JP2005043235A - Device and program for inspecting printed matter - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an inspection device which uses an inspection image itself made by photographing a printed matter, and decides the quality of the print state. <P>SOLUTION: The device for inspecting the printed matter photographs the printed matter and obtains the inspection image. A pixel region where black pixels aggregate is extracted as an object pixel region from the inspection image. Then, feature values of each object pixel region are computed, and a decision as to whether the object pixel region is an incorrect printed portion is made, on the basis of the feature values. Total pixel numbers, complexity of border lines, edge intensity of contours, distribution of pixel values and the like are used as the feature values. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a printed matter inspection apparatus and a printed matter inspection program for determining the quality of a printed matter.

  Conventionally, there has been a demand for an inspection apparatus that determines whether a desired image is printed on a print medium after a printing process. Such an inspection apparatus is proposed in Japanese Patent Laid-Open No. 11-78183.

  In the inspection apparatus disclosed in Japanese Patent Application Laid-Open No. 11-78183, a reference image, which is document information, is printed on preprinted paper on which a pre-image such as a form has been printed in advance, and a printed matter obtained thereby is imaged and an inspection image is obtained. Capture. Then, the print state of the printed matter is inspected by comparing an image obtained by masking the preprint portion of the inspection image with a reference image. Also, in this document, the structural features of the reference image are extracted based on the format information of the reference image, while the structural features are extracted in the same manner for the inspection image, and by comparing these structural features, Inspecting printed materials.

Further, there is a technique disclosed in Japanese Patent Laid-Open No. 9-259220, which is not directly related to the printed matter inspection apparatus. In this technique, a printed matter obtained by printing a reference image on preprinted paper is imaged, an inspection image is captured, and a density histogram is generated. Then, character information is extracted from the printed matter by removing points below a predetermined threshold from the density histogram.
JP-A-11-78183 JP-A-9-259220

  In the inspection apparatus disclosed in Japanese Patent Application Laid-Open No. 11-78183, a reference image and a prior image are required to inspect a printed matter. Therefore, there is a problem that in the absence of these image data, the printed matter cannot be inspected. Similarly, when examining the structural features of the inspection image, similarly, the printed matter cannot be inspected without the format information of the reference image.

  The present invention has been made in view of the above problems, and an object of the present invention is to determine the quality of a printing state from an inspection image itself obtained by imaging a printed matter.

  In order to achieve the above-described object, a printed matter inspection apparatus according to the present invention captures a printed matter on which a reference image, which is document information, is printed on a print medium on which a pre-image is printed, and obtains an inspection image. An image reading unit, a difference image generation unit that generates a difference image obtained by subtracting the reference image from the inspection image, and a pixel region including one or more pixels from the difference image as a target pixel region A target pixel area extracting unit for extracting; and a print state determining unit for calculating a feature amount of the target pixel region and determining whether the print state of the printed material is good based on the feature amount. According to this configuration, the present invention can detect a defect existing in a printed matter from the inspection image itself and determine whether the printing state is good or bad.

  Here, the feature amount of the target pixel area is an amount calculated from the data of the target pixel area itself, and represents the feature of the target pixel area. Typical features include, for example, the total number of pixels in the target pixel region, feature values related to contours such as contour complexity and edge strength, and distribution of pixel values.

  Further, it is preferable that the target pixel area extraction unit extracts a pixel area having a predetermined characteristic and is within a predetermined distance as a target pixel area. In general, since normal print portions are often located adjacent to each other, normal print portions that are within a predetermined distance from each other are connected and extracted as a target pixel region having a relatively large area. On the other hand, defective defective portions often exist in isolation, and are extracted as target pixel regions having a relatively small area. Therefore, this process makes it easy to distinguish between a normal printed part and a defective part.

  Further, it is preferable that the printing state determination unit does not determine a printing state defect for a target pixel region having registered registered pixel pattern data and having a similarity with the registered pixel pattern greater than a predetermined threshold. . This process can also simplify the process and prevent erroneous detection. In the above processing, the reference image and registered pixel pattern data are used in addition to the inspection image, but this does not hinder the object of the present invention. That is, the reference image or the like is used in an auxiliary manner to perform the processing of the present invention more suitably, and does not contradict the purpose of the present invention for determining pass / fail from the inspection image itself.

  Further, when the total number of pixels of the target pixel area is a number of pixels that cannot be visually identified, the print state determination unit may not determine a print state defect for the target pixel area. This simplifies the processing.

  The printed matter is a printed matter on which a reference image as document information is printed on a plurality of print media on which a common pre-image is printed. From the inspection image, a target pixel region included in the reference image And a pre-image generation means for generating the pre-image by removing the target pixel area determined to be defective, and whether the print state of the printed matter to be printed thereafter is good or bad using the pre-image generated. It is preferable to include simple determination means for determining.

  The present invention also provides an inspection image acquisition step for acquiring an inspection image obtained by imaging a printed matter on a computer, and a pixel area in which pixels having predetermined characteristics are gathered from the inspection image as a target pixel area. A printed matter inspection program for executing a target pixel region extracting step and a print state determining step of calculating a feature amount of the target pixel region and determining the quality of the print state of the printed matter based on the feature amount .

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram illustrating an example of a printing system including a printed matter inspection apparatus according to the present embodiment.

  In this printing system, an image is printed on the paper fed from the paper feed cassette 10 by the print processing unit 20, and the printed reference image is fixed on the paper by the fixing roller 25. In the case of an electrophotographic system, the print processing unit 20 includes a raster output scanner, a photosensitive drum, a developing device, a transfer roll, and the like. When printing a standard document, preprinted paper on which a standard image is printed is set in any of the paper feed cassettes 10, and the cassette 10 is selected as a paper supply source to start printing.

  A reading scanner 30 is provided downstream of the fixing roller 25 in the paper conveyance path. The printed surface of the printed material on which the image is fixed is optically read by the reading scanner 30. The reading scanner 30 irradiates the printing surface with an irradiation lamp 35, focuses the reflected light with a lens 36, and detects it with an imaging device 37. As the imaging device 37, for example, a line sensor in which cells are arranged in a line in a direction perpendicular to the paper conveyance direction in a plane parallel to the upper surface of the paper conveyance path can be used. By repeatedly reading the line sensor in synchronization with the sheet conveyance, a two-dimensional printed surface image can be acquired. When inspecting a color image, for example, a line sensor may be provided for each of R, G, and B colors. The image on the printing surface acquired by the imaging device 37, that is, the inspection image is sent to the image processing apparatus 50. Note that the reading resolution of the line sensor is determined in consideration of the accuracy and processing speed necessary for the inspection, but about 100 to 600 dpi is appropriate.

  The printed matter inspection apparatus according to the present embodiment includes a reading scanner 30 and an image processing apparatus 50. The image processing apparatus 50 is configured to include a CPU, a memory, and the like, and performs a printing state determination process on the inspection image captured by the reading scanner 30 when the CPU executes a printed matter inspection program. FIG. 2 is a flowchart showing the print state determination process.

  The image processing apparatus 50 takes in the reference image data used for the printing process. Since image data with a resolution of about 300 to 1200 dpi is usually used for printing and is often higher in resolution than the inspection image, it is often necessary to convert the resolution according to the inspection image when capturing the reference image. . In this embodiment, a business document is targeted and the reference image data is binary image data. However, if the image data is multi-value image data such as a color image or a grayscale image, it is binarized (S201). . Hereinafter, the pixel value of the binary image is 1 for black pixels and 0 for white pixels. As shown in FIG. 3, the reference image is an image including characters, numbers, and the like.

  Next, the image processing apparatus 50 takes the inspection image data from the reading scanner 30 as grayscale multi-valued image data and binarizes it (S202). For example, when the inspection image is taken in as multi-valued image data with a luminance value of 0 to 255, it may be binarized with an appropriate threshold (for example, 160 in this case). The inspection image is an image obtained by printing the reference image of FIG. 3 on preprinted paper on which ruled lines P1, logos P2, characters (addresses) P3, etc. are printed as shown in FIG. Further, due to an abnormality in the printing system or an abnormality in the preprinting apparatus, erroneously printed portions D1 to D4 are included in the detected image. Further, an image missing portion D5 due to missing printing of a part of the reference image occurs. In the present embodiment, the printed matter is A4 size, and the resolution of the reference image and the inspection image is 200 dpi.

  Next, a process of extracting normal printing portions and erroneous printing portions D1 to D4 as target pixel regions is performed (S203). In this processing, the image processing apparatus 50 checks whether each pixel of the inspection image has a predetermined characteristic indicating that it is a normal print portion or an erroneous print portion D1 to D4, and a pixel having the predetermined characteristic is determined. Pixel regions connected to each other are extracted as target pixel regions. Here, the characteristic of the pixel is a characteristic given to each pixel, for example, a pixel value such as a luminance or a color of the pixel. In the present embodiment, a target pixel region in which black pixels are connected and gathered is extracted, and thereby all normal printed portions and erroneous printed portions are extracted.

  The black pixel region thus connected may be the target pixel region, but in the present embodiment, the following processing is further performed in order to appropriately determine the erroneous printing portions D1 to D4. First, the black pixel region of the inspection image is expanded by a predetermined pixel, for example, 5 pixels. When the black pixel region is expanded, as shown in FIG. 5, the black pixel region expands to a thin line L1 that encloses each character and the like, and the black pixel regions such as the character P3 at adjacent positions are connected to each other. Then, each connected black pixel region is set as a target pixel region. By this processing, the connected black pixel area becomes a relatively large target pixel area. In the above processing, the number of pixels to be expanded may be appropriately determined according to the printed matter.

  Next, processing for determining the presence or absence of an image missing portion of the reference image is performed on the inspection image (S204). In this process, first, the black pixel region of the inspection image is expanded by about one pixel or about two pixels. Then, the corresponding pixel value of the inspection image is subtracted from each pixel value of the reference image. Here, “subtraction” is to obtain a difference by subtracting the pixel value of the corresponding pixel of the inspection image from the pixel value of each pixel of the reference image. If the difference is smaller than 0, the pixel value is set to 0. In the subsequent processing, “subtraction” is the same processing. The pixel value of each pixel after subtraction is checked, and a pixel having a pixel value greater than 0 is detected as an image missing portion, and the data is stored. When there is an image missing portion, as shown in FIG. 6, an image missing portion D5 'appears in the subtracted difference image and is detected. In addition, since the black pixel area of the inspection image is expanded by about 1 pixel or 2 pixels, the positional deviation and the binarization error between the inspection image and the reference image are compensated, and the pixel values of almost all the pixels are 0. . In this image missing portion determination process, in order to further check whether or not the image missing portion D5 ′ is a serious image missing, the black pixel area of the image missing portion D5 ′ is reduced by about one pixel or two pixels, Even when the image missing portion D5 ′ still remains, the data of the image missing portion D5 ′ may be stored (S205).

  Next, a difference image is obtained by subtracting the pixel value of the corresponding pixel of the reference image from the pixel value of each pixel of the inspection image (S206). By this process, the target pixel area that is included in common with the reference image is removed from the target pixel area extracted from the inspection image, so that the process of determining the quality of the subsequent printing state is simplified. In this process as well, it is preferable that the black pixel region of the reference image is expanded by about one pixel or two pixels before subtraction to compensate for positional deviation and binarization error. FIG. 7 shows the difference image.

  Next, a process of removing the registered registered pixel pattern registered from the difference image is performed (S207). In order to perform this process, pixel pattern data to be printed on the printed material is registered in the memory of the image processing apparatus 50. Examples of pixel pattern data include a company logo and an insignia. Then, pattern matching is performed on all target pixel areas included in the difference image to calculate the similarity with the registered pixel pattern. If the similarity is equal to or greater than a predetermined threshold, the target pixel area is registered as the registered pixel pattern. Detect as. When the registered pixel pattern is detected, the target pixel region is removed from the difference image. This process also reduces the target pixel area, and simplifies subsequent quality determination processing. In this embodiment, the company logo P2 is detected and removed.

  Next, a process of subtracting and removing a target pixel area that cannot be visually identified from the difference image is performed (S208). Here, the level that cannot be identified visually is a size that is not noticeable by a person who sees the printed matter without being noticed, and is, for example, 0.1 mm or less. In this process, the total number of pixels is counted for each target pixel area, and if the total number of pixels is equal to or less than a predetermined value that cannot be visually identified, the target pixel area is removed from the difference image. This process also reduces the target pixel area, and simplifies subsequent quality determination processing. The threshold value for the total number of pixels may be determined according to the resolution of the image. Further, the size of the target pixel region may be obtained by another method such as counting the number of pixels in the vertical or horizontal direction.

  Next, a feature amount of each target pixel region is calculated, and processing for determining whether the printed state of the printed matter is good or not based on the feature amount is performed (S209 to S212). Here, the feature amount of the target pixel area is an amount calculated from the data of the target pixel area itself, and represents the feature of the target pixel area. As will be described below, in the present embodiment, the total number of pixels in the target pixel region, the complexity of the contour shape, and the edge strength of the contour are calculated and set for each feature amount as the feature amount of the target pixel region. Compared with the determination criterion, it is determined whether or not each target pixel area is erroneously printed.

  When the total number of pixels in the target pixel area is used as the feature amount, the total number of pixels is counted for each target pixel area, and it is determined whether the total number of pixels is equal to or less than a predetermined threshold. If the total number of pixels is equal to or less than the threshold, the target pixel area is determined as an erroneous printing portion, and the information is stored in the memory (S209). Further, the target pixel region is removed from the difference image. In the present embodiment, a relatively small erroneous printing portion D1 is detected by this process. Note that, by the processing in step 203, adjacent small black pixel regions such as characters are connected to each other to be a large target pixel region, and therefore these black pixel regions are not detected as erroneous printing portions.

  When the complexity of the contour shape of the target pixel region is used as the feature amount, first, as shown in FIG. 8, a contour line L2 of each target pixel region is generated (S210). This generation process will be described with reference to FIG. If the target pixel region R to which the black pixels are connected has the shape shown in FIG. 9A, the center positions of the black pixels that are located at the periphery of the target pixel region R and are in contact with the white pixels are connected. Thus, the contour line L2 shown in FIG. 9B is obtained. Next, for this contour line L2, the relationship between the number of moving pixels starting from the predetermined pixel A and the direction change is obtained. If this relationship is represented by the number of moving pixels on the horizontal axis and the direction change on the vertical axis, the graph shown in FIG. 10 is obtained. Next, a numerical value serving as an index of the complexity of the contour line L2 is calculated based on the relationship between the number of moving pixels and the direction change. Here, in order to calculate the complexity of the contour line L2, for example, Fourier transformation is performed on the relationship between the number of moving pixels and the direction change, or the frequency of the distribution of pixel intervals in which the direction change occurs, What is necessary is just to obtain | require the magnitude | size of the high frequency component of a direction change. When the complexity of the contour line L2 is equal to or greater than the predetermined threshold, the target pixel area is determined as an erroneous printing portion, and the information is stored in the memory (S211). Further, the target pixel region is removed from the difference image. By this processing, erroneous printing portions D3 and D4 shown in FIG. 3 are detected.

  When the edge strength in the contour of the target pixel region is used as a feature amount, the edge strength is calculated by applying a spatial frequency filter to the pixel values before binarization processing in each target pixel region and its surroundings. . For example, the product-sum calculation is performed by applying the Laplacian operator shown in FIG. 11B to the pixel value of the inspection image before the binarization processing shown in FIG. Is calculated. When the contour of the target pixel region is clear and the pixel value changes abruptly, the large pixel value extends with a narrow width along the contour of the target pixel region in the edge strength calculation result. On the other hand, when the contour of the target pixel region is unclear and the pixel value changes slowly, the pixel value of the intermediate region extends widely along the contour of the target pixel region. Therefore, when the pixel value of the intermediate region is greater than or equal to a predetermined width in the vicinity of the contour of the target pixel region, the target pixel region is determined as an erroneously printed portion with an unclear contour, and the information is stored in the memory (S212). . Further, the target pixel region is removed from the difference image. By this processing, erroneous printing portions D2, D3, and D4 shown in FIG. 3 are detected. Note that the edge strength may be calculated by other methods. For example, the edge strength may be calculated using a Sobel operator shown in FIG. 12 as a spatial frequency filter.

  In the present embodiment, as described above, the feature amount of the target pixel region is calculated by various methods, and it is determined based on the feature amount whether or not the inspection target region is a misprinted portion. The quality is judged. Accordingly, it is possible to determine whether the print state of the printed material is good or not from the detected image itself without requiring special reference data, pre-image data, or the like.

  In addition, when pre-printing or printing by a printing system is in color, a pixel value distribution may be adopted as the feature amount of the target pixel region. In other words, the target pixel area that is erroneously printed may have a color distribution that is not printed in a normal printed part, and therefore, by examining the distribution of pixel values of the captured inspection image, the target pixel area is erroneously printed. It is determined whether or not there is. In addition, since the colors that can be printed differ depending on the type of printing device, by examining the distribution of pixel values of the inspection image, it is determined whether the target pixel region is due to preprinting or according to the printing system of the present embodiment, This result can also be used for quality determination of printed matter. Specifically, the distribution of pixel values in the target pixel region may be obtained, and determination may be performed based on representative values such as the maximum value and the average value.

  In addition, when a large number of preprinted sheets on which a common pre-image is printed are set in the paper feed cassette 10 and different reference images are continuously printed for each preprinted sheet, the processing described below is performed. Preferably it is done. First, the above-described feature amount calculation is performed on the captured inspection image to detect a target pixel region that is a misprinted portion, and the target pixel region included in the reference image is detected from the inspection image as a faulty error. A preliminary image is generated by removing the target pixel area which is a print portion. And about the test | inspection image imaged after that, the quality of a printing state is determined using this prior image. For example, for the subsequent inspection image, the reference image and the preliminary image are subtracted from the inspection image, and the remaining portion is determined as an erroneously printed portion, or the similarity between the preprinted portion of the inspection image and the preliminary image is obtained. Check for abnormal print paper. By performing this simple determination process, it is possible to reduce the time required for calculating the feature amount and to inspect the print state of the printed matter more quickly.

  The preferred embodiments of the present invention have been described above. However, the present invention is not limited to the above embodiments, and various modifications can be made within an equivalent range.

1 is a configuration diagram of a printer system. FIG. It is a flowchart of a process of an image inspection apparatus. It is explanatory drawing which shows a reference | standard image. It is explanatory drawing which shows a test | inspection image. It is explanatory drawing which shows the test | inspection image by which the black pixel area | region was expanded. It is explanatory drawing which shows the image obtained by subtracting a test | inspection image from a reference | standard image. It is explanatory drawing which shows the image obtained by subtracting a reference | standard image from a test | inspection image. It is explanatory drawing which shows the image from which the outline of the object pixel area was extracted. It is explanatory drawing for demonstrating the arithmetic processing regarding the complexity of an outline. It is explanatory drawing for demonstrating the arithmetic processing regarding the complexity of an outline. It is explanatory drawing for demonstrating the calculation processing regarding the edge strength of an outline. It is explanatory drawing which shows a Sobel operator.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Paper cassette, 20 Print processing part, 25 Fixing roller, 30 Reading scanner, 35 Irradiation lamp, 36 Lens, 37 Imaging device, 50 Image processing apparatus.

Claims (11)

An inspection image reading unit that captures a printed material on which a reference image, which is document information, is printed on a print medium on which a preliminary image is printed;
Difference image generation means for generating a difference image obtained by subtracting the reference image from the inspection image;
A target pixel region extracting means for extracting, from the difference image, a pixel region composed of one or more pixels as a target pixel region;
A printing state determination unit that calculates a feature amount of the target pixel region and determines whether the print state of the printed material is good or not based on the feature amount;
A printed matter inspection apparatus comprising: The printed matter inspection apparatus according to claim 1,
The feature amount of the target pixel region includes the total number of pixels of the target pixel region,
The printed matter inspecting device, wherein the printing state determining unit determines a defective printing state when the total number of pixels is equal to or less than a predetermined reference value. The printed matter inspection apparatus according to claim 1,
The printed matter inspection apparatus, wherein the feature amount of the target pixel region includes a feature amount related to an outline of the target pixel region. The printed matter inspection apparatus according to claim 3,
The printed matter inspection apparatus, wherein the feature amount related to the contour of the target pixel region includes an edge strength in the contour of the target pixel region. The printed matter inspection apparatus according to claim 3,
The feature inspection related to the contour of the target pixel region includes the complexity of the contour shape of the target pixel region. The printed matter inspection apparatus according to claim 1,
The printed matter inspection apparatus, wherein the feature amount of the target pixel region includes a distribution of pixel values of the target pixel region. The printed matter inspection apparatus according to claim 1,
The object pixel area extracting means extracts a pixel area that is a collection of predetermined characteristics and that is within a predetermined distance from each other as a target pixel area. The printed matter inspection apparatus according to claim 1,
The printing state determination means includes
Having registered registered pixel pattern data,
A printed matter inspection apparatus that does not determine a print state defect for the target pixel region having a similarity to the registered pixel pattern greater than a predetermined threshold. The printed matter inspection apparatus according to claim 1,
The printed matter determination device does not determine a defective print state for a target pixel region that cannot be visually identified. The printed matter inspection apparatus according to claim 1,
The printed matter is a printed matter on which a reference image as document information is printed on a plurality of print media on which a common pre-image is printed,
A preliminary image generation means for generating the preliminary image by removing the target pixel area determined to be defective from the difference image;
Simple determination means for determining the quality of the print state of the printed matter to be printed thereafter using the generated preliminary image;
A printed matter inspection apparatus comprising: On the computer,
An inspection image acquisition step of acquiring an inspection image obtained by imaging a printed matter on which a reference image as document information is printed on a print medium on which a preliminary image is printed;
A difference image generation step of generating a difference image obtained by subtracting the reference image from the inspection image;
A target pixel region extraction step of extracting, from the difference image, a pixel region composed of one or more pixels as a target pixel region;
A printing state determination step of calculating a feature amount of the target pixel region and determining whether the print state of the printed material is good or not based on the feature amount;
Printed matter inspection program to execute.

Images