JP2008117248A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor simplifying smoothing of binary image data by reducing a collation frequency. <P>SOLUTION: An input pattern data selection part 21 sequentially selects a collation area with 3×3 pixels from the image data to be processed and inputs it into a collation part 22, and the collation part 22 sequentially collates an inputted monochrome pattern and a plurality of monochrome pattern images stored in a matching pattern memory 24. When one monochrome pattern is matched an output selection part 23 reads an output pattern image corresponding to the detected monochrome pattern from an output pattern memory 25 and outputs it. At that time, patterns having all white pixels and all black pixels are registered as the monochrome pattern used for the first collation, the collation is completed in the early stages of collation processing when an area has all the white pixels or all the black pixels having high appearance frequency, so that the overall collation frequency can be reduced. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing apparatus, and more particularly to an image processing apparatus having a function of performing a smoothing process by pattern matching on image data.

When binary image data read by the scanner is enlarged in the main scanning direction and sub-scanning direction, or when line density conversion is performed when the resolution of the facsimile reception image data is lower than the resolution of the printer, the enlarged image or the line density There is a step in the converted image. As one method for smoothing the stepped portion (jaggy) generated when the image data is enlarged or the line density is converted, smoothing processing by pattern matching is performed (for example, see Patent Document 1).
JP 2003-32495 A

  In the enlargement smoothing process by pattern matching, a black and white pattern in which a step is generated when the enlargement process is performed and a pattern image output when the enlargement process is performed on the black and white pattern are stored in advance in association with each other. If the image data to be processed includes a monochrome pattern image stored in advance, an output pattern image corresponding to the image data is read and output to obtain an image subjected to the enlarged smoothing process. .

  As described above, when performing an enlarged smoothing process on a binary image, a plurality of black and white patterns registered in advance in the pattern table are collated with the image to be subjected to the enlarged smoothing process, and the result of the matching is the same. In this case, an output pattern corresponding to the black and white pattern is output, but the fact that it does not apply to any of the black and white patterns registered in advance is compared with all the black and white patterns registered in the pattern table. I don't know until I try.

On the other hand, in general images, when the pattern reference area is all white pixels or all black pixels, it occurs quite often, but conventionally, black and white patterns such as all white pixels and all black pixels are registered in the pattern table. In the case of all white pixels or all black pixels, it was found that after matching with all registered black and white patterns sequentially in a predetermined order, they did not match either, all white pixels or all black pixels The pixel image, that is, the enlarged content is output as it is.
In this way, all white pixels or all black pixel reference areas that occur with considerable frequency are collated with all the black-and-white patterns stored in the pattern table, and it is necessary to perform a considerable amount of unnecessary collation. Because there was, there was a problem that processing did not proceed.

  The present invention has been made in view of the above problems, and provides an image processing apparatus capable of simplifying the processing by reducing the number of collations when performing smoothing processing of binary image data. Objective.

  In order to achieve the above-described object, an image processing apparatus according to a first aspect of the present invention includes a storage unit that stores a plurality of monochrome patterns in association with output patterns in advance, and stores the input monochrome pattern and the storage unit in advance. An image processing apparatus comprising: a collating unit that collates with a monochrome pattern that has been matched; and an output unit that outputs an output pattern corresponding to the black and white pattern when the collation results in the collation by the collating unit; As a monochrome pattern stored in advance in the storage means, a monochrome pattern of all white pixels and / or all black pixels is registered at a position referred to at an earlier stage of the collation.

According to a second aspect of the present invention, there is provided the image processing device according to the first aspect, wherein the black and white pattern of all white pixels and / or all black pixels is first referred to at the time of collation. It is registered,
According to a third aspect of the present invention, there is provided the image processing apparatus according to the first aspect, wherein a frequent black-and-white pattern is provided at a position referred to first at the time of collation as a black-and-white pattern stored in advance in the storage means. A black and white pattern of all white pixels and / or all black pixels is registered at a position to be registered and referred to next.

According to the image processing apparatus of the first to third aspects of the present invention, as a monochrome pattern stored in advance in the storage unit, a position referred to at an early stage of collation, for example, first referred to during collation. Or a black and white pattern of all white pixels and / or all black pixels is registered at a position referenced next to a frequent black and white pattern, or the pattern reference area is all white pixels or all black pixels Since the collation is completed at an early stage of the collation process, the total number of collations can be reduced.
Theoretically, when matching a pattern that does not match any pattern, the number of times of matching increases by two. However, when looking at the whole, most of the matching regions are often all white pixels or all black pixels. It is possible to reduce the number of collations as a whole.

Hereinafter, an embodiment in which the image processing apparatus of the present invention is applied to a digital multifunction machine having a copying function, a facsimile function, a printing function, a scanner function, and the like will be described.
FIG. 1 is a block diagram showing the hardware configuration of a digital multi-function peripheral. CPU 1, ROM (Read Only Memory) 2, RAM (Random Access Memory) 3, display / operation unit 4, reading unit 5, image memory 6, recording A unit 7, a codec 8, a modem 9, a network control unit NCU 10, a LAN interface (I / F) 11, and an enlarged smoothing processing unit 12 are connected to each other via a bus 13.

The CPU 1 controls each hardware unit of the digital multi-function peripheral via the bus 13 and executes various programs based on the programs stored in the ROM 2. The ROM 2 previously stores various programs and operation messages necessary for the operation of the digital multi-function peripheral, and includes a pattern storage area 14 for storing a monochrome pattern and a corresponding output pattern used for pattern matching during the enlarged smoothing process. Yes. Note that patterns of all white pixels and all black pixels are stored at the first read position of the monochrome pattern used for pattern matching in the pattern storage area 14.
The RAM 3 is composed of an SRAM or the like, and stores temporary data generated when the program is executed.

  The display / operation unit 4 includes a display unit that displays an operation state of the digital multifunction peripheral or displays an operation screen of various functions, and a plurality of keys for operating the digital multifunction peripheral. An original document reading table such as an auto document feeder (ADF) or a flatbed scanner (FBS) is provided. The original is read by a scanner using a CCD or the like, and dot image data is output.

  The image memory 6 is configured by using a DRAM or the like, and stores image data to be transmitted or received image data or image data read by the reading unit 5, and a recording unit 7 is a printer device such as an electrophotographic system. The received data, copy original data, or print data transmitted from an external personal computer or the like is printed out.

  The codec 8 encodes and decodes in accordance with a predetermined protocol. In order to transmit the image data of the read original, the codec 8 is encoded by the MH, MR, or MMR system, and the image data received from the outside is decoded. Also, encoding and decoding are also performed in correspondence with the TIFF method, which is an image format generally used as a file that can be attached to an e-mail.

The modem 9 is connected to the bus 13 and has a function as a fax modem capable of facsimile communication. The modem 9 is also connected to the NCU 10 connected to the bus 13. The NCU 10 is hardware that closes and opens an analog line, and connects the modem 9 to a public switched telephone network (PSTN) as necessary.
The LAN interface 11 is connected to the LAN, receives signals from the LAN, and transmits signals and data to the LAN, and executes interface processing such as signal conversion and protocol conversion.

  Further, as shown in the functional block diagram of FIG. 2, the enlarged smoothing processing unit 12 that performs the enlarged smoothing process by pattern matching includes an input pattern data selection unit 21, a matching unit 22, and an output selection unit 23. Is constituted by a CPU 1, a ROM 2, a RAM 3, and the like, and the functions of each unit are executed by a software program.

  The matching pattern memory 24 and the output pattern memory 25 are memories in the pattern storage area 14 of the ROM 2, and the matching pattern memory 24 stores a black and white pattern in which a step is generated when simple enlargement processing is performed. Reference numeral 25 denotes a pattern image output when the enlarged smoothing process is performed on the black and white pattern stored in the matching pattern memory 24 in association with each black and white pattern.

FIG. 3 is a diagram showing an example of a black and white pattern image stored in the matching pattern memory 24. A white square indicates a white pixel, and a black square indicates a black pixel.
As shown in FIGS. 3A and 3B, black and white patterns of all white pixels and all black pixels are registered at the first reference position in the matching pattern memory 24, and are subsequently referred to. A predetermined number of monochrome patterns as shown in FIGS. 3C and 3D are registered.

  FIG. 4 shows an example of an output pattern image. In the output pattern images shown in FIGS. 4C and 4D, the selected monochrome pattern matches the monochrome pattern shown in FIGS. 3C and 3D. It is a pattern image that is sometimes output. Such an output pattern image is stored in the output pattern memory 25 corresponding to each monochrome pattern stored in the matching pattern memory 24.

  Then, the input pattern data selection unit 21 sequentially selects a 3 pixel × 3 pixel verification region from the image data decoded by the codec 8 and inputs it to the verification unit 22, and the verification unit 22 receives the input pattern data from the input pattern data selection unit 21. By sequentially collating the input monochrome pattern with a plurality of monochrome patterns stored in the matching pattern memory 24, the matching pattern memory 24 is searched for a monochrome pattern that matches the input monochrome pattern.

  If the input monochrome pattern matches any of the plurality of monochrome patterns stored in the matching pattern memory 24, the collation unit 22 sends a signal to the output selection unit 23, and the output selection unit 23 detects the detected monochrome pattern. Are read out from the output pattern memory 25 and output. If none of the monochrome patterns match, the output selection unit 23 enlarges and outputs the input monochrome pattern as it is.

  The digital multi-function peripheral has the above-described configuration. When the facsimile is transmitted, the image data of the original is read by the reading unit 5, compressed by the codec 8, and stored in the image memory 6. The compressed image data is read from the image memory 6, modulated by the modem 9, and transmitted from the NCU 10 to the communication partner through the PSTN. At the time of facsimile reception, the received image data is demodulated by the modem 9 and stored in the image memory 6 and then decoded by the codec 8 as it is or after being subjected to the enlargement smoothing processing by the enlargement smoothing processing unit 12 and then recorded. Printed by part 7.

Next, the operation when the CPU 1 implements the function of the above-described enlarged smoothing processing unit will be described with reference to the flowchart of FIG.
When the program for the enlargement smoothing process shown in the flowchart of FIG. 5 is started, the CPU 1 first selects the first 3 pixels × 3 pixels processing target from the image data decoded by the codec 8 (step 101), and then the ROM 2 The first monochrome pattern is extracted from the pattern storage area 14 (step 102), and it is determined whether or not they match by comparing the selected processing target data with the extracted monochrome pattern (step 103).
If it is determined that the selected processing target data matches the extracted black and white pattern, the CPU 1 replaces the output with the output pattern corresponding to the extracted black and white pattern stored in the pattern storage area 14 of the ROM 2 (step). 104).

On the other hand, if it is determined that the processing target data selected in step 103 does not match the extracted black and white pattern, the CPU 1 determines whether or not collation has been completed for all the black and white patterns stored in the pattern storage area 14 of the ROM 2. If it is determined (step 105) and it is determined that collation with all the black and white patterns has not been completed, the next black and white pattern is extracted from the pattern storage area 14 of the ROM 2 (step 106). The processed processing data is compared with the extracted monochrome pattern.
If it is determined in step 105 that collation has been completed for all the monochrome patterns stored in the pattern storage area 14 of the ROM 2, the CPU 1 enlarges the processing target data as it is and outputs it (step 107). .

  Then, after outputting an output pattern corresponding to the matched black and white pattern in step 104, or after outputting the processing target data as it is in step 107 as it is, the CPU 1 determines whether or not collation of the entire image has been completed. (Step 108), if it is determined that the verification of the entire image has not been completed, the next 3 pixel × 3 pixel processing target data is selected from the image data decoded by the codec 8 (Step 109), and similarly. When the pattern matching process is executed and it is determined in step 108 that the entire image has been collated, the enlargement smoothing process is terminated.

  The enlargement smoothing process is executed as described above. Since the pattern of all white pixels shown in FIG. 3A is first extracted as the monochrome pattern used for pattern matching, the image data decoded by the codec 8 is extracted. The processing target data of 3 pixels × 3 pixels selected from the above are first collated with the pattern of all white pixels, and if they match, the pattern of all white pixels enlarged twice is output.

If all the processing target data of 3 pixels × 3 pixels does not match the pattern of white pixels, then the pattern is checked with the pattern of all black pixels shown in FIG. A pattern of all black pixels enlarged twice is output.
In addition, when the processing target data of 3 pixels × 3 pixels does not match the pattern of all white pixels or all black pixels, it is sequentially compared with the subsequent monochrome patterns stored in the pattern storage area 14 of the ROM 2, If it matches any of the monochrome patterns, an output pattern corresponding to the matched monochrome pattern is output from the pattern storage area 14 of the ROM 2, and if it does not match any of the monochrome patterns, the collation target is doubled as it is. Is enlarged and output.

  As described above, the pattern of all white pixels and all black pixels is registered in the matching pattern memory as the black and white pattern used for collation first. Since the collation is completed at an early stage of the collation process and the total number of collations can be reduced, the processing time of the expanded smoothing process can be shortened.

  In the above-described embodiment, the white and black patterns of all white pixels and all black pixels are added to the black and white pattern in which a step is generated when the simple enlargement process is performed. However, since it is considered that the probability of matching the pattern of all white pixels is higher, only the pattern of all white pixels may be added.

  In the above embodiment, the black and white pattern of all white pixels and all black pixels is registered as the black and white pattern to be referred to at the time of the first collation. It is also possible to register such a frequent pattern and register a monochrome pattern of all white pixels and all black pixels as a monochrome pattern referred to after the frequent pattern matching.

  Further, for a plurality of black and white patterns in the matching pattern table used for collation, the reference order can be changed according to the document type setting, for example, character, photo, character / photo, etc. For a plurality of patterns in the table, the reference order may be changed manually.

In addition, a learning function for each user can be provided, and the reference order of patterns in the matching pattern table used for matching can be changed based on the learning result.
That is, the matching result (history) when pattern matching is actually performed for each user is stored, the reference order is determined according to the number of times (frequency) of actual matching for a plurality of patterns, and the number of times of matching If collation is performed in the order of patterns having a large number of patterns, the collation time can be further shortened.

  In addition, it is possible to set a pattern (non-reference pattern) that is not used for collation for a plurality of patterns in the table. In this way, the image quality decreases, but the more non-reference patterns are processed Can be simplified and the verification time can be reduced.

Further, in the above-described embodiment, the example in which the present invention is applied to the enlarged smoothing process has been described. However, the present invention can also be applied to the linear density conversion or the simple image smoothing process, and the matching target The number of pixels is not limited to 3 × 3, and an arbitrary number of pixels can be set, and the enlargement magnification is not limited to two, and an arbitrary magnification can be set.
In the above embodiment, the image processing apparatus of the present invention is applied to a digital multi-function peripheral. However, the image processing apparatus of the present invention is also applicable to other image processing apparatuses such as a facsimile machine and a copier. can do.

2 is a block diagram illustrating a hardware configuration of a digital multifunction peripheral. FIG. It is a functional block diagram which shows the function of an expansion smoothing process part. It is an example of the monochrome pattern memorize | stored in the pattern memory area of ROM. It is an example of the output pattern memorize | stored in the pattern memory area of ROM. It is a flowchart which shows the effect | action at the time of execution of an expansion smoothing process.

Explanation of symbols

1 CPU
2 ROM
3 RAM
4 Display / Operation Unit 5 Reading Unit 6 Image Memory 7 Recording Unit 8 Codec 9 Modem 10 NCU
11 LAN I / F
12 Enlarged Smoothing Processing Unit 13 Bus 14 Pattern Storage Area 21 Input Pattern Data Selection Unit 22 Verification Unit 23 Output Selection Unit 24 Matching Pattern Memory 25 Output Pattern Memory

Claims (3)

Storage means for storing a plurality of monochrome patterns in association with output patterns in advance, collation means for collating an input monochrome pattern with a monochrome pattern stored in advance in the storage means, and collation by the collation means As a result, an image processing apparatus including an output unit that outputs an output pattern corresponding to the black and white pattern when they match,
An image in which a monochrome pattern of all white pixels and / or all black pixels is registered as a monochrome pattern stored in advance in the storage means at a position referred to at an earlier stage of the collation. Processing equipment. The image processing apparatus according to claim 1,
An image processing apparatus, wherein a black and white pattern of all white pixels and / or all black pixels is registered at a position to be referred to first at the time of collation. The image processing apparatus according to claim 1,
As a monochrome pattern stored in advance in the storage means, a frequent monochrome pattern is registered at the first reference position at the time of collation, and a monochrome pattern of all white pixels and / or all black pixels is registered at the next reference position. An image processing apparatus.

Images