JPH0380771A - Image processor - Google Patents

Abstract

PURPOSE:To prevent the resolution of an image from being lowered by discriminating whether or not the arranging pattern of the maximum value or the minimum value of an image signal coincides with a pattern set in advance, and correcting a discrimination signal by detecting the periodical appearance of the discrimination signal in main and sub scanning directions. CONSTITUTION:As a first step, the maximum value or the minimum value is detected by remarking the fact that the maximum value and the minimum value of gradation data is dispersed periodically and regularly from the characteristic of a dot image. Twelve patterns can be predicted as the arranging pattern of an extreme value as shown in figure. When the arranging pattern coincides with one of the patterns, it is designated as the candidate of a dot. And a picture element at an left end is set as the candidate, and is sent to a main scan dot detecting part 51 as a signal A. At an erroneous judgement eliminating part 52 at the next stage, only one middle picture element enclosed with every non-dot picture element is eliminated. Next, in correction in the subscan direction, a certain number of lines between a dot line and the next dot line is corrected when it exists similarly as in the main scan.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing device applied to a digital copying machine.

[Conventional technology]

In the above-mentioned image type t2 apparatus, the moiré generated in the reading system 2 image processing system is caused by the period of the halftone original, the sampling pitch of reading, or the pitch of a group of pixels serving as a unit of image processing.

Conventionally, in order to prevent this, moiré has been prevented by applying smoothing processing, error diffusion processing, etc. to all images.

[Problem to be solved by the invention]

However, for text images and solid photographic images that are not involved in moire generation, these processes have the disadvantage that the resolution of the images is reduced, and even solid photographs and characters in the photographs become blurred by these processes.

On the other hand, although various techniques for distinguishing between text areas and photo areas have been proposed in the past when there is no halftone dot document and only text and solid photo areas, there is a problem that the addition of halftone dots complicates the technology. Ta.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image processing apparatus that can reliably detect a halftone dot document area in a halftone document with a period in which moiré may occur in a reading device.

[Means to solve the problem]

The above purpose is to improve image signal processing in an image processing apparatus applied to an image forming apparatus that reads document image information using a photoelectric conversion element such as a CCD sensor, processes this image signal, and outputs it to a recording device such as a printer. a discriminating means for detecting an array pattern of local maximum value or local minimum value pixels and determining whether this array pattern matches a preset pattern; This is achieved by comprising a correction means for detecting the appearance and correcting the discrimination signal.

[Effect]

It is determined whether the arrangement pattern of the maximum value or minimum value of the image signal matches a preset pattern, and the periodic appearance of this discrimination signal in the main scanning direction and the sub-scanning direction is detected to generate a discrimination signal. Correct.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 12 is a block diagram of the original reading device part of a digital copying machine to which the present invention is applied, in which a contact glass 1 on which a read original is placed is illuminated by light sources 2a and 2b, and an image of the read original is illuminated. The reflected light from the surface is imaged on the light receiving surface of the CCD image sensor 9 via the mirrors 3, 4, 5, 6.7 and the lens 8. Further, the light source 2 and the mirror 3 are mounted on a traveling body 11 that moves on the lower surface of the contact glass 1 in parallel to the contact glass 1.

Main scanning is performed by solid-state scanning of a COD (image sensor) 9. The document image is read one-dimensionally by the CCD 9, and the entire surface of the document is scanned by moving the optical system (sub-scanning).

In this example, the reading density is set to 16 pixels/mm for both main and sub-scanning.
It is possible to read originals up to 1.5 mm.

FIG. 13 is a block diagram of a document reading section of a digital copying machine according to the present invention, in which 9 is a CCD, 31 and 33 are amplifiers, 32 is a signal combining section, and 34 is an A/D conversion section.

The image signal read at a sampling density of 16 pixels/mm is as shown in the block diagram of FIG.
First, the amplifiers 31 and 33 amplify the voltage amplitude to a predetermined voltage amplitude, and then the A/D converter 34 converts it into digital data with one pixel number of gradations (64 gradations in the embodiment) (the number of gradations is 2). In this example, the A/D converted signal becomes a 6-bit signal, and is then processed by shading to correct uneven illuminance of the light source and variations in sensitivity between elements of the CCD 9. Make corrections.

After shading correction, MTF correction, spatial filter processing such as smoothing, scaling processing, write modulation processing, etc. are generally performed. The write modulation process is to match the output form of the output device such as a printer.If the printer is a one-pixel binary printer, it is a binarization process, and if the printer is a printer that can express multiple gradations per pixel, it is adjusted in advance with the printer. Modulate the signal into an agreed upon form.

In the present invention, as described above, after shading correction, halftone dots are detected and smoothing processing is first performed only on the halftone dots. The halftone dot density at this time is the density of the reading device of 16 dots/mm, and the target range is 1001/inch to 2001/inch where moiré occurs. If the density of halftone dots is coarser than that range, the reading device will be able to resolve it sufficiently and moiré will not occur, and if there are halftone dots that are denser than that range, the CC
This is because it is determined that moiré will not occur due to integration when reading D.

1 and 2 are block diagrams of each embodiment of the image processing apparatus according to the present invention, and in both figures, 40 is a shading correction section, 41 is a halftone dot detection section, 42 is a smoothing circuit,
43 is a selector. Also, in Figure 1, 44 is MT
This is an F correction section. On the other hand, in FIG. 2, 45 and 46 are processing units for text images and photo images, 47 is a text/photo area separation unit, and 48 is a selector.

As shown in these figures, halftone point detection is performed using the data after shading, and a selector 43 is used to switch in real time whether to detect the data after shading as it is or to output it after smoothing.

The halftone area is the output after smoothing.

FIGS. 5 fa to 5 d are explanatory diagrams of spatial filters. Various types of spatial filters can be considered at this time, and an example as shown in this figure is conceivable. In order to remove moiré, a size of approximately C1 is appropriate as shown in Figure 5 (al, (b), (b). At this point, the halftone image signal is equivalent to an image with halftones such as an ordinary solid photograph. can be handled.

Two examples of subsequent processing are shown in FIGS. 1 and 2.

In the example shown in Fig. 1, the MTF correction section 44 then performs MTF correction, and then outputs to the printer. It should be used when the same output process can be used to express both images.

MTF correction may be performed before selection, that is, immediately after shading correction. In that case, whether the signal used for halftone detection is after shading or after MTF correction depends on the halftone detection means, but in this example, either signal may be used.

The example shown in FIG. 2 is an example in which a character or line image and a photographic halftone image are processed separately. As mentioned above, at this point halftone images can be treated in the same way as photographic images, so it is possible to separate text images with no halftones and photographic images with many halftones, and apply appropriate processing to each. By outputting the image to a printer, the overall image quality can be improved.

Figure 3 (al, (bl) and Figure 4 (al, (bl
are image processing process diagrams of a binary printer and a multi-value (multi-gradation) printer, where (al is a process diagram for character images and (bl is a process diagram for photographic images), respectively.

First, when outputting to the binary printer shown in FIG. 3, the character image processing unit 45 uses the MTF as shown in FIG.
The image is corrected, and then magnified in the main scanning direction, and then binarized depending on whether it is larger or smaller than a threshold value. As shown in Figure 5(d), the photographic image processing unit 46 performs smoothing (in this case, using a small filter as shown in Figure 5(d)) and scales it to create a pseudo halftone expression. Therefore, it is binarized by dither processing.

In the case of the multi-gradation printer shown in FIG. 4, the character processing section 4
5, as shown in the same figure (al), filter processing is first performed using the spatial filter ■.This spatial filter ■ is a filter with frequency characteristics that gives priority to high frequencies in the sense of correcting the MTF as in the binary case. After that, the signal is scaled to match the output gradation of the printer, and a signal for direct driving is created, or as shown in FIG. 4, it is encoded and output according to the agreement with the printer. As shown in the same figure (bl), signal noise is first removed using a spatial filter ■.The filter ■ therefore has a frequency characteristic that suppresses high frequency components.Then, it is scaled and output to match the printer in the same way as in the case of characters. But in this case,
In order to obtain an image with better gradation, the printer performs area gradation, which expresses gradations using multiple gradations and multiple pixels.

Although there are processes suitable for text and photographs other than those shown in FIGS. 3 and 4, only two examples will be shown here.

Various methods have been devised in the past for separating characters from photographic images that do not include halftone dots.

Although not described in detail here, in the embodiment, the density gradient is detected two-dimensionally, and the processing unit 4 and one of the outputs obtained from the processing unit 46 is selected.

I mentioned zooming earlier in Figures 3 and 4, but the methods for zooming in this type of device are: (1) changing the magnification in the main scanning direction by changing the reduction ratio of the optical system including the lens; Changing the magnification in the sub-scanning direction by changing the optical system movement speed and changing the reading density in the sub-scanning direction. (2) Using the data at the same magnification, calculate the data at the sampling point during the magnification change by interpolation calculation. There is scaling in the main scanning direction, (same scaling in the sub-scanning direction as in 11), and (3) scaling by performing two-dimensional interpolation simultaneously in main and sub-scanning using data at the same magnification. In the example, (2) is used.

Further, although not shown, the γ characteristic, which is the input/output density characteristic, may be converted after scaling processing, after MTF correction, or the like.

Here, we return to the explanation of the halftone dot detection section. From the characteristics of halftone images,
Focusing on the fact that the maximum and minimum values of the gray scale data are periodically and regularly distributed, the first step is to detect the maximum or minimum value.

FIG. 6 is a pixel matrix pattern diagram, and the halftone dot detection method is to use the matrix as shown in FIG. The pixel of interest X0 is other than x, , x, , x3 .
If it is larger (smaller) than all of x, or equal to one of the four and larger (smaller) than the other three, Xo is determined to be the local maximum value. The maximum value (minimum value) signal is set to 1 for the maximum value (minimum value) and 0 for the others.

FIG. 7 is a halftone dot detection block diagram, in which 5o is a pattern comparison section, 51 is a main scanning halftone detection section, 52 is an erroneous judgment removal section, and 53 is a sub-scanning halftone correction section.

This circuit processes the process from the maximum value (minimum value) signal to the halftone dot signal. First, in one dimension in the main scanning direction, for 100I!/inch to 2001/inch, which targets the array of extreme value signals. Predict.

FIG. 8 is a diagram showing the arrangement pattern of extreme value signals, and 12 patterns as shown in this diagram are predicted. When this pattern matches, it is selected as a halftone dot candidate. There is also a method in which all the pixels of the matching pattern are used as halftone dot candidates, but in this example, only the leftmost pixel (in FIG. 8) is used as a candidate, and the signal A
It is sent to the main scanning halftone dot detection section 51 as a. At this time, if the pattern detected once among the 12 types of halftone dot patterns is extremely different from the pattern detected next, erroneous detection can be prevented by identifying each one and checking the continuity. Check for a match with the pattern at an extreme value, then check again for a match at the next extreme value. Non-extremities between the two candidates are not candidates for halftone dots, so if the number of pixels between the two candidates is within a certain number of pixels, the area between the two candidates is reviewed as a candidate. That is main scanning halftone detection. Changing the predetermined number of pixels depending on the type of pattern is also a measure to prevent false positives.

FIG. 9 is a block diagram from the extreme value signal to main scanning halftone detection, in which 60 is a pattern match circuit and 61 is a DQ flip-flop.

Due to the characteristics of halftone dot originals, the pattern does not appear in a very localized area, but should appear over a very wide range from the pixel level. The next erroneous determination removal unit 52 focuses on this point and reviews the pixels that are erroneously detected as non-halftone areas in the halftone signal and the isolated areas that are erroneously detected as halftone areas. Remove judgment.

FIG. 1O is a diagram showing an example of an erroneous judgment removal circuit,
0 is a shift register.

In the example shown in this figure, if eight pixels at the left and right ends of the 32 pixels in the main scan are continuous halftone dots, the central area surrounded by these is also a halftone dot. Note that the number of pixels, 32.8 (!), is not limited to this, but is determined experimentally.

At the same time, isolated points in the center are removed, but in this example, only one pixel surrounded by one non-halftone dot is removed, but from the non-halftone dots surrounding two or three pixels. It is also possible to improve accuracy by removing the inside. With this, the signal C is activated, halftone detection in the main scanning direction is completed, and then correction in the sub-scanning direction is performed. Since the halftone dots are spread two-dimensionally, the distribution of extreme values may or may not exist for each line. As in the case of main scanning, if the distance between one halftone dot line and the next halftone dot line is a certain number of lines, the gap between them is corrected.

FIG. 11 is a block diagram showing an example of a sub-scanning halftone correction section, and 71 is a line memory.

In this way, correction in the sub-scanning direction is performed using a large number of line memories 71.

In this way, the halftone dot signal is finally extracted, as shown in FIG.

It is applied as shown in Figure 2. Although this embodiment shows an example in which the smoothing process is applied to the halftone dot area, it is also an effective method to apply a process aimed at removing moiré, such as an error diffusion method, only to the halftone dot area.

The outline of a laser printer to which the present invention is applied will be described below.

FIG. 14 shows the constituent countries of the laser printer part of the digital copying machine to which the present invention is applied.
It is equipped with a laser writing system, image reproduction system, paper feeding system, etc. The laser writing system is a laser output crab unit 12. Imaging lens 13. It is equipped with a mirror 14. Inside the laser output unit 12, there is a polygonal mirror (which is rotated at a constant speed at high speed by a laser diode as a laser light source and a motor).
Polygon mirror). Laser light output from the laser writing system is irradiated onto a photosensitive drum 15 provided in the image reproduction system. Around the photosensitive drum 15,
Electric charger 16. Eraser 1f, development unit 18
．． Transfer charger 19, separation charger 201 separation claw 2
1. It is equipped with a cleaning unit 22 and the like. still,
A beam sensor (not shown) that generates a main scanning synchronization signal (MSYNC) is arranged near one end of the photosensitive drum 15 at a position that is irradiated with a laser beam.

The process of image reproduction will be briefly explained. Photosensitive drum 15
The surface of the battery is uniformly charged to a high potential by the charger 16. When that surface is irradiated with laser light, the potential of the irradiated portion decreases. Since the laser light is controlled on/off depending on whether the recorded pixel is black or white, a potential distribution corresponding to the recorded image, that is, an electrostatic latent image, is formed on the surface of the photoreceptor by irradiation with the laser light. When the portion on which the electrostatic latent image is formed passes through the developing unit 18, toner adheres to the portion according to the process of the potential, and a toner image that visualizes the electrostatic latent image is formed. A recording sheet is fed into the area where the toner image is formed at a predetermined timing and overlaps the toner image. This toner image is transferred to the recording sheet by the transfer charge roller 9,
It is separated from the photosensitive drum 15 by a separation charger 20 .

The separated recording sheets are conveyed by a conveyor belt 23, thermally fixed by a fixing roller 24 having a built-in heater, and then discharged to a paper discharge tray 25.

In the embodiment, there are two paper feeding systems. The one paper feeding system is equipped with a paper feeding cassette 26, and the other paper feeding system is equipped with a paper feeding cassette 27. The recording sheets in the paper feed cassette 26 are fed by paper feed rollers 28 . Recording sheets in the paper feed cassette 27 are fed by paper feed rollers 29 . The fed recording sheet temporarily stops in contact with the registration rollers 30, and then is moved to the photosensitive drum 1 with timing synchronized with the progress of the recording process.
Sent to 5. Although not shown, each paper feed system includes
It is equipped with a size sensor that detects the sheet size of the cassette.

I mentioned earlier that moiré may or may not occur depending on the density of the halftone dots, but this is also related to the magnification ratio.

That is, as enlarged copies are made, moiré becomes less likely to occur at 100 A/inch, and conversely, there is a possibility that moire will occur even at 2,001 inches or more. Furthermore, as the image size is reduced, moire appears even below 100 f/inch, and moiré no longer appears at around 175 Il/inch.

Therefore, in this embodiment, pattern generation can be made variable according to the magnification ratio. For example, 1 in Figure 8
A method of selecting from among 12 patterns is also conceivable.

Note that the discrimination means described in the claims is the pattern comparison section 50, and the correction means is the main scanning halftone dot detection section 51. Misjudgment removal unit 52. A sub-scanning halftone correction section 53 constitutes this.

〔Effect of the invention〕

As explained above, according to the present invention, it is possible to reliably detect the halftone dot document area with a period where moiré may occur. , it is possible to provide an image processing device that can prevent sharpness differences from decreasing.

[Brief explanation of drawings]

1 and 2 are block diagrams of each embodiment of the image processing apparatus according to the present invention, FIGS. 3 and 4 are image processing process diagrams of binary and multivalue printers, and FIG. 5 is an explanation of the spatial filter. Figure 6 is a pixel matrix pattern diagram, Figure 7 is a halftone detection block diagram, Figure 8 is an extreme value signal arrangement pattern diagram, and Figure 9 is a block diagram from extreme value signal to main scanning halftone detection. 10 is a circuit diagram for removing false judgments, FIG. 11 is a block diagram showing an example of a sub-scanning dot correction section, FIG. 12 is a block diagram of a document reading device, and FIG. 13 is a block diagram of a document reading section. , FIG. 14 is a block diagram of the laser printer. 50... Pattern comparison section, 51... Main scanning halftone dot detection section, 52... Misjudgment removal section, 53... Sub-scanning halftone dot correction section. No.! Figure 2 Figure 1 Figure 3 (σ) 5 (b) 6 N4 diagram (a an 1.. (b) Figure 5 (C) '!H6 diagram'jJl/vA

Claims (1)

[Claims] In an image processing device applied to an image forming apparatus that reads document image information using a photoelectric conversion element such as a CCD sensor, processes this image signal, and outputs it to a recording device such as a printer, the local maximum or minimum value of the image signal is A discriminating means detects an array pattern of value pixels and determines whether this array pattern matches a preset pattern, and a discriminating means detects the periodic appearance of a discriminating signal in the main scanning direction and the sub scanning direction. , and a correction means for correcting a discrimination signal.