entire ACP.

FIG. 4 shows an outline of the functional configuration of the CDIC. The image data input / output control 161 receives the RGB image data output by the output I / F 12 and outputs the RGB image data to the IPP. The IPP is the scanner image processing 190 (FIG. 6) described later.
Then, it is sent to the image data input control 162 of the CDIC.

The data received by the image data input control 162 is subjected to the primary compression of the image data in the data compression unit 163 in order to improve the transfer efficiency on the parallel bus Pb. The compressed image data is converted into parallel data by the data conversion unit 164 and sent to the parallel bus Pb via the parallel data I / F 165. The image data input from the parallel data bus Pb via the parallel data I / F 165 is serially converted by the data conversion unit 164. This data is primarily compressed for bus transfer, and is expanded by the data expansion unit 166. The decompressed image data is transferred to the IP by the image data output control 167.
Forwarded to P. In the IPP, RGB image data is converted into YMCK image data by printer image processing 200 (FIG. 6) described later, and converted into image data YpMpCpKp for image output of the printer 100 and output to the write I / F 134.

The CDIC has a conversion function of parallel data transferred by the parallel bus Pb and serial data transferred by the serial bus Sb. System controller 1
Transfers data to the parallel bus Pb, and the process controller 131 transfers data to the serial bus Sb. A data conversion unit 164 and a serial data I / F 169 for communication between the two controllers 1 and 131.
Then, parallel / serial data conversion is performed. The serial data I / F 168 is for IPP and transfers serial data together with the IPP.

FIG. 5 shows an outline of the structure of the IMAC. I
The MAC includes an access control 172, a memory control 173, a compression / expansion module 176, and an image editing module 17.
7, system I / F 179, local bus control 18
0, parallel bus control 171, serial port control 17
5, serial port 174 and network control 17
Eight.

The compression / expansion module 176, the image editing module 177, the parallel bus control 171, the serial port control 175, and the network control 178 are connected to the access control 172 via the DMAC (Direct Memory Access Control).

The system I / F 179 sends and receives commands or data to and from the system controller 1. Basically, the system controller 1 controls the entire ACP. Further, the system controller 1 manages the resource allocation of the MEM, and the control of other units is controlled by the system I / F.
179, parallel bus control 171, and parallel bus Pb.

Each unit of the ACP is basically connected to the parallel bus Pb. Therefore, the parallel bus control 171 manages transmission / reception of data to / from the system controller 1 and the MEM by controlling bus occupancy.

The network control 178 controls the connection with the LAN. The network control 178 manages transmission / reception of data to / from an external expansion device connected to the network. Here, the system controller 1 does not participate in the operation management of the connected devices on the network, but the IM
It controls the interface in AC. Although not particularly limited, in this embodiment, 100BASE
A control for -T is added.

The serial port 174 connected to the serial bus has a plurality of ports. The serial port control 175 has a number of port control mechanisms corresponding to the types of buses prepared. Although not particularly limited, in this embodiment, port control for USB and 1284 is performed. In addition to the external serial port, it also controls the reception and transmission of data relating to command reception or display with the operation unit.

The local bus control 180 interfaces with the local serial bus Rb to which the RAM 4, the non-volatile memory 5, and the font ROM 6 for expanding the printer code data necessary for starting the system controller 1 are connected.

As for operation control, command control by the system controller 1 from the system I / F 179 is carried out. Data control focuses on the MEM and manages MEM access from external units. The image data is transferred from the CDIC to the IMAC via the parallel bus Pb. Then, the image data is taken into the IMAC by the parallel bus control 171.

The memory access of the captured image data is separated from the management of the system controller 1. That is, the memory access is performed by direct memory access control (DMAC) independent of system control. Regarding access to the MEM, the access control 172 arbitrates access requests from a plurality of units. Then, the memory control 173 controls the access operation of the MEM or the reading / writing of data.

When accessing the MEM from the network, the data fetched from the network into the IMAC via the network control 178 is transferred to the MEM by the direct memory access control DMAC. The access control 172 arbitrates access to the MEM in a plurality of jobs. The memory controller 173 reads / writes data from / to the MEM.

When accessing the MEM from the serial bus, the serial port 1 is controlled by the serial port control 175.
The data fetched into the IMAC via 74 is transferred to the MEM by the direct memory access control DMAC. The access control 172 is a MEM for multiple jobs.
Arbitrate access to. The memory control 173 is
Read / write data from / to MEM.

The print output data from the PC connected to the network or the serial bus is expanded by the system controller 1 into the memory area in the MEM using the font data on the local bus.

The system controller 1 manages the interface with each external unit. IMAC
Regarding the data transfer after being taken in, each DMAC shown in FIG. 5 manages the memory access. In this case, since the DMACs execute data transfer independently of each other, the access control 172 collides jobs related to access to the MEM or prioritizes each access request.

Here, each DM is used to access the MEM.
In addition to the access by the AC, the access from the system controller 1 via the system I / F 179 for developing the bitmap of the stored data is also included. In access control 172, D permitted to access the MEM
The MAC data or the data from the system I / F 179 is directly transferred to the MEM by the memory control 173.

The IMAC has a compression / expansion module 176 and an image editing module 177 regarding data processing inside the IMAC. The compression / decompression module 176
Data is compressed and expanded so that image data or code data can be effectively stored in the MEM. compression/
The decompression module 176 controls the interface with the MEM by the DMAC.

The image data once stored in the MEM is called by the direct memory access control DMAC from the MEM to the compression / expansion module 176 via the memory control 173 and the access control 172. The image data thus converted is returned to the MEM or output to the external bus by the direct memory access control DMAC.

The image editing module 177 controls the MEM by the DMAC and processes the data in the MEM. Specifically, in addition to clearing the memory area, the image editing module 177 performs image data rotation processing, combining different images, etc. as data processing. The image editing module 177 performs editing for converting data to be processed by address control on the memory.

The image editing module 177 processes the bitmap image developed on the MEM.
The image editing module 177 cannot edit the compressed code data and printer code data. Therefore, image compression for effective memory storage is performed on the data after image editing.

FIG. 6 shows an outline of the image processing function of IPP. Separation and generation (determination of whether the image is a character region or a photo region: image region separation) 192, scanner γ conversion 193, filter 194, color correction 202, scaling 203, image processing 2 shown in FIG.
Of the processes of 04, printer γ conversion 205 and gradation process 206, the processes performed by the processor array 144 of FIG. 8, which will be described later, are separation generation 192, filter 194, scaling 203, image processing (create) 204, and This is a process in which some pixels can be processed at the same time by a part of the gradation process 206, and a merit can be obtained by changing the process contents by a program. In addition, it is useful in performing processing other than sequential processing such as error diffusion which is affected by the previous processing. The sequential processing is performed by special processing 148 and 149 (FIG. 7).

The special processing (circuits) 148 and 149 shown in FIG. 7, which will be described later, of the IPP is the error diffusion processing in the scanner γ conversion 193, the color correction 202, the printer γ conversion 205, and the gradation processing 206. Quantization of multi-gradation image data into low-gradation data. This quantization will be described later in detail.

The original image to be copied is separated into R, G, and B by the color original scanner 10 and is read. As an example, each bit has an n-bit structure (n is 1 to 14 and has various gradation bit numbers at present). R, G, B
The image data is output from the scanner 10. Scanner 1
The shading correction 181 of the output I / F 12 of 0 corrects the unevenness of the brightness level due to the unevenness of illumination (irregularity of illumination) in the main scanning direction and the unevenness of the photoelectric conversion characteristic.

In the scanner γ conversion 193, the scanner 10
The shading-corrected image data of each color is converted from reflectance (luminance) data (luminance) to brightness data. In the separation generation 192, the character portion and the photograph portion are determined, and the chromatic color and achromatic color are determined.

The filter 194 performs a so-called MTF filter process, and performs edge enhancement, smoothing, etc. according to the user's preference such as a sharp image or a soft image.
In addition to the process of changing the frequency characteristic of the image data, the edge enhancement process (adaptive edge enhancement process) according to the edge degree of the image data is performed. For example, so-called adaptive edge enhancement in which edge enhancement is performed on a character edge and edge enhancement is not performed on a halftone image is performed on each of the RGB image data.

Referring again to FIG. 6, in the color correction 202, the difference between the color separation characteristics of the input system (10) and the spectral characteristics of the color material of the output system (100) is corrected, and the color required for faithful color reproduction is corrected. Color correction processing for calculating the amount of material YMC, and U for replacing the portion where the three colors of YMC overlap with Bk (black)
It consists of CR processing.

The printer γ conversion 205 corrects the image data according to the image quality mode detected by the separation generation 192 such as characters and photographs and the image forming characteristics of the printer 100. Also, in the copy mode in which high image quality output is designated, background removal and the like are performed at the same time. Printer γ conversion 20
Reference numeral 5 denotes a plurality of γ conversion tables (lookup table L that can be switched according to the image area separation result (text / photo).
UT).

In the gradation processing 206, multi-value gradation image data is converted into binary or extremely small gradation (minor gradation) recording image data (quantized data).

FIG. 7 shows an outline of the internal structure of the IPP.
The IPP receives image data from the output I / F 10 of the color original scanner 10 and inputs / outputs image data to / from the CDIC at the input / output port 141. The input / output image data is once stored in a data buffer including a bus switch and a buffer memory or the input / output data register 142, and then input to the processor array 144 via the memory control of the SIMD type processor 143 or output to the CDIC. It The data for controlling the IPP and the image processing program (the program and the processing parameters) of the IPP are given from the system controller 1, CDIC or the process controller 131 to the host buffer 147 and written in the data RAM 146 and the program RAM 145.

FIG. 8 shows the structure of a part of the inside of the processor array 144. In this embodiment, the processor array 144 uses SIMD (Single Instruction stream Mul).
tiple Data stream) type processor, which executes a single instruction in parallel for a plurality of data and includes a plurality of processor elements PE1 to PE8 and the like. Each processor element PE (hereinafter simply referred to as PE) stores an input / output register 150 for storing data addressed to itself, a multiplexer (hereinafter referred to as MUX) 151 for accessing a register of another PE, and a barrel shifter ( Shift Expand) 152, logical operation unit (A in the following)
LU) 153, an accumulator (A) 154 for storing a logical result, and a temporary register (F) 155 for temporarily saving the contents of the accumulator.

Each input / output register 150 is connected to an address bus and a data bus (read line and word line), and stores an instruction code defining a process, data to be processed, and the like. The contents of the input / output register 150 are input to the logical ALU 153, and the arithmetic processing result is stored in the accumulator 154. In order to take the result out of the PE, it is temporarily saved in the temporary register 155. By taking out the contents of the temporary register 155,
The processing result for the target data is obtained.

The instruction code is given to each PE in the same content, the data to be processed is given in a different state for each PE, and the adjacent P
By referring to the contents of the E input / output register 150 in the MUX 151, the operation result is processed in parallel and output to each accumulator 154.

For example, if the contents of one line of image data are arranged in the PE for each pixel and arithmetic processing is performed with the same instruction code, the processing result for one line can be obtained in a shorter time than sequential processing for each pixel. can get. In particular, in the spatial filter processing and the shading correction processing, the instruction code for each PE is an arithmetic expression itself, and the processing can be performed commonly to all PEs.

FIG. 9 shows a functional configuration of the special processing 148 (FIG. 7) for performing the sequential image data processing. The function illustrated in FIG. 9 is for performing an error diffusion process, and the sub-scanning held by the SIMD processor 143 for the error diffusion in the sub-scanning direction is applied to the pixel data of the pixel included in the current pixel line. Image data to which the error diffusion value in the direction is added is input. The error diffusion processing of the special processing 148 quantizes the added value (added value data) of the input image data and the error data. The input image data is the image data of the target pixel to be processed this time, and the error data is the error data of the pixel processed before the target pixel.

The error diffusion processing of the special processing 148 is performed by the error data calculation 1481 and the multiplexer 148 that selects one from the error data calculated by the error data calculation 1481.
4 and error data addition 1485 for processing the error data selected by the multiplexer 1484 and adding it to the data input from the SIMD type processor 143.

Further, in the sequential image data processing of the special processing 148, the logic circuit 1483 for inputting the signal necessary for selecting the error data to the multiplexer 1484 and the preset error diffusion mode (binary error diffusion, There is provided an error diffusion processing hardware register group 1482 capable of setting information as to which of three-value error diffusion and four-value error diffusion should be used for error diffusion and a calculation coefficient used for the error diffusion processing. Further, the sequential image data processing of the special processing 148 is performed by the blue noise signal generator 14
The error diffusion processing hardware register group 1482 is also provided with 86 to determine whether or not to use blue noise for the error diffusion processing.
It is configured to be selectable by the setting of.

The SIMD type processor 143 adds the error diffusion value in the sub-scanning direction held for error diffusion in the sub-scanning direction to the pixel data of the pixel included in the current pixel line, and adds it to the input image data. 1487 adds the error diffusion value in the main scanning direction. The error data calculation 1481 calculates error data which is a difference between the image data to which the error diffusion value is added and the predetermined threshold value. Error data calculation 148
1 is three quantization reference value storage units 3a, 3c, 3e,
It is provided with three comparators 3b, 3d, 3f and three threshold value tables 2a to 2c.

In this embodiment, the quantization reference value storage unit 3a,
The comparator 3b and the threshold value table 2a operate as a set. Further, the quantization reference value storage unit 3c, the comparator 3d, and the threshold value table 2b operate as one set, and the quantization reference value storage unit 3e, the comparator 3f, and the threshold value table 2 are operated.
c operates as one set. The threshold table 2a stores a dither threshold for determining whether the density is high, the threshold table 2b stores a dither threshold for determining whether the density is medium, and the threshold table 2c is a density low. It stores a dither threshold for determining whether or not it is. For the threshold values read from these threshold value tables 2a to 2c, after the blue noise is subtracted, the quantization reference value storage units 3a, 3c, 3
e and the comparators 3b, 3d, 3f.
The output of the blue noise signal generator 1486 is "0".
When blue noise is not added in, the threshold table 2a
The threshold values read from the data storage devices 2a to 2c are directly stored in the quantization reference value storage units 3a, 3c and 3e and the comparators 3b and 3d,
Given to 3f.

The comparators 3b, 3d and 3f input threshold values from the respective connected threshold value tables 2a to 2c. Then, each generates a binary signal indicating whether the added value data output by the addition 1485 is equal to or more than a threshold value, and outputs the binary signal to the logic circuit 1483. The subtraction 1488 subtracts each threshold value from the added value data and outputs each subtracted value data to the multiplexer 1484. Create image data. As a result, a total of three error data are simultaneously input to the multiplexer 1484.

When the blue noise is used for the error diffusion processing, the blue noise signal generator 1486 turns on and off the blue noise data at a relatively high cycle to generate the blue noise. The thresholds are the comparators 3b, 3d,
It is subtracted from the blue noise before inputting into 3f.
By the processing using blue noise, it is possible to prevent the image from becoming monotonous by giving an appropriate variation to the threshold value.

The threshold value tables 2a to 2c store threshold value groups having different values. In the present embodiment, among the threshold value tables 2a to 2c, the threshold value group having the largest threshold value table 2a is stored, and the threshold value table 2b and the threshold value table 2c are stored in the order of decreasing threshold values. The quantization standard value storage units 3a, 3c and 3e read out from the threshold value tables 2a to 2c and the comparators 3b and 3b.
The respective threshold values given to d and 3f are latched.

The logic circuit 1483 includes a comparator 3b,
The binary signals (1 bit data) output from each of 3d and 3f are encoded into 2 bit data, output as quantized data, and input to the multiplexer 1484. The multiplexer 1484 selects any one of the three error data as the error data of the target pixel according to the quantized data. The selected error data is
In order to reflect the quantization of the subsequent lines in the sub-scanning direction,
SIMD type processor 143 processor array 144
It is input to the RAM of the local memory 142 via the processor element (PE).

Further, the quantized data output from the logic circuit 1483 is input to one of the PEs of the processor array 144 of the SIMD type processor 143. In this embodiment, the image data is data represented by 2 bits of upper bits and lower bits.

The selected error data is also input to the error data addition 1485. Error data addition 1485
Is three in the main scanning direction with respect to the pixel indicated by the circled numbers 1, 2, and 3 shown in FIG. 10A, that is, the pixel of interest (small square with a downward-sloping leftward diagonal). The error data of the previously processed pixel (denoted as error data 3 in error data addition 1485 in FIG. 9), the error data of the previously processed pixel (denoted as error data 2 in FIG. 9), one The error data of the previously processed pixel (denoted as error data 1 in FIG. 9) is stored.

The error data addition 1485 is the error data 3
Is multiplied by a calculation coefficient of 0 or 1. In addition, the error data 1 is multiplied by 1 or 2 which is the operation coefficient to obtain the error data 1
Is multiplied by a calculation coefficient of 2 or 4. Then, the three multiplication values are added, and this value (weighting error data) is added to the addition value data input next from the SIMD processor 143. As a result, a pixel located closer to the target pixel has a greater effect on the error diffusion processing of the target pixel, and the error of the pixel can be appropriately diffused to form an image closer to the original image.

Note that the above-described quantized data is created in the sequential image data processing of the special processing 148 by using a configuration generally called an IIR type filter system. Below, the image data of a pixel is called pixel data.

FIG. 10B shows the arithmetic function of the IIR type filter system. The mathematical expression calculated by this calculation function is as follows: ODn = (1−K) × ODn−1 + K · IDn (1) ODn: Quantized density after calculation (quantized data) ODn−1: Quantization of the previous pixel It is expressed as density IDn: current pixel data K: weighting factor.

As is clear from the equation (1) and FIG. 10B, the quantized density ODn after calculation is obtained from the quantized density ODn-1 of the previous pixel and the value of the current pixel data IDn. To be Generally, the IIR type filter system is a dedicated circuit for performing a so-called sequential conversion, in which an operation is performed on a current pixel by using an operation result using a pixel processed before the current pixel. The sequential image data processing by the special processing 148 of the image processing apparatus according to the present embodiment should be used for the overall sequential conversion as shown in FIG. 10B regardless of the functional configuration shown in FIG. You can

FIG. 11A shows the SI in IPP.
An outline of the functional configuration of the gradation processing 206 (FIG. 6) by the cooperation of the MD processor 143 and the special processing 148 is shown. This gradation processing receives multi-gradation image data and outputs the quantized data. The error data calculation 1481, the error data addition 1485, etc. of the special processing 148 and the image feature extraction by the SIMD type processor 143 are performed. (Edge detection 143a, area expansion processing 143b), and a signal delay 142 for timing adjustment between the quantization processing in the special processing 148 and the image feature extraction in the SIMD type processor 143. The signal delay 142 is provided as needed, and is, for example, a line memory having a required number of lines.

The input multi-tone image data is 8-bit / pixel data read by the scanner 10 at 600 dpi, for example. Generally, such multi-tone image data is input to the printer image quality processing 200 (FIG. 6) after passing through a smoothing filter (for example, 194 in FIG. 6) to smoothly express halftones. Usually 150L
Since it is smoothed from an image period of about pi (Line Pair per Inch), 175 Lpi used in gravure printing or the like.
The above-described periodic component of the high frequency halftone dot image does not remain in the input image data of the gradation processing 206.

The error diffusion process, that is, the quantization process in the special process 148 shown in FIG. 11A is performed by the quantization threshold generation (dither threshold generation 2) in the error data calculation 1481.
a to 2c, multiplication 4a, addition 4b), the multi-gradation image data is quantized by the error diffusion method using the quantization threshold value generated by the quantization threshold value.
Error data calculation 1481 including the quantizer 4c, error calculation 1488, error storage 1485a, error diffusion matrix 1
485b and error addition 1487. Note that FIG.
The multiplication 4a, the addition 4b, and the quantizer 4c, which are surrounded by a two-dot chain line in (a) of 1, are comparators 3b and 3 shown in FIG.
The function of the combination of d and 3f and the logic circuit 1483 is shown, and the subtraction 1488 of FIG. 11A shows the combination of the multiplexer 1484 and the subtraction 1488 shown in FIG.

In FIG. 11A, the input image data is adjusted in timing by the signal delay 142 and input to the error addition 1487. The image data to which the diffusion error has been added by the error addition 1487 is input to the quantizer 4c. The quantizer 4c quantizes the input image data using the quantization threshold value given by the addition 4b, and outputs the quantization result as quantized data.

In this embodiment, 2-bit error diffusion processing is performed. That is, 2-bit quantized data is output. More specifically, dither threshold generation (2a-
2c), multiplication 4a and addition 4b are combined to generate quantization thresholds 1 to 3 (th1 to th3). The mutual relation between the quantization threshold values th1 to th3 is the quantization threshold value 1 (th
1) ≤ quantization threshold 2 (th2) ≤ quantization threshold 3 (th
3). The quantization threshold value 1 (th1) is one of the threshold value group (for example, DT1 in FIG. 14) generated based on the data group of the threshold value table 2c, and the quantization threshold value 2 (th).
2) is one of a threshold value group (for example, DT2 in FIG. 14) generated based on the data group of the threshold value table 2b,
The quantization threshold 3 (th3) is, for example, the threshold table 2
a threshold group generated based on the data group of a (see FIG.
One of four DT3).

The quantizer 4c (3b, 3d, 3f) compares the image data to which the diffusion error is added with th1 to th3, generates a binary signal indicating whether the image data is larger than a threshold value, and outputs the logic circuit. Encodes these three binary signals into 2-bit quantized data. That is, if the image data to which the diffusion error is added is larger than th3, it is "3", and if it is less than th3 and larger than th2, it is "2".
Is equal to or smaller than th2 and larger than th1, "1" is set to th
When it is smaller than 1, 2-bit quantized data representing “0” is output.

The error calculation 1488 is performed by three kinds of quantization errors of the quantizer 4c, "image data-th which is added with diffusion error-th.
1 ”,“ image data with diffusion error-th2 ”, and“ image data with diffusion error-th3 ”are calculated. Then, from the multiplexer 1484, if the quantized data is "3", "the image data with the diffusion error added-
th3 "error data, if the quantized data is" 2 "," image data with diffusion error added-th2 "error data, if the quantized data is" 1 "," image with diffusion error added " The error data "data-th1" and the error data "image data to which diffusion error is added" when the quantized data is "0" are output.

The calculated error data is stored in the error storage 1485a.
Temporarily store in. The error storage 1485a is for storing the quantization error regarding the processed pixels around the target pixel. In the present embodiment, since the quantization error is diffused up to the peripheral pixels two lines ahead as described below, the line memory of three lines stores the error storage 1485, for example.
Used for a.

The error diffusion matrix 1485b calculates the diffusion error to be added to the next pixel of interest from the quantized error data stored in the error storage 1485a. In this embodiment, the error diffusion matrix 1485b is (b) in FIG.
3 pixels in the sub-scanning direction y (vertical direction in the figure) as shown in
Diffusion error data is calculated using an error diffusion matrix having a size of 5 pixels in the main scanning direction x (horizontal direction in the drawing). In FIG. 11B, the asterisk * corresponds to the position of the next pixel of interest, a, b ,. ． ． , K, l are coefficients (total sum is 32) corresponding to the positions of 12 peripheral processed pixels.
In the error diffusion matrix 1485b, a value obtained by dividing the product sum of the quantization errors for these 12 processed pixels and the corresponding coefficients a to l by 32 is given to the error addition 1487 as the diffusion error for the next pixel of interest.

The image feature extraction by the SIMD type processor 143 is performed by edge detection 1 as shown in FIG.
43a and area expansion processing 143b. Edge detection 1
43a is for detecting the edge of the input image data,
In this embodiment, level 0 (maximum edge degree) to level 8
Outputs 4-bit edge data representing edge levels up to (non-edge). More specifically, for example, by using four types of 5 × 5 differential filters DF1 to DF4 shown in FIG. 12, ± 4 from the main scanning direction, the sub-scanning direction, and the main scanning direction.
The edge amount is detected in four directions inclined by 5 °, the edge amount having the maximum absolute value is selected, and the absolute value of the edge amount is quantized into four level levels from level 0 to level 3. Output. Area expansion processing 143b
Performs a region expansion process with a width of 7 pixels on the edge detected by the edge detection 143a. The edge data output from the edge detection is referred to and the 7
The minimum edge level (maximum edge degree) in the area of 7 pixels (3 pixels before and after the main scanning direction, 3 pixels before and after the sub scanning direction) is set as the edge level of the pixel of interest, Output as edge data. This edge data is the multiplication 4a of the error data calculation 1481 (see FIG.
A).

The dither threshold generation (2a ...
The multiplication 4a multiplies the dither threshold data output by 2c) by the coefficient corresponding to the edge data, and the addition 4b adds +128 to the product data of the multiplication. The sum data of this addition is given to the quantizer 4c as a quantization threshold value (for example, FIG. 14).

The multiplication 4a generates a quantization threshold value that periodically oscillates in the image space with a vibration width corresponding to the edge level represented by the edge data output from the area expansion processing 143b. In the present embodiment, the dither threshold value generation (2a to 2c) in FIG. 11A is 1 as shown in FIG.
Using a 4 × 4 dither threshold matrix arranged so as to grow the lines from 1 to 6 (1 is the minimum and 6 is the maximum), the dither threshold that periodically oscillates from 1 to 6 in the image space is output. . The numerical values shown in (a) of FIG. 13 are sequential numerical values, not dither thresholds. The dither threshold at a certain position of 1 is the minimum value, and a larger threshold is assigned as the order numerical value becomes larger. Dither thresholds of the same value are used for the same ordinal numerical values. The dither threshold period is
This is 168 Lpi in the case of 600 dpi image formation.
Equivalent to. Such dither threshold generation (2a to 2c)
Is composed of a ROM that stores the dither threshold matrix and a counter that counts main and sub-scanning timing signals of image data and generates a read address of the ROM.

The multiplication 4a shown in FIG. 11A uses coefficient 3 when the edge level indicated by the edge data from the image feature extraction (143) is level 0 (non-edge), and coefficient 2 when level 1. , Level 1 with coefficient 1 and level 3
At the time of (maximum edge degree), the coefficient 0 is multiplied by the output value of dither threshold generation (2a to 2c).

Here, by arranging the dither threshold values corresponding to the order numerical value of 1 in FIG. 13A in the main scanning direction x, dots having two pixels arranged in the main scanning direction x are first formed. For example, it represents PM1 in FIG. in this way,
With the intention that stable dot formation is performed, two 1-value pixels that are writing levels with low energy are arranged.
The screen angle and the line growth direction in this case are shown in FIG. The line growth direction is shown in "Line growth direction 1" in the figure.

The quantized data processed as described above is output to the printer 100 to print out an image. As a result, the resolution of letters, image change points and halftone dot images with a relatively low number of lines is good, and that of photos, images with few changes and halftone images with a high number of lines is smooth and stable. Is good
It is possible to print out a high-quality image in which those areas are aligned without any discomfort. The basis for this will be described below.

A change such as an edge portion of a character or a line drawing in the image is sharp and the edge level is level 3 (the edge degree is the highest).
In the portion where is, the quantization threshold generated by the generation of the medium and high density quantization thresholds (2a to 2c, 4a, 4b) has a coefficient value of 0 given by the area expansion processing 143b, and therefore the output of the multiplication 4a. Becomes zero, but in addition 4b, +128
Since a fixed value (FIG. 18) different from the above is given and the quantization processing by a pure (principle) error diffusion method using a fixed threshold value is performed, an image with good resolution can be formed.

In a portion having a low edge degree such as a flat portion of a photograph or an image, quantization threshold generation (2a to 2c, 4a, 4) is performed.
Since the oscillation width of the quantization threshold generated in b) becomes large, the quantization processing of the quantization processing (1481) is a dither-based processing, and the image data is halftone-dotted at the dither threshold cycle. Since the dither threshold matrix in which the thresholds are arranged in the order shown in FIG. 13A is used to generate the quantization threshold, as the density level of the image data increases,
The output dots grow spirally from the central portion of the line grown along the screen angle within the dither threshold period.

FIG. 13B shows the order of line growth as another modification of the example of FIG. 13A, that is, a second example. This indicates the order of arrangement of the dither thresholds, and the dither threshold at a certain position of 1 is the minimum value, and the larger threshold value is assigned as the order numerical value increases. The relationship between the order in which the lines grow and the dither threshold is shown. The numerical values indicating the order are the data values of the dither threshold generation (2a to 2c) of FIG. Although the order of the pixels formed when the line grows is different from the case of FIG. 13A, even if such a value is set, a line along the screen angle is formed and a smooth image is formed. Can be obtained.

When reading a document image with a scanner, the main scanning is normally performed from left to right and the sub-scanning is performed from top to bottom, as indicated by the thin arrows in FIG. Are input in the order in which they are read, the direction of the quantization processing, that is, the propagation direction of the quantization error is the direction from the upper left to the lower right as indicated by the thick arrow in (d) of FIG. In this case, 4 pixels ×
Since the quantization threshold as shown in FIG. 14 is generated in the output of the addition 4b within the 4-pixel dither threshold cycle, as the density level increases from the low level, the order as shown in FIG. Output dots occur. That is, the output dots grow so that the line becomes thicker from left to right in the main scanning direction x.

FIG. 15 shows how the output dots (recording dots) are generated in the low density portion, the medium density portion and the high density portion of the image. As seen in FIG. 15, from low to medium concentration,
The recorded image grows in a line along the screen angle.
Therefore, it is possible to form a smooth image in which the flatness of the image flat portion of low and medium density is good. In addition, in the low density area,
As shown in PM1, the output dots (recording dots) first grow in a direction in which two pixels are arranged in the main scanning direction, and subsequently grow in a line shape along the screen angle. That is, in the low-density portion, dot growth is a horizontal tone in which two pixels are preferentially grown side by side in the main scanning direction x. Such laterally-based dot growth improves the reproducibility of the toner image, especially when an electrophotographic printer is used for image formation, as compared with a case where two pixels are arrayed in the sub-scanning direction y with one value. However, the roughness of the image is reduced, and a high-quality image having excellent stability can be formed.

In the present embodiment, the quantization parameter is changed so that the number of lines is increased as the edge degree increases in the boundary portion between the area having a large edge degree and the area having a small edge degree, and the dither-based processing is changed to the error diffusion-based processing. The characteristics of the quantization processing can be smoothly switched to the processing or vice versa. Therefore, it is possible to form an image in which the boundary portions of both image regions are aligned without a feeling of strangeness.

16 to 18, edge degree 1 (closer to photograph; main subject to photograph; halftone image) to 3 (closer to letter; main subject to character;
An example of the quantization threshold (the output of the addition 4b) of the line drawing) is shown respectively. The quantization threshold values of edge degrees 1 and 2 shown as an example in FIGS. 16 and 17 are 268 Lpi as the number of lines.
And the quantization threshold value of edge degree 3 shown in FIG. 18 is the number of lines of 300 Lpi in the low image density region,
In the medium to high image density area, the error diffusion processing is performed.

FIG. 19A shows, as a third example, an example of the order of threshold arrays for growing recording dots in a line along different screen angle directions. Numerical values indicate the order of dither threshold arrangement. The dither threshold at a certain position of 1 is the minimum value, and a larger threshold is assigned as the order numerical value becomes larger. The screen angle in this case is shown in FIG.
(B) of. FIG. 20 shows an example of the quantization threshold (output of addition 4b).

In FIG. 21A, as a fourth example, 4 × 4
An example of the order in which the lines grow in which the quantization threshold is the smallest among the pixels is shown. The numerical values in this order also indicate the arrangement order of the dither threshold. The dither threshold at 1 is the minimum value,
As the numerical value increases, a larger threshold value is assigned. FIG. 21B shows the screen angle in that case. FIG. 22 shows an example of the quantization threshold value (output of addition 4b), and FIG. The following is a description of the occurrence of.

Also in this way, a smooth image can be obtained. Further, in the case of an optical system using two beams in which latent images are formed alternately with different beams for each main scan, even if either one of the two beams is changed first for each sheet output, There is an advantage that the influence of the beam output difference is reduced.

As shown by thin arrows in FIG. 24A, the main scanning is performed from right to left and the sub-scanning is performed from top to bottom to read the original image, and the image data is input in the order of reading. In this case, the direction of the quantization process, that is, the propagation direction of the quantization error is the direction from the upper right to the lower left as shown by the thick arrow in (a) of FIG. In this case, in the image flat portion, within the dither threshold period of 4 pixels × 4 pixels, as shown in FIG.
The quantization threshold shown in FIG. 25 that grows in the order as shown in FIG. The numerical values in this order also indicate the arrangement order of the dither threshold. The dither threshold at a certain position of 1 is the minimum value, and a larger threshold is assigned as the order numerical value becomes larger.

Here, the order in which the pixels grow is that the first to third pixels are common in comparison with FIG. 13A.
4-6 are different. By setting in this way, in the case where the scanning is performed from right to left in the main scanning direction x, that is, (c) of FIG. 13, (b) of FIG. 21, and (b) of FIG. As shown in "Growth direction 2", output dots are formed so that the line grows from right to left in the main scanning direction x, and a smooth image can be obtained.

FIG. 26 (a) shows an example in which the order of the quantization thresholds is set in 4 × 4 pixels so that the first dots formed are 2 pixels in the sub-scanning direction. The numerical values in this order also indicate the arrangement order of the dither threshold. The dither threshold at a certain position of 1 is the minimum value, and a larger threshold is assigned as the order numerical value becomes larger. FIG. 26B shows an example of the quantization threshold value (output of addition 4b). An example of dot growth with low image density, medium image density, and high image density,
It is shown in FIG.

[0137]

The gradation processing for converting multi-gradation image data into low-gradation quantized data by the error diffusion method also
It is possible to obtain quantized data that can easily reproduce a smooth image.

[Brief description of drawings]

FIG. 1 is a front view showing an overview of a multi-function full-color copying machine according to an embodiment of the present invention.

2 is an enlarged vertical cross-sectional view showing an outline of an image forming mechanism of the printer 100 shown in FIG.

3 is a block diagram showing a system configuration of image data processing of the copying machine shown in FIG.

FIG. 4 is a block diagram showing an outline of a functional configuration of the image data interface control CDIC shown in FIG.

FIG. 5 is a block diagram showing an outline of a functional configuration of an image memory access control IMAC.

FIG. 6 is a block diagram showing an outline of a functional configuration of an image data processor IPP.

7 is a block diagram showing an outline of a functional configuration of the image data processing device IPP shown in FIG.

8 is a block diagram showing an outline of a functional configuration of a processor array 144 shown in FIG.

9 is a block diagram showing a functional configuration of quantization processing of multi-value gradation image data using error diffusion processing into low gradation data in the special processing 148 shown in FIG. 7.

FIG. 10A is a plan view schematically showing pixel divisions on an image, in which a small square with a diagonal line to the left is a pixel to be quantized (pixel of interest), circled 1, 2, 3 Is a preceding pixel that gives a quantization error to the pixel of interest. (B) is a block diagram showing a functional configuration for performing an error diffusion processing calculation.

11A is a functional division of the SIMD type processor 143 and the special processing 148 shown in FIG. 7 at the time of quantization processing of multi-value gradation image data using error diffusion processing to low gradation data. It is a block diagram showing an outline of. (B)
Is the pixel of interest indicated by * in the error diffusion matrix and a, b ,. ． ． FIG. 3 is a plan view schematically showing k, l.

FIG. 12 is an edge detection 143a shown in FIG.
6 is a plan view showing filter coefficients of a differential filter used for calculating an edge amount in FIG.

13A is a first dither threshold (dither threshold matrix) of the dither threshold generators 2a to 2c shown in FIG. 11A in the order of arrangement from a small value to a large value.
An example plan view, (b) is a plan view showing a second example, (c)
FIG. 3A is a plan view schematically showing pixel divisions and screen angles on an image, and FIG. 3D is a plan view showing a scanning direction (thin line arrow) of image reading and a direction of error diffusion processing (thick line arrow) on a document. is there.

14 is a plan view showing an example of a quantization threshold given to the quantizer 4c in FIG. 11 (a).

FIG. 15 is an image represented by quantized data when low density image data, medium density image data and high density image data having the same value are quantized with the quantization thresholds DT1, DT2 and DT3 shown in FIG. PM1, PM2 and PM3
FIG.

16 is an image feature extraction 143 shown in FIG. 11 (a).
When the edge degree 1 is given to the multiplication 4a, the quantizer 4
It is a top view which shows the example of the quantization threshold value given to c.

FIG. 17 is an image feature extraction 143 shown in FIG.
When the edge degree 2 is given to the multiplication 4a, the quantizer 4
It is a top view which shows the example of the quantization threshold value given to c.

FIG. 18 is an image feature extraction 143 shown in FIG.
When the edge degree 3 is given to the multiplication 4a, the quantizer 4
It is a top view which shows the example of the quantization threshold value given to c.

19A is a third dither threshold (dither threshold matrix) of the dither threshold generators 2a to 2c shown in FIG. 11A in the order of arrangement from a small value to a large value.
FIG. 3B is a plan view showing an example, and FIG. 3B is a plan view schematically showing a pixel section on an image and a screen angle.

20 is a plan view showing an example of a quantization threshold value given to the quantizer 4c of FIG. 11 (a).

21A is a fourth dither threshold (dither threshold matrix) of the dither threshold generators 2a to 2c shown in FIG. 11A in the order of arrangement from a small value to a large value.
FIG. 3B is a plan view showing an example, and FIG. 3B is a plan view schematically showing a pixel section on an image and a screen angle.

22 is a plan view showing an example of a quantization threshold value given to the quantizer 4c of FIG. 11 (a).

FIG. 23 is a diagram showing quantization thresholds DT16 and DT1 shown in FIG.
7 is a plan view showing images PM4, PM5, and PM6 represented by quantized data when low density image data, medium density image data, and high density image data having the same value are quantized by 7 and DT18.

FIG. 24A is a plan view showing the scanning direction of image reading (thin line arrow) and the direction of error diffusion processing (thick line arrow) on a document. 11B is a plan view showing a fifth example of the arrangement order of the dither thresholds (dither threshold matrix) included in the dither threshold generations 2a to 2c shown in FIG. 11A from the smallest value to the largest value. .

25 is a plan view showing an example of a quantization threshold given to the quantizer 4c of FIG. 11 (a).

26A is a sixth dither threshold (dither threshold matrix) of the dither threshold generators 2a to 2c shown in FIG. 11A in the order of arrangement from a small value to a large value.
It is a top view which shows an example. FIG. 11B is a plan view showing an example of the quantization threshold given to the quantizer 4c in FIG.

FIG. 27 is a quantization threshold value DT2 shown in FIG.
2, the images PM7 and P represented by the quantized data when low density image data, medium density image data and high density image data of the same value are quantized by DT23 and DT24.
It is a top view which shows M8 and PM9.

[Explanation of symbols]

10: Color manuscript scanner 20: Operation board 30: Automatic manuscript feeder 80: Power supply 90: Motherboard 100: Color printer PC: Personal computer PBX: Exchanger PN: Communication line 102: Optical writing unit 103, 104: Paper feed cassette 105 : Registration roller pair 106: transfer belt unit 107: fixing unit 108: paper discharge trays 110M, 110C, 110Y, 110K: photoconductor units 111M, 111C, 111Y, 111K: photoconductor drums 120M, 120C, 120Y, 120K: developing device 160: Transfer / Conveying Belt ACP: Image Data Processor CDIC: Image Data Interface Control IMAC: Image Memory Access Control IPP: Image Data Processor

Claims (6)

[Claims] 1. Threshold generating means for generating a quantized threshold value that periodically oscillates, and gradation processing means for converting multi-gradation image data into quantized data of low gradation by an error diffusion method,
In the image processing apparatus, the quantization is performed so that an image recorded according to the quantized data grows along a screen angle within a specific cycle in an image space as the density level of the multi-tone image data increases. An image processing apparatus, comprising: a threshold value generation means for generating a threshold value. 2. The conversion to the quantized data is performed from the upper left to the lower right in the image space, and the quantization threshold is increased substantially from the left to the right within the specific period. The image processing apparatus according to claim 1, which is switched. 3. The conversion to the quantized data is performed from the upper right to the lower left in the image space, and the quantization threshold is switched so as to be substantially increased from the right to the left within the specific period. The image processing apparatus according to claim 1, wherein: 4. In the low density level region of multi-tone image data, the quantization threshold is switched so that output dots are preferentially grown in the main scanning direction; The image processing device described. 5. In the low density level region of multi-tone image data, the quantization threshold value is switched so that output dots are preferentially grown in the sub-scanning direction; The image processing device described. 6. An image processing apparatus according to claim 1, and an image is formed on a transfer sheet based on the quantized data output by the gradation processing means of the image processing apparatus. An image forming apparatus including a printer.