JP2007024669A - Inspection device and inspection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an inspection device or the like capable of detecting with high accuracy the defects of printed matter. <P>SOLUTION: A plurality of patterns are arranged on an object to be inspected (printed matter or the like). This inspection device 1 divides an inspection object image into small regions, designates one small domain image, and extracts a similar small region image having close correlation with the small region image. The inspection device 1 divides furthermore the small region image and the similar small region image, and re-extracts a region image, having a high correlation coefficient in a sub-pixel unit, relative to the divided regions. Position correction of a divided small region image having a low correlation coefficient is performed by using the mean value of the correlation quantity of each divided small region image, having the high correlation coefficient. The inspection device 1 extracts a defect (dirt or chip) from the difference with a reference region relative to the divided small domain image after correction. The inspection device 1 can inspect with high accuracy defects in the inspection object having patterns. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an inspection apparatus, an inspection method, and the like for detecting defects (such as chipped defects, stain defects, blurring, and color tone defects) of printed materials. More specifically, the present invention relates to an inspection apparatus, an inspection method, and the like for detecting a defect such as a printed matter in which a plurality of identical patterns are arranged.

  Conventionally, the inspection for determining the quality of a printed material is performed by dividing an inspection object image obtained by imaging a printed material on which the same pattern or the like, which is the inspection object, is multi-faced, to generate a plurality of small area images. There is known a printed matter inspection method and apparatus for extracting a small region having a maximum correlation coefficient in each image, subtracting these small region images from each other, and detecting a defect of the printed matter from the calculation result (for example, [[ (See Patent Document 1]).

JP 2003-123056 A

  However, for a small area image with a small feature in the pattern of the printed matter, a small area image with a low correlation coefficient may be searched, and a small area image with a shifted pattern position may be extracted. At this time, a difference occurs between the small area images whose pattern positions are shifted, which is a cause of deterioration in defect detection accuracy. There is also a risk of erroneous detection of defects.

  In addition, since the presence / absence of a defect is determined based on the absolute value of the difference between the small area images, the type of defect (for example, dirt on the printed material, missing characters on the printed material, etc.) cannot be determined. was there.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide an inspection apparatus or the like that can detect defects of printed matter with high accuracy.

  In order to achieve the above-described object, the first invention is an inspection apparatus that performs image processing on an inspection image obtained by imaging an inspection object in which a plurality of identical arrangement units are arranged, and inspects the quality. A dividing unit that divides the inspection image into a plurality of small region images; an extraction unit that extracts a similar small region image having the highest similarity for each small region image in the inspection image; and the similarity is a predetermined value An inspection apparatus comprising: correction means for correcting the position of the similar small region image having the similarity smaller than the predetermined value based on the position of the larger similar small region image.

The inspection apparatus irradiates the inspection object with light, captures reflected light from the inspection object, acquires an inspection image, performs image processing on the inspection image, and determines whether the inspection object is good or bad.
The inspection object is, for example, a printed material. An inspection object can also be applied to a printing paper with a sticking material (such as a seal), a coating material (such as printing), or a vapor deposition material.
An arrangement unit indicates one unit such as a symbol or a character string. It is assumed that the same arrangement unit (design etc.) is arranged on the inspection object (printed article etc.) according to a predetermined rule.

The small area image is an image obtained by dividing the inspection image into predetermined areas. For example, the small area image is obtained as a plurality of rectangular area images by equally dividing the inspection image vertically and horizontally.
The similarity is an index indicating the similarity between image regions. For example, a correlation coefficient or the like is used as the similarity.
When there is a feature in the pattern of the small area image, the degree of similarity with the extracted similar small area image tends to be large. That is, the region image having a characteristic pattern has a tendency that the degree of similarity is larger than a predetermined value (for example, a correlation coefficient of 70%).
When there are few features in the pattern of the small area image, the degree of similarity with the extracted similar small area image tends to be small.

  The inspection apparatus according to the first aspect of the invention divides the inspection image into a plurality of small region images, and extracts a similar small region image having the highest degree of similarity with the small region image. Based on the position of the similar small region image whose similarity is larger than the predetermined value, the position of the similar small region image whose similarity is smaller than the predetermined value is corrected, and the small region image is compared with the similar small region image to be inspected. Judge the quality of the.

Further, the correction means performs correction using an average value related to the position of the similar small region image having a similarity higher than a predetermined value.
That is, the positional deviation of the small area image (especially, the similar small area image whose similarity is smaller than the predetermined value has a large tendency of positional deviation), for example, the small area image of the similar small area image whose similarity is larger than the predetermined value. It corrects using the average value of the relative position with respect to. By correcting the positional deviation, it is possible to improve the defect detection accuracy of the inspection image.

Further, the dividing unit may further perform at least one re-division, and the extracting unit, the correcting unit, and the determining unit may perform processing on the sub-region image and the similar sub-region image after the re-division.
By re-dividing the inspection image of the inspection object, the defect detection accuracy of the inspection image can be further improved.

The similarity may be calculated with subpixel accuracy.
The subpixel accuracy is, for example, 1/10 pixel. By calculating the similarity with sub-pixel accuracy, it is possible to further improve the defect detection accuracy of the inspection image.

In addition, regarding the determination of pass / fail of the inspection object, the determination unit of the inspection apparatus performs a difference process from the small region image to the similar small region image, binarizes with a predetermined threshold value, and stains the region exceeding the predetermined number of pixels. You may make it determine as a defect.
In addition, the determination unit of the inspection apparatus may perform a difference process on the small area image from the similar small area image, binarize with a predetermined threshold value, and determine an area exceeding a predetermined number of pixels as a defective defect.

  The extraction / correction means and the determination means may be provided in one processing device (computer), or the extraction / correction means and the determination means are provided in different processing devices (computers), so that the entire system is provided. The above-described inspection apparatus may be configured.

  A difference between image areas is calculated through a difference process, a binarization process, and the like based on pixel values (density values) of corresponding pixels between image areas. The pixel value (density value) is a value that is set to be larger as it is brighter. For example, in the RGB color system, the pixel value of the darkest place (black, etc.) is set as (0, 0, 0), and the pixel value of the brightest place (white, etc.) is set as (255, 255, 255).

The inspection apparatus performs, for example, a difference calculation for each of the corresponding pixels for each of the R component, the G component, and the B component, and sets the maximum value among the R component, the G component, and the B component of the difference pixel value. A binarization process is performed on the difference value, and a binary difference value as a difference is calculated.
The difference calculation is performed in the saturation mode, and when the difference calculation result is negative (minus), the value is set to “0” (0 clip).

That is, the inspection apparatus compares the pixels of the similar small region image with the pixels of the corresponding small region image, and determines a pixel having a lower (darker) density value (binary difference value) than that of the small region image. It is determined as a stain defect, and a pixel having a higher density value (binary difference value) (brighter) than the small area image is determined as a defect defect of the inspection object.
Note that by binarizing the pixel difference with a predetermined threshold, extra information is filtered to emphasize the features.
The inspection apparatus performs similar small region image extraction, correction, and determination processing on all the small region images of the inspection image, and determines the quality of the inspection object.

  According to the inspection apparatus of the first invention, the inspection image is divided or subdivided into small region images, and a similar small region image having a maximum correlation coefficient is extracted for each small region image. For small similar small area images, the coordinate position is corrected based on the average value of the coordinate positions of similar small area images having a large correlation coefficient, and each small area image is compared with the corresponding similar small area image after the correction. Therefore, the defect inspection of the inspection object can be performed with high accuracy.

A second invention is an inspection method for inspecting pass / fail by performing image processing on an inspection image obtained by imaging an inspection object in which a plurality of identical arrangement units are arranged, and the inspection image is divided into a plurality of small images. A division step of dividing the image into region images, an extraction step of extracting a similar small region image having the highest similarity for each of the small region images in the inspection image, and the similarity small region image having a similarity greater than a predetermined value. And a correction step of correcting the position of the similar small area image whose similarity is smaller than the predetermined value based on the position.
2nd invention is invention regarding the inspection method which the inspection apparatus of 1st invention performs.

  ADVANTAGE OF THE INVENTION According to this invention, the inspection apparatus etc. which can perform the defect detection of printed matter with high precision can be provided.

  DESCRIPTION OF EMBODIMENTS Hereinafter, preferred embodiments of an inspection apparatus and the like according to embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description and the accompanying drawings, the same reference numerals are given to components having substantially the same functional configuration, and redundant description will be omitted.

(1. Configuration of the inspection apparatus 1)
First, the configuration of the inspection apparatus 1 according to the embodiment of the present invention will be described with reference to FIG.
FIG. 1 is a schematic configuration diagram of an inspection apparatus 1.

  The inspection apparatus 1 includes a line camera 3, an illuminator 5, an image processing unit 9, a display unit 11, an image storage unit 13, an operation unit 15, and the like. The line camera 3, the illuminator 5, the image processing unit 9, the display unit 11, the image storage unit 13, the operation unit 15, and the like are connected to each other via a connection line such as a bus or a cable.

  The inspection apparatus 1 performs image processing such as comparison processing based on the small region image 31 and the similar small region image 35 (see FIG. 2 and the like), determines whether the printed material 7 to be inspected is acceptable, and displays the inspection result on the display unit. 11 is displayed.

  The small area image 31 is an area image obtained by equally dividing image data obtained by imaging the printed material 7 to be inspected (printed material image data 39: see FIG. 4) vertically and horizontally. The small area image 31 is a reference area image, and is sequentially used as an inspection reference image for the entire area of the printed image data 39.

  The similar small region image 35 is a small region image having the largest correlation coefficient with the small region image 31 that is the inspection reference image. Note that “high correlation” is synonymous with “high image similarity”.

  The inspection apparatus 1 can acquire image data of the printed matter 7 by imaging with the line camera 3. The inspection apparatus 1 images the printed matter 7 with the line camera 3 while conveying the printed matter 7 in the direction of movement 17.

  The line camera 3 is a device that images the printed matter 7. The line camera 3 is a camera (imaging device) including, for example, an optical sensor such as a CCD (Charge Coupled Device) in which light receiving elements (pixels) are arranged one-dimensionally, a drive circuit, an imaging optical system, and the like.

  The illuminator 5 is a light source that irradiates the printed material 7 with light. The illuminator 5 is, for example, an illuminator in which a plurality of LEDs are arranged in a planar shape. Further, a fluorescent lamp or the like can be used as the illuminator 5.

The image processing unit 9 performs image processing on the image data of the printed matter 7 and detects defects such as chipping and dirt. The image processing unit 9 can be configured by, for example, a computer and a program that causes the computer to operate as the image processing unit 9 of the inspection apparatus 1.
The configuration and operation of the image processing unit 9 will be described later.

The display unit 11 displays a defect detection result or the like in the image processing unit 9. The display unit 11 is, for example, a CRT (Cathode Ray Tube) or an LCD (Liquid Crystal Display).
The image storage unit 13 is a device that stores image data and the like. The image storage unit 13 is, for example, a hard disk or a storage medium drive.
The operation unit 15 is a device that receives an operation instruction from an operator. The operation unit 15 is a pointing device such as a keyboard and a mouse, for example.

(2. Hardware configuration of image processing unit 9)
Next, the hardware configuration of the image processing unit 9 will be described with reference to FIG.
FIG. 2 is a diagram illustrating a hardware configuration of the image processing unit 9.

The image processing unit 9 shown in FIG. 2 is configured by connecting a CPU 21, a memory 23, and the like to each other via a system bus (not shown).
The image processing unit 9 is connected to the line camera 3 via a network interface (not shown) or the like, and is connected to the display unit 11 via a video adapter (not shown) or the like.
Note that the hardware configuration in FIG. 2 is an example, and various configurations can be adopted depending on the application and purpose.

  A CPU 21 (Central Processing Unit) stores a storage device, a ROM (Read Only Memory), an execution program stored in a recording medium, an OS (operation system) program, an application program, and the like in a work memory on a RAM (Random Access Memory). Called and executed in the area, performs arithmetic processing (four arithmetic operations, comparison operations, etc.), operation control of hardware and software, and implements various functions (see FIG. 3 and the like) described later.

The memory 23 is a storage device such as a RAM or a ROM. In the RAM, a small area image memory 25, a similar small area image memory 27, a display memory 29, a printed material imaging memory (not shown) for storing imaging data of the printed material 7 by the line camera 3, and a work memory for executing a processing program ( (Not shown).
The small area image memory 25 holds a small area image 31, a subdivided small area image 33, and the like. The similar small area image memory 27 holds a similar small area image 35, a subdivided similar small area image 37, and the like.

  The small area image memory 25 and the similar small area image memory 27 are considered as a printed material imaging memory, and the inspection apparatus 1 first stores imaging data of the printed material 7 by the line camera 3 in a predetermined area of the printed material imaging memory. You may do it. The inspection apparatus 1 divides the image data and stores the reference small region image 31 in the small region image memory 25 according to the processing steps of the present embodiment, and a similar small region image having a high correlation with the small region image 31. 35 may be stored in the similar small area image memory 27.

The display memory 29 is a memory that holds display data to be displayed on the display unit 11, and holds, for example, image data indicating inspection results and defective portions.
The small area image memory 25, the similar small area image memory 27, the display memory 29, the printed matter imaging memory, and the like may be provided independently in the image processing unit 9, or areas may be appropriately allocated on the RAM. May be.

When the image data of the printed material 7 acquired by the line camera 3 is input to the image processing unit 9, the image processing unit 9 stores the image data in the memory 23 (printed material imaging memory or the like).
The CPU 21 divides the imaged data (printed product image data 39: see FIG. 4), extracts the reference small region image 31 and the similar small region image 35, and further subdivides each of the subdivided small region image 33 and the re-divided small region image 33. Based on the divided similar small area image 37, the quality of the entire print 7 to be inspected is determined, and the inspection result is output to the display memory 29.
The contents of the display memory 29 are input to the display unit 11 and displayed.

(3. Processing steps of the inspection apparatus 1 of the present embodiment)
Next, a processing procedure of the inspection apparatus 1 according to the present embodiment will be described with reference to FIGS.
FIG. 3 is a flowchart showing an inspection processing procedure of the inspection apparatus 1 according to the present embodiment.
This inspection process is executed by the CPU 21 of the image processing unit 9 by calling a predetermined program or the like stored in the ROM or the like of the memory 23 to the work memory area on the RAM. By executing this processing, the CPU 21 functions as an (inspection image) dividing unit, a (similar small region image) extracting unit, a correcting unit, a defect determining unit, and the like of the present invention.

(3-1. Region division processing and correction processing)
With reference to FIG. 3 (Step 1001 to Step 1008) and FIGS. 4 to 10, the region division processing and the correction processing executed by the inspection apparatus 1 will be described.
FIG. 4 is a diagram showing small area division of the printed material image data 39 and extraction of the temporary similar small area image 34 based on the small area image 31.
FIG. 5 is a diagram illustrating subdivision of the small area image 31.
FIG. 6 is a diagram illustrating re-division of the temporary similar small area image 34.
FIG. 7 is a diagram showing a search (correlation coefficient calculation) for the similar region 41 of the subdivision small region image 33 in the peripheral region 40 of the temporary subdivision similar subregion image 36.
FIG. 8 is a diagram illustrating correction of the re-divided similar small region images 37-1 to 37-6.
FIG. 9 is a diagram illustrating the generation of the similar small area image 35.
FIG. 10 is a diagram for explaining the generation process of the similar small region image 35.

  First, it is assumed that the operator images the printed matter 7 with the line camera 3, and the imaging data is sent to the image processing unit 9 of the inspection apparatus 1 and stored in the memory 23. In addition, the basic design is repeatedly arranged on the printed matter 7.

  The inspection apparatus 1 equally divides the imaging data (printed image data 39), which is an inspection target image, vertically and horizontally to generate a plurality of small area images (step 1001). FIG. 4 shows printed material image data 39 divided into small areas. In FIG. 4, the printed product image data 39 is divided into 9 parts vertically and 6 parts horizontally to generate a total of 54 small areas. Note that the inspection apparatus 1 displays the printed material image data 39 shown in FIG. 4 on the display unit 11, and the operator can confirm the imaging data of the printed material 7 and the small area division.

  The inspection apparatus 1 first designates the small area image 31 as the area image. Further, an area having the highest correlation coefficient with the small area image 31 is extracted as the temporary similar small area image 34 from the printed image data 39 divided into the small areas (step 1002). According to the present embodiment, since the area of the printed image data 39 is not divided according to the design, the correlation between the reference small area image 31 and the temporarily similar small area image 34 is sufficient. I can't say that. Therefore, if the defect inspection is performed by comparing the two area images at this stage, there is a possibility that the inspection accuracy is deteriorated or a defect detection error is caused.

  Next, the inspection apparatus 1 divides the small region image 31 and the temporary similar small region image 34 into predetermined numbers, respectively, and re-divides the small sub-region images 33-1 to 33-6 and the temporary re-divided similar small region image 36. -1 to 36-6 are generated (step 1003). In the present embodiment, each of the small region image 31 and the temporary similar small region image 34 is divided into six parts. FIG. 5 shows subdivision small region images 33-1 to 33-6, and FIG. Images 36-1 to 36-6 are shown.

  The inspection apparatus 1 extracts the region 41-1 having the highest correlation coefficient with the subdivision subregion image 33-1 in the peripheral region 40-1 of the temporary subdivision similar subregion image 36-1 (step 1004). The extraction of the region 41-1 is performed in units of subpixels (for example, 1/10 pixel), and high detection accuracy is realized.

  FIG. 7 shows extraction of an area 41-1 corresponding to the subdivision small area image 33-1 as an example. Note that the inspection apparatus 1 expands the region to the peripheral region 40 of the temporary subdivision similar small region image 36 and extracts the region 41 having the highest correlation coefficient for all the subdivision small region images 33.

  The enlargement amount of the peripheral region 40 is set as a default value in advance by a predetermined amount (for example, 30% enlargement with respect to the region of the temporarily subdivided similar small region image 36 in both the vertical and horizontal directions) or set in advance by the operator. It is assumed that

  Hereinafter, the inspection apparatus 1 extracts regions 41-1 to 41-6 having the highest correlation coefficient with the re-divided small region images 33-1 to 33-6 in the peripheral regions 40-1 to 40-6, respectively. The correlation coefficient between the subdivision small area images 33-1 to 33-3 and the areas 41-1 to 41-3 is higher than a predetermined value (for example, 90%) (the pattern has many features). The correlation coefficient between the region images 33-4 to 33-6 and the regions 41-4 to 41-6 will be described as being lower than a predetermined value (for example, 90%) (having little feature in the pattern).

  The inspection apparatus 1 extracts the regions 41-1 to 41-3 having high correlation coefficients as they are as the re-divided similar small region images 37-1 to 37-3 (Step 1005). The inspection apparatus 1 calculates the movement amounts from the temporary subdivision similar small region images 36-1 to 36-3 to the subdivision similar subregion images 37-1 to 37-3 as correction amounts 38-1 to 38-3. (Step 1006).

  The inspection apparatus 1 does not use the regions 41-4 to 41-6 where the correlation coefficient is low, and uses the average values of the correction amounts 38-1 to 38-3 when the correlation coefficient is high as the correction amounts 38-4 to 38-. 6 (step 1007). Based on the correction amounts 38-4 to 38-6 for the temporary subdivision similar small region images 36-4 to 36-6 calculated in the processing of step 1007, the inspection apparatus 1 resubdivides the similar subregion images 37-4 to 37-4. 37-6 is extracted (step 1008).

Here, the processing in steps 1004 to 1008 is summarized as follows (see FIG. 10).
The inspection apparatus 1 has the highest correlation coefficient with the re-divided small region images 33-1 to 33-6 in the peripheral regions 40-1 to 40-6 of the temporary re-divided similar small region images 36-1 to 36-6. Large areas 41-1 to 41-6 are extracted.

The inspection apparatus 1 directly adopts the regions 41-1 to 41-3 having a large correlation coefficient as the subdivision similar small region images 37-1 to 37-3, and calculates correction amounts 38-1 to 38-3 later. .
For calculating the correction amounts 38-1 to 38-3, the following calculation formula for the coordinate position can be used.
(Correction amount 38-1) = (subdivision similar small region image 37-1)-(temporary subdivision similar small region image 36-1),
(Correction amount 38-2) = (subdivision similar small region image 37-2)-(temporary subdivision similar small region image 36-2),
(Correction amount 38-3) = (subdivision similar small region image 37-3)-(temporary subdivision similar small region image 36-3).

The inspection apparatus 1 calculates the correction amounts 38-4 to 38-6 first without adopting the regions 41-4 to 41-6 having a small correlation coefficient.
For the calculation of the correction amounts 38-4 to 38-6, the following calculation formula relating to the coordinate position can be used.
(Correction amount 38-4) = (correction amount 38-5) = (correction amount 38-6)
= [(Correction amount 38-1) + (correction amount 38-2) + (correction amount 38-3)] / 3 (average).

The inspection apparatus 1 extracts subdivision similar small region images 37-4 to 37-6 based on the correction amounts 38-4 to 38-6.
In addition, regarding the extraction of the subdivision similar small region images 37-4 to 37-6, the following calculation formula regarding the coordinate position can be used.
(Subdivision similar small region image 37-4) = (temporary subdivision similar small region image 36-4) + (correction amount 38-4),
(Subdivision similar small region image 37-5) = (temporary subdivision similar small region image 36-5) + (correction amount 38-5),
(Subdivision similar small region image 37-6) = (temporary subdivision similar small region image 36-6) + (correction amount 38-6).

  As described above, the inspection apparatus 1 corrects the positions of the temporary subdivision similar small area images 36-1 to 36-6, and extracts the subdivision similar small area images 37-1 to 37-6. FIG. 9 shows a similar small region image 35 formed from the subdivision similar small region images 37-1 to 37-6. The inspection apparatus 1 registers the similar small region image 35 and the subdivided similar small region images 37-1 to 37-6 in the similar small region image memory 27.

  In the above-described method, the relative movement amount based on the coordinate position of the temporary subdivision similar small region images 36-1 to 36-6 is calculated and corrected. The relative movement amount based on the coordinate position of 33-6 may be calculated as the correction amount, and the re-divided similar small region images 37-1 to 37-6 may be extracted.

(3-2. Defect detection processing and determination processing)
Next, a defect detection process and a determination process executed by the inspection apparatus 1 will be described with reference to FIG. 3 (steps 1009 to 1012), FIG. 11, and FIG.
FIG. 11 is a diagram illustrating a flow of operation of the inspection apparatus 1 in the defect detection process and the determination process.
In FIG. 11, it is assumed that the re-divided similar region image 37 (inspection image) includes a dirt image 42 and a missing image 43 as defects.

  The inspection apparatus 1 performs a difference operation on all the pixels of the subdivision small region image 33 and the subdivision similar subregion image 37 to extract defect candidates (step 1009 in FIG. 3, step 2001 and step 3001 in FIG. 11). ).

  Dirt candidates are extracted based on a difference result obtained by subtracting the re-divided similar small area image 37 (inspection image) from the re-divided small area image 33 (reference image). For all corresponding pixels of the subdivision subregion image 33 and subdivision similar subregion image 37, the subdivision similar subregion image 37 is subjected to differential processing from the subdivision subregion image 33, and RGB processing is combined to perform MAX processing. Then, the dirt candidate image 45 is extracted (step 2001 in FIG. 11).

  On the contrary, a missing candidate is extracted based on a difference result obtained by subtracting the subdivision small region image 33 (reference image) from the subdivision similar subregion image 37 (inspection image). For all the corresponding pixels of the subdivision subregion image 33 and subdivision similar subregion image 37, the subdivision subregion image 33 is subjected to differential processing from the subdivision similar subregion image 37, and the RGB components are combined to perform MAX processing. The missing candidate image 47 is extracted (step 3001 in FIG. 11).

  The inspection apparatus 1 binarizes the dirt candidate image 45 and the missing candidate image 47 with a predetermined threshold and extracts defective pixels (step 1010 in FIG. 3, step 2002 and step 3002 in FIG. 11). The inspection apparatus 1 performs a binarization process on the stain candidate image 45 with a predetermined stain threshold value, and extracts a stain failure pixel (step 2002 in FIG. 11). The inspection apparatus 1 performs binarization processing on the missing candidate image 47 using a predetermined missing threshold value, and extracts missing defective pixels (step 3002 in FIG. 11).

  By performing binarization processing with a predetermined threshold, all pixel values are set to either the maximum density (density value “255”: white) or the minimum density (density value “0”: black). . Since the binarization process removes excess information, the difference feature (defect) can be emphasized.

Specifically, the above difference calculation is performed for each of the R component, the G component, and the B component for all corresponding pixels of the subdivision small region image 33 (reference image) and the subdivision similar subregion image 37 (inspection image). The difference value of each pixel is set to the maximum value among the R component, G component, and B component of the difference pixel value, and the binarized difference value is calculated for each pixel through the binarization process.
The difference calculation is performed in the saturation mode, and when the difference calculation result is negative (minus), the value is set to “0” (0 clip).

  For example, the pixel value of a pixel of the subdivision small region image 33 (reference image) is (R component, G component, B component) = (20, 100, 155), and the subdivision similar subregion image corresponding thereto Assume that the pixel value of 37 (inspection image) is (R component, G component, B component) = (60, 20, 100). The difference pixel value of (subdivision subregion image 33 (reference image)) − (subdivision similar subregion image 37 (examination image)) at the pixel is (−40, 80, 55). Further, the difference value in the pixel is “80”, which is the maximum value in each component of the difference pixel value. When the threshold value of the binarization process is “50” and binarized to “0” or “255”, the binarized difference value in the pixel exceeds the threshold value “50”, and thus becomes “255”.

  When the stain candidate image 45 and the missing candidate image 47 in FIG. 11 are binarized by threshold values in steps 2002 and 3002, respectively, an area having a binarized difference value of “255” (an area having a large difference: a defect candidate portion). ) Is indicated by “white (blank part)”, and an area where the binarized difference value is “0” (area where the difference is small) is indicated by “black (hatched part)”.

The inspection apparatus 1 detects a difference exceeding the predetermined number of defective pixels as a defect (step 1011 in FIG. 3, step 2003 and step 3003 in FIG. 11).
The inspection apparatus 1 detects an area where the number of dirty defective pixels extracted in the process of step 2002 exceeds a predetermined number of dirty pixels as a dirty defect. The inspection apparatus 1 creates a dirt detection image 49 and displays the detected dirt defect as the detected dirt 53 (step 2003).
The inspection apparatus 1 detects an area where the number of defective pixels extracted in the process of step 3002 exceeds a predetermined number of defective pixels as a defective defect. The inspection apparatus 1 creates the chipped detection image 51 and displays the detected chipping defect as the detected chipping 55 (step 3003).
Since an area exceeding the predetermined number of pixels is detected as a dirty defect or a chipped defect, by specifying the number of detected pixels, it is possible to eliminate erroneous detection of defects and adjust the defect detection accuracy.

  The inspection apparatus 1 extracts a defect by performing a difference between all of the subdivision subregion images 33 (reference images) constituting the subregion image 31 and the corresponding subdivision similar subregion images 37 (inspection images). (Step 1009 to Step 1011 in FIG. 3).

  The inspection apparatus 1 returns to step 1002 in FIG. 3 described above, and repeats the defect extraction processing in steps 1002 to 1011 for all the small area images constituting the inspection target image.

FIG. 12 is a diagram showing a defect determination display.
The inspection apparatus 1 displays the defect extraction result of the inspection target image on the display unit 11 and ends the process (step 1012 in FIG. 3).
The inspection apparatus 1 displays, for example, the detection stain 53 detected as a stain defect in “red” and the detection defect 55 detected as a defect defect in “blue” on the result display image 57.

(3-3. Overview of operation of inspection apparatus 1)
Through the above process, the inspection apparatus 1 divides an inspection image including a repeated pattern or the like into small region images, extracts similar region images for each small region image, and further extracts the small region image and the similar small region image. Subdivide. The inspection apparatus 1 extracts a subdivision similar small region image having the highest correlation coefficient in units of sub-pixels for each of the subdivision subregion images, and the subdivision similar small region whose correlation coefficient is higher than a predetermined value. The average value of the coordinate position correction amounts of the region image is calculated as the coordinate position correction amount of the subdivision similar small region image whose correlation coefficient is lower than a predetermined value. The inspection apparatus 1 performs a difference process between the subdivision small region image (reference image) and the subdivision similar subregion image (inspection image) after the coordinate position correction, and extracts a defect (dirt, chipped).

(4. Effect)
As described above, according to the embodiment of the present invention, it is possible to detect a printing defect with high accuracy even in a printed matter to be inspected, particularly in a region having a small feature in a pattern.

Further, according to the present invention, the correlation coefficient between the reference image of the region obtained by dividing the image to be inspected into small region images and further subdividing the small region image and the inspection image is calculated in subpixel units. It is possible to detect a defect with high accuracy and to reduce the erroneous detection of a defect.
In addition, since it is possible to detect defects in printed matter with high accuracy, yield, printing efficiency, printing cost, and the like can be improved.

Further, since it is not necessary for the operator to set an image as a reference area, there is an effect of reducing the operator's work load.
In addition, according to the present invention, it is possible to detect a stain defect and a chip defect in a printed material.

(5. Other)
In the above-described embodiment, the acquisition of the inspection target image of the printed material has been described as using a line camera. However, the present invention is not limited to this, and a scanner device or the like may be used.
In addition, the printed image data 39 may be acquired by imaging or scanning with a camera or the like. Not limited to this, digital data to be output to a printing plate or the like when printing by CTP (Computer To Plate) is used. The printed image data 39 may be used.

  In the above-described embodiment, the inspection apparatus 1 performs region division to obtain a small region image and a similar small region image, and performs division again to obtain a subdivision small region image and a subdivision similar small region image. Then, after correcting the coordinate position of the subdivision similar small region image having a small correlation coefficient based on the average coordinate position of the subdivision similar small region image having a large correlation coefficient, a defect is detected by performing a comparison process. As described above, the number of divisions is not limited to this, and the position correction processing of the present invention can be applied regardless of whether one-stage division processing is performed or multiple-stage division processing is performed. Can do.

The inspection apparatus 1 performs region division to obtain a small region image and a similar small region image. At this stage, the inspection device 1 has a small small correlation coefficient based on an average of coordinate positions of similar small region images having a large correlation coefficient. A defect may be detected by performing comparison processing after correcting the coordinate position of the region image. In this case, since the re-division process is not performed, the processing load can be reduced.
The inspection apparatus 1 performs region division to obtain a small region image and a similar small region image, performs division again to obtain a re-divided small region image and a re-divided similar small region image, and further performs re-division. Re-subdivision subregion images and sub-subdivision similar subregion images are acquired, and the coordinate position of the sub-subdivision similar subregion image with a small correlation coefficient is obtained based on the average coordinate position of the sub-subdivision similar subregion image with a large correlation coefficient After correction, a defect may be detected by performing comparison processing. In this case, defect detection with high accuracy can be realized by performing a plurality of division processes.

  The preferred embodiments of the inspection apparatus and the like according to the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to such examples. It will be apparent to those skilled in the art that various changes or modifications can be conceived within the scope of the technical idea disclosed in the present application, and these are naturally within the technical scope of the present invention. Understood.

Schematic configuration diagram of the inspection apparatus 1 The figure which shows the hardware constitutions of the image process part 9. Flow chart showing processing steps of inspection device 1 The figure which shows extraction of a small area division and temporary similar small area image 34 The figure which shows the subdivision of the small area image 31 The figure which shows the subdivision of the temporary similar small area | region image 34 The figure which shows the search (calculation of a correlation coefficient) of the similar area | region 41 of the subdivision subregion image 33 in the peripheral region 40 of the temporary subdivision similar subregion image 36. The figure which shows correction | amendment of the subdivision similar small area | region images 37-1 to 37-6. The figure which shows the production | generation of the similar small area | region image 35 The figure explaining the production | generation process of the similar small area | region image 35 The figure which shows the flow of operation | movement of the inspection apparatus 1 in the detection process and determination process of a defect. Diagram showing defect judgment display

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... …… Inspection apparatus 3 ......... Line camera 5 ......... Illuminator 7 ......... Printed material 9 ......... Image processing part 11 ......... Display part 13 ......... Image memory | storage part 15 ......... Operation part 17 ……… Moving direction 21 ……… CPU
23 ......... Memory 25 ......... Small area image memory 27 ......... Similar small area image memory 29 ......... Display memory 31 ......... Small area image 33, 33-1-33-6 ......... Subdivision small Area image 34... Temporary similar small area image 35... Similar small area image 36-1 to 36-6... Temporary subdivision similar small area image 37, 37-1 to 37-6. Similar small region images 38-1 to 38-6... Correction amount 39... Printed image data 40... Peripheral region 41... Region (region extracted in peripheral region 40)
42 ......... Dirty image 43 ... …… Deficient image 45 ... …… Dust candidate image 47 ... …… Deficient candidate image 49 ... …… Dust detected image 51 ... …… Deficient detected image 53 ... …… Detected stain 55 ... …… Missing detection 57 ... …… Result display image

Claims (16)

An inspection apparatus that performs image processing on an inspection image acquired by imaging an inspection object in which a plurality of identical arrangement units are arranged, and inspects the quality.
Dividing means for dividing the inspection image into a plurality of small region images;
In the inspection image, an extraction unit that extracts a similar small region image having the largest similarity for each small region image;
Correction means for correcting the position of the similar small area image whose similarity is smaller than the predetermined value based on the position of the similar small area image whose similarity is larger than a predetermined value;
An inspection apparatus comprising:   The inspection apparatus according to claim 1, further comprising a determination unit that compares the small area image with the similar small area image to determine whether the inspection object is good or bad.   The inspection apparatus according to claim 1, wherein the correction unit performs correction using an average value related to a position of the similar small region image in which the similarity is greater than the predetermined value. The dividing means further performs at least one subdivision,
4. The inspection apparatus according to claim 1, wherein the extraction unit, the correction unit, and the determination unit perform processing on the sub-region image and the similar sub-region image after re-division. .   5. The inspection apparatus according to claim 1, wherein the similarity is calculated with sub-pixel accuracy.   The determination means performs difference processing of the similar small area image from the small area image, binarizes with a predetermined threshold value, and determines an area exceeding a predetermined number of pixels as a contamination defect. The inspection device according to claim 5.   The determination means performs difference processing of the small area image from the similar small area image, binarizes with a predetermined threshold value, and determines an area exceeding a predetermined number of pixels as a defective defect. The inspection apparatus according to claim 6.   The inspection apparatus according to claim 1, wherein the processing according to claim 1 is performed on all the small area images of the inspection image. An inspection method for inspecting pass / fail by performing image processing on an inspection image acquired by imaging an inspection object in which a plurality of identical arrangement units are arranged,
A dividing step of dividing the inspection image into a plurality of small area images;
In the inspection image, an extraction step of extracting a similar small region image having the largest similarity for each small region image;
A correction step of correcting the position of the similar small area image whose similarity is smaller than the predetermined value based on the position of the similar small area image whose similarity is larger than a predetermined value;
An inspection method comprising:   The inspection method according to claim 9, further comprising a determination step of comparing the small region image and the similar small region image to determine whether the inspection object is good or bad.   11. The inspection method according to claim 9, wherein the correction step performs correction using an average value related to a position of the similar small region image in which the similarity is greater than the predetermined value. The dividing step further performs at least one subdivision,
The inspection method according to claim 9 or 11, wherein the extraction step, the correction step, and the determination step perform processing on the sub-region image and the similar sub-region image after re-division.   The inspection method according to claim 9, wherein the similarity is calculated with sub-pixel accuracy.   11. The determination step includes performing a difference process on the similar small region image from the small region image, binarizing with a predetermined threshold value, and determining a region exceeding a predetermined number of pixels as a stain defect. The inspection method according to claim 13.   11. The determination step includes performing a difference process on the small region image from the similar small region image, binarizing with a predetermined threshold value, and determining a region exceeding a predetermined number of pixels as a defective defect. The inspection method according to claim 14.   The inspection method according to claim 9, wherein the processing according to claim 9 is performed on all the small area images of the inspection image.

Images