JP2002216112A - Dot printing inspection device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a dot printing inspection device with hardiness even for unevenness of illumination and its method in the case of inspecting a fault of dot omission of a dot printing surface. SOLUTION: The dot printing inspection device is provided with a line sensor to photograph the dot printing surface and to acquire an image, a filter image creating means to create a maximum value filter image and a minimum value filter image from the acquired original image, a differential image creating means to create a differential image between the maximum value filter image and the minimum value filter image, a binarized image creating means to create a binarized image from the differential image by a prescribed threshold and a featured value analyzing means to analyze featured value of a dot omission area based on the binarized image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dot printing inspection apparatus and method for inspecting a dot printing surface.

[0002]

2. Description of the Related Art As a method of inspecting a halftone dot printing surface, conventionally,
The halftone dot printing of a predetermined density is performed on a suitable printing paper, and the operator enlarges the halftone dot with a loupe or a magnifying microscope, etc., finds a dot missing area over the entire surface to be inspected, and manually plots the dot. The quality and quality of halftone printing were evaluated by analyzing the distribution and number of missing dots. Such manual operation is not preferable because the burden on the eyes of the operator and the time required for the operation increase.

In order to solve such a problem, Japanese Patent Publication No. 4-2104 discloses a dot printing evaluation apparatus for automatically inspecting a dot printing surface by image processing. This device binarizes image data obtained by photographing a printing surface into two types, one with a halftone dot and one without a halftone dot, using a predetermined threshold value, and searches for a position where a halftone dot should be based on the binary image. Then, a plurality of pixel points are designated in a predetermined area including the position, and the number of halftone dot data obtained at the plurality of points is compared with a set value. When the value is less than the set value, an operation of determining a defective halftone dot is performed for each position where the halftone dot should be, and an operation of counting the total number of halftone dots determined to be defective halftone dots is performed.

[0004]

In the above-mentioned known technique, the presence or absence of a halftone dot is determined from a binarized image obtained by simply binarizing an original image. To obtain this binarized image, it is necessary to set a threshold, but it is generally not easy to determine the threshold stably. For example,
An illumination device that illuminates a halftone dot printing surface has temporal or spatial non-uniformity in terms of its illumination intensity. If there is such unevenness, it is difficult to uniquely determine the threshold value.

In the above-mentioned known technique, the threshold value is determined between the peak value of the white level and the peak value of the black level in the density histogram of the original image. The difficulty in determining the threshold is described in the above-mentioned known technology.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a halftone dot printing inspection apparatus which is robust against uneven illumination when inspecting a dot defect on a halftone dot printing surface. And a method.

[0007]

In order to solve the above-mentioned problems, a dot printing inspection apparatus according to the present invention comprises: an image pickup means for picking up an image by photographing a dot printing surface; Filter image creating means for creating a value filter image and a minimum value filter image; difference image creating means for creating a difference image between the maximum value filter image and the minimum value filter image; A binarized image generating unit for generating a binarized image according to the above, and a characteristic amount analyzing unit for analyzing a characteristic amount of the dot missing area based on the binarized image. . The operation and effect of this configuration are as follows. (1) First, an image is obtained by photographing a halftone dot printing surface as an object by an imaging unit. The imaging means is a line sensor,
Appropriate devices such as an area sensor and an image scanner can be adopted.

(2) A maximum value filter image and a minimum value filter image are created from the obtained original image. The maximum value filter process is a process of replacing the density value of the target pixel with the density value of the maximum density in the filter region, and is a process generally performed in image processing. The same applies to the minimum value filter processing. If there is unevenness in the lighting device,
A similar effect appears on the maximum value filter image and the minimum value filter image.

(3) Next, a difference image between the maximum value filter image and the minimum value filter image is created. Taking the difference cancels out the effects of the non-uniformity of the lighting device. When the influence of the non-uniformity of the lighting device is small, a similar image can be obtained only by the minimum value filter image. Whether to use only the minimum value filter image or the difference image between the maximum value filter image and the minimum value filter image can be selected according to the degree of non-uniformity of illumination in the shooting environment to be used. .

(4) A binary image is created from the difference image by a predetermined threshold value. The threshold value may be a preset value or a value obtained in real time each time.

(5) Based on the above-mentioned binarized image, the characteristic amount of the missing area is analyzed. As a method of analyzing feature values,
An appropriate method may be used. For example, analysis and evaluation can be performed based on the number of dot-missing regions and the area of the dot-missing region.

As described above, it is possible to analyze the characteristic amount of the spot missing area with the illumination non-uniformity of the illumination device removed. As a result, it is possible to provide a halftone dot printing inspection apparatus that has robustness against uneven illumination when inspecting a dot printing defect on a halftone dot printing surface.

As a preferred embodiment of the present invention, the size of the maximum value filter and the minimum value filter is substantially equal to the arrangement period of the halftone dots. When setting the size of the filter, if it is too large, it may not be possible to detect missing dots. Therefore, it is possible to reliably detect missing dots by making them substantially match (including perfect match) with the arrangement cycle of the halftone dots.

As another preferred embodiment of the present invention, when any of the center value, the mode, and the center of gravity of the density histogram of the difference image is a, the standard deviation is σ, and the proportional constant is k, A threshold value setting means for obtaining the threshold value by the expression (a + kσ). The distribution of the density histogram of the difference image is expected to be close to a normal distribution. On the other hand, the distribution peculiar to the missing area occurs at a position apart from the original normal distribution. Therefore, by setting the threshold value based on the above equation, the threshold value can be set at an intermediate density position between the original normal distribution curve and the dot missing area. By binarizing the difference image with the threshold value, the missing area can be reliably detected.

[0015] In still another preferred embodiment of the present invention,
A threshold value setting unit that sets an image obtained by offsetting the neighborhood average processed image of the difference image by a predetermined amount as the threshold value may be used. What is neighborhood averaging?
This is a process of replacing the density value of the target pixel with the average density value of all pixels in the filter area. In the difference image, it is necessary to set a threshold value at an intermediate level between the density level of the dot missing area and the level of the other area. By taking a neighborhood average, it is possible to set the intermediate level. Note that the offset is performed to set the threshold to a more appropriate intermediate level.

In order to solve the above-mentioned problems, a dot printing inspection method according to the present invention comprises the steps of: capturing an image by printing a dot printing surface; and obtaining a maximum value filter image and a minimum value filter from the obtained original image. Creating an image; creating a difference image between the maximum value filter image and the minimum value filter image; creating a binary image from the difference image with a predetermined threshold value; Analyzing the feature value of the halftone dot missing area based on the digitized image.

The operation and effect of this configuration are as already described in the above (1) to (5). As a result, it is possible to provide a halftone dot printing inspection method that is robust against uneven illumination when inspecting a dot dot defect on a halftone dot printing surface.

[0018]

Preferred embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic diagram illustrating a configuration of a halftone dot printing inspection device.

<Structure of Halftone Printing Inspection Apparatus> A printed material 2 having a halftone printing surface to be inspected is conveyed to a predetermined position by a conveyor 3. An illumination light source 5 and a line sensor 6 (corresponding to an imaging unit) are installed in the photographing box 4, and photograph an image of a halftone printing surface of the printed matter 2. Note that the imaging means is not limited to a specific structure, but may employ an appropriate type such as an area sensor or an image scanner. Image data captured by the line sensor 6 is sent to the image analysis device 7.

Halftone printing is used for gravure printing and the like, and the density of the halftone dots represents the density contrast.
When the halftone dot printing surface is enlarged, portions where the dots that should originally exist are missing (these defects are referred to as “dot missing”) are seen. The halftone printing inspection apparatus according to the present invention evaluates such halftone printing.

In order to perform such an evaluation, the image analysis device 7
Incorporates a program for performing image processing on the captured image data. Hereinafter, the image analysis device 7
The following describes the functions provided by the above, but this function can be realized mainly by an image processing program.

FIG. 2 is a block diagram mainly showing each function realized by the image processing program of the image analyzer 7. The image memory 70 is a memory that stores original image data acquired by the line sensor 6. The filter image creation means 71 creates a maximum value filter image and a minimum value filter image from the original image. The maximum value filter is a filter often used in image processing, and is a filter for replacing the density value of the target pixel with the density value having the highest density among the pixels in the filter area. The minimum value filter is a filter for replacing the density value of the target pixel with the density value having the lowest density among the pixels in the filter area.

The difference image creation means 72 has a function of obtaining a difference image between the maximum value filter image and the minimum value filter image obtained by the filter image creation means 71. The binarized image creating unit 74 creates a binarized image from the difference image obtained by the difference image creating unit 72 using a predetermined threshold. The threshold value setting means 73 sets a threshold value for creating the binarized image.

The characteristic amount analyzing means 75 analyzes the characteristic amount of the dot missing area based on the binarized image obtained by the binarized image generating means 74, and performs quality inspection and evaluation of halftone printing.

<Dot Printing Inspection Procedure> Next, a procedure for inspecting halftone printing by the image analyzer 7 will be described.
First, FIG. 3 shows a part of an original image obtained by photographing a halftone dot printing surface. In the drawing, dots are shown by ●, which are regularly arranged at a predetermined pitch P. The area indicated by a broken line in the figure indicates the size of the filter area of the maximum value filter and the minimum value filter. The size of these filters is
It almost matches (or completely matches) the arrangement period of the halftone dots. The area indicated by R in the figure is a state in which halftone dots are printed correctly, and the area indicated by Q is a state in which halftone dots are completely missing (dot missing state). It is preferable that the size of each pixel of the line sensor 6 be as small as possible when photographing such a dot printing surface.

FIG. 4 is a diagram showing a procedure of image processing.
FIG. 4A shows image data of a cross section of the line L in an original image obtained by photographing the halftone printing shown in FIG. The vertical axis is
Shows the photometric value. Halftone dots are indicated by ●, and the photometric value becomes smaller. When a halftone dot exists, a predetermined concave portion is seen in the original image. The area where the halftone dot is missing is indicated by the dotted line B
There is no concave portion which should originally exist as shown in FIG. Also,
Looking at the original image, a large sinusoidal undulation W is seen, which is due to the spatial non-uniformity of the illumination light source 5.

Next, a maximum value filter image of the original image is created. This is shown in FIG. The concave portion due to the existence of the halftone dot disappears from the waveform. The waveform has undulations due to unevenness of the illumination light source. Further, a minimum value filter image is created, which is shown in FIG. The portion of the dot missing area described above is detected as the protruding area C. As in (b), a similar swell due to the non-uniformity of the illumination light source remains. If there is no unevenness of the illumination light source, it is possible to extract only the protruding region C by binarizing the minimum value filter image by a predetermined threshold value. If so, it is difficult to detect a missing area only from this minimum value filter image.

Next, a difference image is created by taking the difference between the maximum value filter image and the minimum value filter image. This difference image is shown in FIG. In creating the difference image, the difference between the maximum value filter image and the minimum value filter image may not be immediately adopted as the difference image, but may be offset by a predetermined amount.

In the maximum value filter image and the minimum value filter image, undulations due to non-uniformity of the illumination light source were observed. However, by taking the difference, the undulations can be canceled. This is because the non-uniformity of the illumination light source appears in the maximum value filter image and the minimum value filter image as undulations having a similar shape. By creating such a difference image, it becomes possible to inspect the halftone dot printed surface regardless of whether or not the illumination light source is uneven.

Next, a binary image is created by binarizing the difference image with a threshold value. The setting of the threshold will be described. As the threshold value, for example, a fixed value obtained experimentally in advance may be used, but in real time (dynamically) based on the actually captured original image.
It is preferable to obtain it because more accurate inspection can be performed.

FIG. 5 is a diagram showing a density histogram of the difference image shown in FIG. The horizontal axis indicates the density value, and the vertical axis indicates the number of pixels. In FIG. 5, the histogram curve is divided into a large peak S1 having a substantially normal distribution and a small peak S2 distributed in a portion having a low density value. This small mountain S2 is due to the existence of the dot missing area.

First, a is the central value of the large peak S1 of the histogram (mode or barycentric value may be adopted), and σ is the standard deviation of the large peak S1. Set the value obtained in step as a threshold. k is an appropriate proportional constant, and this value can be a fixed value obtained experimentally, or a value manually adjusted for each test object. In FIG. 5, the proportionality constant k is set to −3, and the density value indicated by the broken line is obtained as the threshold value.

FIG. 4D shows this threshold value, which is indicated by a broken line position Y. By binarizing the difference image using this threshold value, a missing area (indicated by D) can be detected.

Another method for obtaining the threshold will be described. A neighborhood average image of the difference image shown in FIG. The neighborhood averaging process is a process of replacing the density value of the target pixel with the average density value of all pixels in the filter area. Note that the neighborhood average is also called a moving average. In this case, as to how close the average value is to be, it is preferable to take a neighborhood average several times or more the arrangement period of the halftone dots. Also, where there is no change in density, even if the neighborhood average is taken, the density value will be the same as that of the pixel of interest, so it is preferable to offset the obtained neighborhood average processed image by a predetermined value.

FIG. 4 shows the threshold values obtained by the neighborhood averaging process.
This is indicated by the broken line Z in (e). It can be easily understood that the dot missing area D can be reliably detected also by this threshold value.

Using the threshold value set as described above, a binary image is obtained by binarizing the difference image. With this binarized image, a missing area can be detected. The feature amount of the detected dot missing area is analyzed. The feature amount includes the number of dot-missing areas, a map of the dot-noting areas (list of coordinate values), the area of the dot-notifying areas, and the density of the dot-noting areas, and these can be used to evaluate halftone dot printing. . Note that the density of the dot missing area can be evaluated by, for example, dividing the area to be inspected into m × n blocks and determining the number of missing dots in each block.

<Embodiment> FIGS. 6 and 7 show actual examples of the inspection procedure described above. FIG. 6 shows an image obtained by performing image processing on an image obtained by photographing an actual halftone dot printing surface. (A) is an original image, (b) is a maximum filter image, (c) is a minimum filter image, (d) is a difference image, and (e) is a binarized image. Since the binarized image contains a noise component, the noise component is removed by appropriately performing expansion and contraction processing on the binarized image, as shown in FIG. The present invention is not limited to the case where the binarized image is directly used for the analysis, but also the case where the binarized image is analyzed using an image subjected to processing for removing a noise component from the binarized image. Analyzing the quantity ”. FIG. 7 shows the missing dot extraction image shown in FIG. 6, and this image is divided into 5 × 5 blocks and the feature amount is analyzed. In the column of the missing spot position information, the number of missing areas (seven), their coordinates and the area (the number of pixels) are shown. In addition, the density of missing dots is shown, and the number of crowds in each block is shown. The evaluation of the halftone dot printing surface can be performed based on these feature amounts. For example, when the area of the dot missing area is larger than a preset value, it can be determined that printing is defective.

It should be noted that a reverse character (a numerical value indicating a product type, etc.) of 40% is found in the image, but this is not an evaluation target. By removing the outside extraction region, erroneous detection can be prevented.

<Another Advantage of the Present Invention> Although the inspection apparatus according to the present invention has already been described as being robust against non-uniformity of the illumination device, it has another advantage, which will be described below. . As described in the section of the related art, in the related art, it is necessary to specify a position where a halftone dot should be. For this purpose, data (a periodic function-type histogram) obtained by integrating the binarized image data (the density value of 0 or 1) in each of the x direction and the y direction is obtained. Each peak of this function represents a row of halftone dots. As shown on the right side of the arrow in FIG. 8A, by obtaining a waveform in which peaks and valleys can be clearly identified, rows and columns of halftone dots can be specified.

However, when the printed matter as the subject is photographed in an inclined state, the halftone dot image is also inclined as shown in FIG. 8B. Then, peaks and valleys do not appear in the waveform of the obtained periodic function histogram, and the position where the halftone dot should be cannot be specified. Therefore, in the related art, if the image is photographed in a state where the image is slightly inclined, there is a problem that the position of the halftone dot cannot be specified, and the inspection that should be performed cannot be performed. FIG.
Although it is inclined by about 3 °, it is difficult to identify a halftone dot even with such an inclination. On the other hand, in the present invention, the inspection is performed by setting a square (or rectangular) filter area as shown in FIG. Therefore, if there is a state in which a halftone dot other than the halftone dot of interest does not enter the inside of the filter area, the inspection can be performed normally. Therefore,
As shown in FIG. 9 (b), even if the inclination is as large as about 30 ° (in practice, the inclination is not so large),
Since the adjacent halftone dots do not enter the filter area, the inspection can be performed.

As described above, the inspection according to the present invention also has robustness against image inclination.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a configuration of a dot printing inspection apparatus.

FIG. 2 is a block diagram illustrating functions of the image analysis device.

FIG. 3 is a diagram illustrating a part of an original image obtained by photographing a halftone dot printing surface;

FIG. 4 is a diagram showing a procedure of image processing by the image analysis device.

FIG. 5 is a diagram showing a density histogram of a difference image.

FIG. 6 is a diagram showing an actual example of an inspection result.

FIG. 7 is a diagram showing an example of an inspection result;

FIG. 8 is a diagram for explaining another problem of the related art.

FIG. 9 is a view for explaining that the inspection according to the present invention has robustness against image inclination.

[Explanation of symbols]

 2 Printed matter 6 Line sensor 7 Image analyzing device 71 Filter image creating means 72 Difference image creating means 73 Threshold setting means 74 Binary image creating means 75 Feature amount analyzing means

 ────────────────────────────────────────────────── ─── Continuing from the front page (72) Inventor Toshiyasu Nishioka 3-1-198 Kasuganaka, Konohana-ku, Osaka-shi, Japan CA07 CA12 CA16 CB07 CB12 CB16 CC01 CE06 CE12 CH08 CH09 CH18 DA03 DA08 DA17 DC01 DC04 5C077 LL11 MP02 PP01 PP47 PP58 RR02 RR11 TT08

Claims (2)

[Claims] An image capturing unit that captures an image by printing a halftone dot printing surface; a filter image creating unit that creates a maximum value filter image and a minimum value filter image from the acquired original image; A difference image creating means for creating a difference image between an image and the minimum value filter image; a binary image creating means for creating a binary image from the difference image by a predetermined threshold value; And a feature value analyzing means for analyzing a feature value of the dot missing area based on the dot pattern inspection apparatus. 2. A step of capturing an image by printing a halftone dot printing surface to obtain an image; a step of creating a maximum value filter image and a minimum value filter image from the acquired original image; Creating a difference image of the value filter image; creating a binarized image from the difference image with a predetermined threshold value; analyzing a feature amount of a dot missing area based on the binarized image And a dot printing inspection method.