Image analysis device, image analysis method, and program

Abstract

To provide an image analysis device, an image analysis method and a program for accurately evaluating a distribution of an ink that has penetrated printed matter.SOLUTION: An image analysis device 1 is provided, comprising: a correction coefficient computation unit 11 (step S13) configured to compute ink penetration correction coefficients kx, ky on the basis of a surface image of a printed surface of printed matter before embedding and a surface image of the printed surface of the printed matter after embedding; and a correction unit 12 (step S14) for correcting a cross-sectional image of the printed matter obtained by cutting and photographing the printed matter after embedding using the correction coefficients kx, ky.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to an image analysis apparatus, an image analysis method, and a program.

Background Art

[0002] A printed image printed by inkjet printing is an aggregate of ink dots, and the ink dot shape is cited as an index of the image quality of a printed matter such as paper. As a method for evaluating this ink dot shape, there is a method of cutting the printed matter to observe the internal ink distribution. As a method for observing the internal ink distribution of a printed matter in this way, after embedding the printed matter in resin, the printed matter is cut using a microtome or a razor, and the cross section of the printed matter is photographed, so that the internal ink distribution of the printed matter is generally observed.

[0003] Also, as a method for observing the internal ink distribution of a printed matter, in order to omit resin embedding, the printed matter is cut using a focused ion beam apparatus, and the cross section of the cut printed matter is imaged with both a scanning electron microscope and an optical microscope, and a method of superimposing the imaging image by the scanning electron microscope and the density image of the ink in the imaging image by the optical microscope has been proposed (for example, see Patent Document 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] When the printed matter is embedded in resin as described above, ink elution into the resin and deformation during resin curing occur, so that even if the printed matter is cut to take a cross-sectional image, an accurate ink distribution cannot be obtained.

[0006] In addition, in order to take a cross-sectional image by omitting the embedding resin as described above and cutting the printed matter using a focused ion beam apparatus or the like, a special apparatus such as a focused ion beam apparatus is required. Further, the procedure is complicated and it is difficult to obtain a three-dimensional ink distribution.

[0007] An object of the present invention is to provide an image analysis apparatus, an image analysis method, and a program that can accurately evaluate the distribution of ink penetrated into a printed matter with a simple configuration.

Means for Solving the Problems

[0008] In one aspect, the image analysis apparatus includes a correction coefficient calculation unit that calculates a correction coefficient for ink penetration based on a surface image of the printing surface of the printed matter before embedding and a surface image of the printing surface of the printed matter after embedding, and a correction unit that corrects a cross-sectional image of the printed matter taken by cutting the printed matter after embedding based on the correction coefficient.

Effects of the Invention

[0009] According to the above aspect, the distribution of ink penetrated into the printed matter can be accurately evaluated with a simple configuration.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Mode for Carrying Out the Invention

[0011] Hereinafter, an image analysis apparatus, an image analysis method, and a program according to an embodiment of the present invention will be described with reference to the drawings.

[0012] FIG. 1 is a block diagram showing an image analysis apparatus 1 according to an embodiment.

[0013] As shown in FIG. 1, the image analysis apparatus 1 includes a control unit 10, a storage unit 20, a display unit 30, an input unit 40, and an interface unit 50.

[0014] The control unit 10 has one or more processors (for example, a CPU: Central Processing Unit) that function as an arithmetic processing device. This processor reads and executes a predetermined program from, for example, the storage unit 20 or a storage medium (non-transitory computer-readable recording medium) detachable from the image analysis apparatus 1, and functions as a correction coefficient calculation unit 11, a correction unit 12, and an evaluation value calculation unit 13. In this way, the control unit 10 (or the image analysis apparatus 1) functions as an example of a computer that executes a program.

[0015] First, in this embodiment, before embedding a printed matter such as paper in resin, a surface image of the printed surface (see Fig. 2(a)) is captured by a microscope apparatus. Then, the printed matter is embedded in resin as a pre-treatment for cutting, and is cut using a microtome or a razor so that a cross-section parallel to the printed surface is exposed. For example, each time it is cut, a cross-sectional image (see Fig. 2(b)) is captured by transmission observation. Note that the surface image and the cross-sectional image are, for example, images of ink for one dot. Also, for capturing the printed surface and the cross-section, any observation method can be employed using an arbitrary microscope apparatus such as an optical microscope apparatus. Also, the embedding is performed using, for example, an arbitrary resin, but it may be performed by a method other than resin embedding.

[0016] Based on the surface image of the printed surface of the printed matter before resin embedding (see Fig. 2(a)) and the surface image of the printed surface of the printed matter after resin embedding (not shown), the correction coefficient calculation unit 11 calculates a correction coefficient for ink penetration. Note that the surface image of the printed surface after resin embedding is, for example, a cross-sectional image of the printed surface where the resin has been cut and exposed. However, if the printed surface after resin embedding can be captured through the resin, the surface image after resin embedding may be an image captured before cutting. Note that deformation during cutting of the printed matter (for example, shrinkage in the cutting direction, etc.) may also occur in the cross-sectional image of the printed matter. Therefore, if deformation during cutting also occurs in the printed matter even when only the resin is cut, in order to calculate a correction value assuming the deformation during cutting, it is desirable that the surface image of the printed surface after resin embedding is a cross-sectional image of the printed surface where the resin has been cut and exposed.

[0017] For example, the correction coefficient calculation unit 11 may calculate, as the correction coefficient, the reduction rate (or magnification rate) of the ink penetration region of the surface image after resin embedding with respect to the ink penetration region where ink has penetrated in the surface image before resin embedding. This correction coefficient may be calculated for each of the x coordinate and the y coordinate orthogonal to each other on the printed surface, but may also be calculated as a common value for these. The correction coefficient can be calculated, for example, based on the change in the coordinates (contour coordinates) of the ink penetration region before and after resin embedding, or the change in the area of the ink penetration region before and after resin embedding.

[0018] The correction unit 12 corrects a cross-sectional image of a printed matter photographed by cutting the printed matter after resin embedding, for example, by using an affine transformation (enlargement) based on a correction coefficient. Correcting the cross-sectional image in this way is because when the printed matter undergoes resin embedding, the ink penetration region changes due to ink bleeding into the resin and deformation during resin curing. This is to reproduce the ink penetration region before the change (before resin embedding). The correction of the cross-sectional image is, for example, expressed by the following formula [Equation 1], using the correction coefficient k x for the x coordinate and the correction coefficient k y for the y coordinate to convert the coordinates (x, y) of the cross-sectional image to coordinates (x’, y’).

[0019]

Equation

[0020] The evaluation value calculation unit 13 calculates an evaluation value of the ink penetration of the printed matter based on a plurality of cross-sectional images corrected by the correction unit 12. The calculation of this evaluation value will be described later with reference to the flowchart of FIG. 4.

[0021] Note that in addition to the correction coefficient calculation unit 11, the correction unit 12, and the evaluation value calculation unit 13, the control unit 10 may have, for example, an acquisition unit that acquires a surface image before resin embedding and a cross-sectional image after resin embedding from external devices such as a microscope device and a user terminal, and an image processing unit that performs image processing on the acquired image data.

[0022] The storage unit 20 includes, for example, a ROM (Read Only Memory), which is a read-only semiconductor memory in which a predetermined control program is recorded in advance, a RAM (Random Access Memory), which is a semiconductor memory that can be written to and read from as needed and is used as a working storage area when the processor executes various control programs, and a hard disk device. For example, the storage unit 20 stores a plurality of cross-sectional images corrected based on the correction coefficients k x , k y as input images, and stores evaluation values calculated as described later.

[0023] The display unit 30 is a display that displays a display screen or the like for an input operation by the user.

[0024] The input unit 40 is, for example, a keyboard, a mouse, a touch panel, or the like for performing various operations.

[0025] The interface unit 50 exchanges various information with external devices such as a microscope device and a user terminal. For example, the interface unit 50 acquires a surface image before resin embedding and a cross-sectional image after resin embedding from the microscope device or the user terminal.

[0026] Here, with reference to FIG. 3, the correction process of the cross-sectional image will be described. Note that descriptions of matters overlapping with the above description will be omitted as appropriate.

[0027] First, the control unit 10 acquires a surface image of the printed surface before resin embedding shown in FIG. 2(a) photographed by the microscope device as described above (step S11).

[0028] Further, the control unit 10 acquires a plurality of cross-sectional images photographed by the microscope device as described above, which are images of a cross-section cut so that a cross-section parallel to the printed surface is exposed after resin embedding in the printed matter (step S12). These plurality of cross-sectional images may include a surface image of the printed surface after resin embedding as described above.

[0029] The control unit 10 may acquire the surface image of the printed surface before resin embedding (step S11) and the cross-sectional image after resin embedding (step S12) simultaneously. Note that since the oil-based ink penetrates and spreads over time after printing, the time difference between the imaging timing of the surface image before resin embedding and the imaging timing of the surface image (cross-sectional image) after resin embedding is preferably short. For example, in the case of printing with oil-based ink, these images are preferably taken after leaving them at a constant temperature for 24 hours after printing.

[0030] Further, the correction coefficient calculation unit 11 compares the surface images before and after resin embedding, and calculates the correction coefficients k x , k y as described above (step S13).

[0031] Further, the correction unit 12 corrects the plurality of cross-sectional images obtained as described above (step S12), that is, the plurality of cross-sectional images of the printed matter taken by cutting the printed matter after resin embedding, based on the correction coefficients k x , k y (step S14). This correction may be performed, for example, by an affine transformation using the above formula (1).

[0032] Further, the control unit 10 stores the plurality of converted cross-sectional images in the storage unit 20 as input images (step S15). Thus, the process shown in FIG. 3 ends.

[0033] Next, with reference to FIG. 4, the calculation process of the evaluation value of the ink penetration of the printed matter will be described. Regarding FIG. 4 as well, descriptions of matters overlapping with the above description will be omitted as appropriate.

[0034] First, the evaluation value calculation unit 13 (control unit 10) reads the input image stored in the storage unit 20 in step S15 of FIG. 3 described above (step S21).

[0035] Further, the evaluation value calculation unit 13 determines whether the read cross-sectional image is a color image (or a gray image) (step S22). If it is a color image (step S22: YES), the color image is converted into a gray image (steps S23 to S25). This conversion to a gray image may be performed by any known method. For example, it is performed by generating an RGB channel image (step S23), performing HSV conversion and generating each converted image (step S24), and specifying the ink color from the HSV image and converting it into a gray image (step S25). As a result, a gray image of sample 1 shown in FIG. 5(a) and a gray image of sample 2 shown in FIG. 5(b) are obtained. Note that the images of samples 1 and 2 shown in FIGS. 5 to 7 are used for convenience of explanation as being all generated from the surface images (color images) before resin embedding, but actually they are generated from the corrected cross-sectional images.

[0036] When the read cross-sectional image is a gray image (step S22: NO), and when the color image is converted into a gray image as described above (steps S23 to S25), the evaluation value calculation unit 13 binarizes the image into black and white using a specified threshold value in order to extract the ink part (ink penetration region) (step S26). As a result, a binarized image of sample 1 shown in FIG. 6(a) and a binarized image of sample 2 shown in FIG. 6(b) are obtained.

[0037] Further, the evaluation value calculation unit 13 extracts the contour of the ink penetration region from the binarized image (step S27). For this contour extraction, for example, a known method such as contour tracing can be used.

[0038] Further, the evaluation value calculation unit 13 determines whether the cross-sectional image having a contour (ink penetration region) is the last one (step S28). If it is not the last one (step S28: NO), the processes of steps S21 to S27 described above are performed on other input images (cross-sectional images corrected based on the correction coefficients k x , k y ).

[0039] If the cross-sectional image with a contour is the last one (step S28: YES), the evaluation value calculation unit 13 calculates the fractal dimension for each contour (step S29). This fractal dimension can be said to be an index used as a value indicating how complex the contour shape of the ink dots is. For example, the fractal dimension is a numerical value representing self-similarity, that is, the characteristics of the details are similar to those of the whole. In image analysis, it is a parameter indicating the shade change of each pixel and the complexity / roughness of the shape. The fractal dimension f can be obtained when a certain figure is reduced to 1 / a and the number of figures required to fill the original figure is b, where a f =b. The more complex the line figure becomes to cover the plane, the closer it gets to the plane, so the fractal dimension f approaches 2. In the example of Fig. 8, regarding the upper and lower contours (an example different from Fig. 7), since the lower contour is more complex than the upper contour, similar to the image in the left column (an image showing the reduced shape approximated as a circle for the virtual ink dot shape (the area surrounded by the line)), the number of circles (b) required for the lower part is larger. Therefore, the fractal dimension f of the lower contour is larger than that of the upper contour.

[0040] Note that the contour image of sample 1 shown in Fig. 7(a) represents a contour with a fractal dimension f of 1.07, and the contour image of sample 2 shown in Fig. 7(b) with a more complex contour represents a contour with a fractal dimension f of 1.21.

[0041] Next, the evaluation value calculation unit 13 calculates the evaluation value of the ink penetration of the printed matter based on the fractal dimension f of the plurality of cross-sectional images (step S30). Thereby, the process shown in Fig. 4 is completed.

[0042] For example, when there are n cross-sections where ink penetration (transfer) can be confirmed and the fractal dimension of the n-th layer is f(n), the evaluation value of the dot penetration behavior can be calculated by the following formula [Equation 2]. As shown in the formula of [Equation 2] below, it is advisable to use the sum of each layer of the product of the fractal dimension (f(n)), which is an example of the bleeding quantification value at a predetermined penetration depth, and the number of layers (n) as the degree of bleeding of the printed matter. Note that in combinations where the ink is likely to bleed, such as the combination of ink and the printed matter (paper), the evaluation value will be larger. In this way, the degree of ink bleeding can be quantitatively evaluated.

[0043]

Number

[0044] If comparing samples with different thicknesses per layer of the cross-sectional image taken by cutting, when the thickness of each layer is t and the ink penetration depth is z, the evaluation value may be calculated by the following formula [Equation 3].

[0045]

Number

[0046] Although it is an example, the evaluation value of ink penetration calculated as described above may be presented to the user by the display unit 30, or transmitted to an external device via the interface unit 50 and presented to the user. Alternatively, a printing device, a printing control device that controls this printing device, an image processing device that performs image processing on the printed image, etc. may perform image processing to correct the printed image based on the above evaluation value. For example, if the evaluation value indicates that there is a lot of ink bleeding, image processing such as reducing the number of dots in the printed image (the entire printed image or only the edges of the printed area, etc.) may be performed.

[0047] Also, in order to calculate the evaluation value of ink penetration, it is not necessary to use the fractal dimension f. For example, the area of the ink penetration region may be calculated using the contour obtained as described above, and the evaluation value may be calculated using this area.

[0048] In the embodiment described above, the image analysis apparatus 1 calculates the correction coefficient k for ink penetration based on the surface image of the printed surface of the printed matter before embedding (see FIG. 2(a)) and the surface image of the printed surface of the printed matter after embedding. x , k y The correction coefficient calculation unit 11 calculates the correction coefficient k for ink penetration based on the surface image of the printed surface of the printed matter before embedding (see FIG. 2(a)) and the surface image of the printed surface of the printed matter after embedding. The correction unit 12 corrects the cross-sectional image of the printed matter (see FIG. 2(b)) taken after cutting the printed matter after embedding based on the correction coefficient k x , k y .

[0049] The image analysis method is an image analysis method performed by a computer (for example, the control unit 10 or the image analysis apparatus 1). Based on the surface image of the printed surface of the printed matter before embedding (see FIG. 2(a)) and the surface image of the printed surface of the printed matter after embedding, the correction coefficient k for ink penetration is calculated (step S13 in FIG. 3). The cross-sectional image of the printed matter (see FIG. 2(b)) taken after cutting the printed matter after embedding is corrected based on the correction coefficient k x , k y (step S14). x , k y

[0050] The program causes a computer (for example, the control unit 10 or the image analysis apparatus 1) to calculate the correction coefficient k for ink penetration based on the surface image of the printed surface of the printed matter before embedding (see FIG. 2(a)) and the surface image of the printed surface of the printed matter after embedding (step S13), and correct the cross-sectional image of the printed matter (see FIG. 2(b)) taken after cutting the printed matter after embedding based on the correction coefficient k x , k y (step S14). x , k y

[0051] By embedding the printed matter, for example, the ink distribution fluctuates due to ink bleeding into the resin or deformation during resin curing. Therefore, the cross-sectional image taken after cutting the printed matter after embedding is corrected with the correction coefficient k obtained from the comparison of the surface images before and after embedding x , k yBy performing correction based on this, the ink distribution in the cross-sectional image taken after embedding can be brought closer to the actual ink distribution before embedding. Furthermore, compared with a mode (comparative example) in which embedding is omitted, a printed matter is cut using a focused ion beam apparatus or the like, and cross-sectional images are taken with both an electron microscope apparatus and an optical microscope, it is possible to evaluate the ink distribution even with a simple configuration that does not use a special apparatus. Therefore, according to the present embodiment, it is possible to accurately evaluate the distribution of ink that has penetrated into a printed matter with a simple configuration.

[0052] Also, in the present embodiment, the evaluation value calculation unit 13 calculates an evaluation value of ink penetration of the printed matter based on a plurality of cross-sectional images corrected by the correction unit 12 (steps S21 to S30 in FIG. 4).

[0053] Thereby, based on the calculated evaluation value, it is possible to quantitatively evaluate the three-dimensional distribution of the ink that has penetrated into the printed matter.

[0054] Note that the present invention is not limited to the above-described embodiment as it is, and components can be modified and embodied without departing from the gist thereof at the implementation stage. Also, various inventions can be formed by appropriately combining a plurality of components disclosed in the above-described embodiment. For example, all the components shown in the embodiment may be appropriately combined. Needless to say, various modifications and applications are possible within the scope not departing from the gist of the invention. Hereinafter, the invention described in the claims of the present application at the time of filing is appended.

[0055] [Appendix 1] A correction coefficient calculation unit that calculates a correction coefficient for ink penetration based on a surface image of the printing surface of the printed matter before embedding and a surface image of the printing surface of the printed matter after embedding, A correction unit that corrects a cross-sectional image of the printed matter taken by cutting the printed matter after embedding based on the correction coefficient An image analysis apparatus, characterized by comprising:

[0056] [Appendix 2] The apparatus further includes an evaluation value calculation unit that calculates an evaluation value of ink penetration based on the plurality of cross-sectional images after correction by the correction unit. The image analysis apparatus according to appended claim 1, characterized in that.

[0057] [Appended claim 3] An image analysis method performed by a computer, calculating a correction coefficient for ink penetration based on a surface image of the printed surface of the printed matter before embedding and a surface image of the printed surface of the printed matter after embedding; correcting a cross-sectional image of the printed matter taken after cutting the printed matter after embedding based on the correction coefficient An image analysis method characterized by including the above.

[0058] [Appended claim 4] a function of calculating a correction coefficient for ink penetration based on a surface image of the printed surface of the printed matter before embedding and a surface image of the printed surface of the printed matter after embedding; a function of correcting a cross-sectional image of the printed matter taken after cutting the printed matter after embedding based on the correction coefficient A program for causing a computer to realize the above.

Explanation of reference numerals

[0059] 1 Image analysis apparatus 10 Control unit 11 Correction coefficient calculation unit 12 Correction unit 13 Evaluation value calculation unit 20 Storage unit 30 Display unit 40 Input unit 50 Interface unit

Claims

1. A correction coefficient calculation unit that calculates a correction coefficient for ink penetration based on a surface image of a printed surface of a printed matter before embedding and a surface image of the printed surface of the printed matter after embedding; A correction unit that corrects a cross-sectional image of the printed matter taken by cutting the printed matter after embedding based on the correction coefficient An image analysis apparatus characterized by comprising:

2. Further comprising an evaluation value calculation unit that calculates an evaluation value of ink penetration of the printed matter based on a plurality of the cross-sectional images corrected by the correction unit The image analysis apparatus according to claim 1, characterized in that:

3. An image analysis method performed by a computer, comprising: Calculating a correction coefficient for ink penetration based on a surface image of a printed surface of a printed matter before embedding and a surface image of the printed surface of the printed matter after embedding; Correcting a cross-sectional image of the printed matter taken by cutting the printed matter after embedding based on the correction coefficient An image analysis method characterized by including:

4. A function of calculating a correction coefficient for ink penetration based on a surface image of a printed surface of a printed matter before embedding and a surface image of the printed surface of the printed matter after embedding; A function of correcting a cross-sectional image of the printed matter taken by cutting the printed matter after embedding based on the correction coefficient A program for causing a computer to realize:

Images