Evaluation method, evaluation apparatus, and evaluation program for evaluating the fiber orientation of a test product.

Abstract

To evaluate fiber orientation highly accurately even for a material with weak fiber orientation.SOLUTION: A method for evaluating fiber orientation of a test article includes the steps of: (A) preparing an image of the test article whose fiber orientation is to be evaluated; (B) dividing the image papered in the prior step into a plurality of gratings; (C) obtaining a fiber orientation angle for each grating for each grating divided in the prior step; (D) creating the relative frequency distribution of the grating from the fiber orientation angle for each grating obtained in the prior step; (E) obtaining an inclined elliptic function with the relative frequency of a fiber portion of each grade as length and with an orientation angle θr as a deflection angle by dividing a noise function due to a grating shape from the relative frequency distribution obtained in the prior step; and (F) calculating an index for evaluating the fiber orientation of the test article from the inclined elliptic function in the prior step.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to an evaluation method, an evaluation apparatus, and an evaluation program for evaluating the fiber orientation of a test article, and more particularly, to an evaluation method, an evaluation apparatus, and an evaluation program for evaluating the fiber orientation of food.

Background Art

[0002] As a technique for evaluating the fiber orientation of an object, for example, Non-Patent Document 1 below discloses a technique for evaluating the internal fiber orientation of industrial materials such as cement. Specifically, in Non-Patent Document 1, a binarized image of a material is divided into a plurality of grids, and for each of the divided grids, the angle of the fiber is linearly approximated to obtain the fiber orientation angle for each grid. A relative frequency distribution of the grids is created from the fiber orientation angles, and a polar coordinate distribution is created by converting the orientation angle to the declination angle in a plane coordinate system with respect to the relative frequency of each class. The fiber orientation is evaluated by approximating the distribution obtained here as an inclined ellipse (fiber orientation angle = inclination angle of the ellipse, fiber orientation strength = ratio of the long side to the short side of the ellipse).

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the prior art described in Non-Patent Document 1, materials such as cement, which can clearly recognize each fiber as a line, are used as the evaluation target. However, when analyzing materials with weak orientation such as food, "noise caused by the grid shape" inherent in the analysis method occurs, and it has been found that the analysis results include large errors. Even in Non-Patent Document 1, in order to reduce the influence of noise, when calculating the angle by linearly approximating each cell, a process of excluding cells with a mean square residual greater than a certain value (cells with too large errors with respect to the linear approximation curve) is performed. However, this removes "noise caused by the blurriness of the image itself" and cannot remove noise inherent in the analysis method. In addition, when the test item is food, since the original fiber orientation is weak, the overall image is unclear, and it is also difficult to define the noise range.

[0005] The present invention has been made to solve the above problems, and an object thereof is to accurately evaluate the fiber orientation even for materials with weak fiber orientation.

Means for Solving the Problems

[0006] In order to solve the above problems, the present invention includes the following aspects. Item [1]. A method for evaluating the fiber orientation of a test item having the following steps: (A) A step of preparing an image of a test item for which fiber orientation is to be evaluated, (B) A step of dividing the image prepared above into a plurality of grids, (C) For each of the grids divided above, a step of obtaining the fiber orientation angle for each grid, (D) A step of creating a relative frequency distribution of the grids from the fiber orientation angles for each grid obtained above, (E) Excluding the "noise function caused by the grid shape" from the relative frequency distribution obtained above, and using "the relative frequency of the fiber portion of each class as the length and the orientation angle θ as the deflection angle to obtain an inclined ellipse function", and (F) A step of calculating an index for evaluating the fiber orientation of the test item from the inclined ellipse function obtained above Item [2]. The evaluation method described in Section 1, wherein the image prepared in step (A) above is an image obtained by imaging a region in the test product in which the fiber orientation is homogeneous. Section 3. The evaluation method described in item 1 or 2, wherein the image prepared in step (A) is an image obtained by binarizing an image obtained by imaging the test product. Section 4. The evaluation method described in item 3, wherein the threshold used in the binarization process is the brightness of pixels in the image that are ranked from the lowest to 55-80% in brightness. Section 5. An evaluation method described in any one of items 1 to 4, wherein the shape of the grid divided in step (B) above is rectangular. Section 6. The evaluation method described in item 5, wherein the rectangle is a square. Section 7. An evaluation method described in any one of items 1 to 6, wherein the index used to evaluate the fiber orientation calculated in (F) above is orientation strength k, orientation strength k' and / or orientation flatness f. Section 8. A method for evaluating the physical properties and / or texture of a test product, comprising a step of linking an index for evaluating fiber orientation obtained in step (F) of the evaluation method described in any one of items 1 to 7 with the physical properties, appearance, and / or texture of the test product. Section 9. An evaluation apparatus for evaluating the fiber orientation of a test product having the following parts: (a) Image preparation unit for preparing images of the test product to be evaluated for fiber orientation, (b) A division unit that divides the image prepared above into a plurality of grids, (c) An angle calculation unit that determines the fiber orientation angle for each of the grids divided above, (d) A relative frequency distribution generation unit that generates a relative frequency distribution of the grid from the fiber orientation angles for each grid obtained above. (e) Subtract the "noise function due to the grid shape" from the relative frequency distribution obtained above, and set the relative frequency of the fiber portion of each class as the length, and the orientation angle θ r A tilt elliptic function calculation unit that calculates a tilt elliptic function with an argument of , (f) An index calculation unit that calculates an index for evaluating the fiber orientation of the test product from the tilt elliptic function obtained above. Section 10. The evaluation apparatus described in item 9, wherein the image prepared by the image preparation unit is an image obtained by imaging a region in the test product in which the fiber orientation is uniform. Section 11. The evaluation apparatus according to item 9 or 10, wherein the image prepared by the image preparation unit is an image obtained by binarizing an image obtained by imaging the test product. Section 12. The evaluation apparatus described in item 11, wherein the threshold used in the binarization process is the brightness of pixels in the image that are ranked from the lowest to 55-80% in brightness. Section 13. An evaluation apparatus according to any one of items 9 to 12, wherein the shape of the grid divided by the dividing section is rectangular. Section 14. The evaluation apparatus described in item 13, wherein the rectangle is a square. Section 15. An evaluation device according to any one of items 9 to 14, wherein the index for evaluating fiber orientation calculated by the index calculation unit is orientation strength k, orientation strength k' and / or orientation flatness f. Section 16. An evaluation device for the physical properties and / or texture of a test product, comprising an index for evaluating fiber orientation obtained by the index calculation unit of an evaluation device described in any of items 9 to 15, and a linking unit for linking the physical properties, appearance, and / or texture of the test product. Section 17. An evaluation program that uses a computer to perform the following steps to assess the fiber orientation of a test sample: (A) A step of preparing images of the test material to be evaluated for fiber orientation, (B) A step of dividing the image prepared above into multiple grids, (C) For each of the grids divided above, a step of determining the angle of fiber orientation for each grid, (D) A step of creating a relative frequency distribution of the grid from the fiber orientation angles for each grid obtained above, (E) Subtract the "noise function due to the grid shape" from the relative frequency distribution obtained above, and set the relative frequency of the fiber portion of each class as the length, and the orientation angle θ r The process of finding a tilt elliptic function with an argument of , (F) A step of calculating an index for evaluating the fiber orientation of the test product from the tilt elliptic function obtained above. [Effects of the Invention]

[0007] According to the present invention, even materials with weak fiber orientation can be evaluated with high accuracy. [Brief explanation of the drawing]

[0008] [Figure 1] This is a block diagram showing the configuration of the evaluation device. [Figure 2] This is a flowchart showing the steps of the evaluation method. [Figure 3] (a) is an example of an image obtained by imaging the test product before binarization, and (b) is an example of an image obtained after binarization of the said image. [Figure 4] This figure shows the fiber orientation angle for each grid in the binarized image shown in Figure 3(b). [Figure 5] (a) is a histogram showing the relative frequency distribution of fiber orientation angles for each grid, and (b) is a figure showing the polar coordinate distribution and tilted ellipse obtained by converting the relative frequency of each class of the relative frequency distribution to planar coordinates with the fiber orientation angle as the argument. [Figure 6] (a) is an example of an image with no orientation whatsoever, and (b) is a histogram of the relative frequency distribution of fiber orientation angles created from the said image using a conventional method. [Figure 7] (a) is a histogram showing the relative frequency distribution after noise reduction from the relative frequency distribution of fiber orientation angles for each grid in the binarized image shown in Figure 3(b), and (b) is a figure showing the polar coordinate distribution and tilted ellipse obtained by converting the relative frequency of each class of the said relative frequency distribution to planar coordinates with the fiber orientation angle as the argument. [Figure 8](a) is a perspective view showing an example of the test product, (b) is a cross-sectional view of the test product, and (c) is a cross-sectional view illustrating the suitability of the analysis area. [Figure 9] (a) shows the values ​​of the long side a, orientation intensity k, and orientation flatness f when the length of the short side b of the ellipse is set to 1, and (b) is a graph showing the relationship between orientation intensity k and orientation flatness f. [Figure 10] This graph shows the relationship between the number of divided pixels, the orientation intensity k, and the orientation angle θr. [Figure 11] This graph shows the relationship between the threshold value used in the binarization process of the image shown in Figure 3(a) and the orientation intensity k obtained by the method of the above embodiment. [Figure 12] (a) is an image obtained by imaging a cross-section of fresh cheese, and (b) is an image obtained by imaging a cross-section of old cheese. [Figure 13] Figure 13 is a graph showing the relationship between the threshold value used in the binarization process and the resulting orientation intensity k. [Figure 14] (a) and (b) are images obtained by binarizing the images shown in Figures 12(a) and (b) with a 75% threshold, respectively. [Figure 15] This is an example of an image that lacks orientation. [Figure 16] (a) is a histogram of the relative frequency distribution of fiber orientation angles created using a conventional method from the image shown in Figure 15, and (b) is a histogram of the relative frequency distribution with further noise removal. [Figure 17] This figure shows the polar coordinate distribution and tilted ellipse obtained by converting the relative frequencies of each class in the relative frequency distribution shown in Figure 16(b) to planar coordinates with the fiber orientation angle as the declination angle. [Figure 18] This is an example of an image with very weak orientation. [Figure 19] (a) is a histogram of the relative frequency distribution of fiber orientation angles created using a conventional method from the image shown in Figure 18, and (b) is a histogram of the relative frequency distribution with further noise removal. [Modes for carrying out the invention]

[0009] Embodiments of the present invention will be described below with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described below, and various modifications are possible without departing from its spirit.

[0010] (Device configuration) Figure 1 is a block diagram showing the configuration of an evaluation device 1 according to one embodiment of the present invention. The evaluation device 1 has the function of evaluating the fiber orientation of a test product, and in this embodiment, the test product is a food product. The evaluation device 1 is connected to an imaging device 2, and in this embodiment, the imaging device 2 is a microscope.

[0011] The evaluation device 1 can be configured as a general-purpose computer and its hardware configuration includes a processor such as a CPU or GPU (not shown), a main memory such as DRAM or SRAM (not shown), and an auxiliary storage device 10 such as an HDD or SSD. The auxiliary storage device 10 stores various programs for operating the evaluation device 1, such as the evaluation program P.

[0012] The evaluation device 1 has, as functional blocks, an image preparation unit 11, a division unit 12, an angle calculation unit 13, a relative frequency distribution creation unit 14, a tilt elliptic function calculation unit 15, an index calculation unit 16, and a linking unit 17. Each of these units may be implemented in hardware by logic circuits or the like, or in software by the processor of the evaluation device 1. In the latter case, each of the above units can be implemented by the processor reading the evaluation program P stored in the auxiliary storage device 10 into the main memory and executing it. The evaluation program P may be downloaded to the evaluation device 1 via a communication network such as the Internet, or it may be installed to the evaluation device 1 via a computer-readable non-temporary recording medium such as a CD-ROM on which the evaluation program P is recorded.

[0013] (Processing procedure) Figure 2 is a flowchart showing the steps of the evaluation method according to this embodiment. Of these steps, step S1 is performed by the imaging device 2, and steps S2 to S8 are performed by the evaluation device 1. Note that all or part of these steps S1 to S8 may be performed manually by a human.

[0014] In step S1, the imaging device 2 images the test product and transmits the obtained image to the evaluation device 1. In this embodiment, the imaging device 2 images the cross-section of the food product, which is the test product. If the fiber orientation of the cross-section is not uniform throughout, it is preferable for the imaging device 2 to image the region in the test product where the fiber orientation is uniform.

[0015] In step S2, the image preparation unit 11 of the evaluation device 1 prepares an image of the test product to be evaluated for fiber orientation. In this embodiment, the image preparation unit 11 acquires an image from the imaging device 2 and then binarizes the image. Step S2 corresponds to step (A) described in the claims.

[0016] Figure 3(a) is an example of an image of the test product before binarization, and Figure 3(b) is an example of an image obtained after binarization. In this example, pixels with a brightness greater than a predetermined threshold are converted to 1 (white), and pixels with a brightness less than a predetermined threshold are converted to 0 (black). However, pixels with a brightness greater than a predetermined threshold may be converted to 0, and pixels with a brightness less than a predetermined threshold may be converted to 1. Binarization makes it possible to clearly distinguish fibers in the image in white or black.

[0017] In this embodiment, the brightness of the pixel ranked 75% from the lowest brightness in the image before binarization is used as the threshold for the binarization process. For example, if the number of pixels is 1 million, the brightness of the pixel ranked 750,000th from the lowest brightness (250,000th from the top) becomes the threshold.

[0018] In step S3, the division unit 12 of the evaluation device 1 divides the image prepared in step S1 into a plurality of grids. The shape and size of the grid are not particularly limited, but in this embodiment, the division unit 12 divides it into 10 pixel × 10 pixel square grids. Step S3 corresponds to step (B) described in the claims.

[0019] In step S4, the angle calculation unit 13 of the evaluation device 1 determines the fiber orientation angle for each grid divided in step S3. In this embodiment, the angle calculation unit 13 determines the fiber orientation angle by linearly approximating the fiber angle in each grid. Step S4 corresponds to step (C) described in the claims.

[0020] Figure 4 shows the fiber orientation angles for each grid in the binarized image shown in Figure 3(b). Note that the fiber orientation angles were not calculated for grids containing only black pixels or only white pixels.

[0021] In step S5, the relative frequency distribution creation unit 14 of the evaluation device 1 creates a relative frequency distribution of grids from the fiber orientation angles for each grid obtained in step S4. In this embodiment, the relative frequency distribution creation unit 14 creates a relative frequency distribution of grids with respect to the number of grids and the fiber orientation angles for each grid. Step S5 corresponds to step (D) described in the claims.

[0022] Figure 5(a) is a histogram showing the relative frequency distribution of fiber orientation angles for each grid, and Figure 5(b) shows the polar coordinate distribution and a tilted ellipse obtained by converting the relative frequency of each class of the relative frequency distribution to a planar coordinate system with the fiber orientation angle as the declination angle. In the prior art described in Non-Patent Literature 1, the polar coordinate distribution obtained here is approximated as a tilted ellipse, thereby approximating the fiber orientation angle θ r The fibrous properties of the image are analyzed by using the tilt angle of the ellipse as the angle of inclination of the ellipse, and expressing the orientation intensity k of the fiber orientation as the ratio of the long side to the short side of the ellipse, k = a / b. Specifically, if the series of the histogram is n (in the case of Figure 5(a), n=20), the relative frequency F n and polar coordinates a, b, θ r The relationship is as follows: [Number]

[0023] For n combinations of x and y, with (x n , y n ), by solving the equation (4), which is the equation of the inclined ellipse, using the least squares method (minimizing the sum of squared residuals), the constants a, b, and θ r can be obtained. Here, assuming a (long side) > b (short side), the orientation strength k of the fiber orientation is calculated as k = a / b. The orientation strength k calculated by the conventional technique from the binary image shown in Fig. 3(b) is 2.40.

[0024] However, in the conventional technique, when analyzing materials with weak orientation such as food, "noise caused by the lattice shape" inherent in the analysis method occurs, and the analysis results include large errors. Also, when the test item is food, since the original fiber orientation is weak, the overall image is unclear, and it is difficult to define the noise range.

[0025] Therefore, in step S6, the inclined ellipse function calculation unit 15 of the evaluation device 1 excludes the "noise function caused by the lattice shape" from the relative frequency distribution obtained in step S5, and uses "the relative frequency of the fiber portion of each class as the length and the orientation angle θ r as the declination angle to obtain the inclined ellipse function". Step S6 corresponds to step (E) described in the claims.

[0026] First, the "noise caused by the lattice shape" will be explained. In this embodiment, the binary image is divided into square cells (10 pixel × 10 pixel), and the fiber orientation angle is linearly approximated for each lattice. This "square cell" is the factor of the "noise caused by the lattice shape".

[0027] Here, Figure 6(b) shows a histogram of the relative frequency distribution of fiber orientation angles created using a conventional method for an image with no orientation whatsoever, as shown in Figure 6(a). In the case of an image without fiber orientation, there should be no distribution by angle, so the histogram should be horizontal. However, in reality, there is noise with peaks at ±45°, as shown by A in Figure 6(b).

[0028] Specifically, F n =B n +A n Therefore, noise A can be expressed as follows:

number

number

[0029] Here, the noise A represented by equations (5) to (8) is a function caused by the shape of the square cell. More specifically, noise A represents the relationship between the difference in transverse distance and the angle θ at that point, for a line passing through the center point of the square cell. If the length of one side of the square cell is 1, then at an angle of 45°, the transverse distance is longest at √2, and therefore the difference in transverse distance is the maximum (√2-1).

[0030] Therefore, this "noise caused by the grid shape" will occur regardless of the type of image being analyzed. When analyzing original images in which individual fibers can be recognized, the "noise caused by the grid shape" is negligibly small. However, when analyzing images of test products such as food with weak fiber orientation, the impact of this noise becomes significant.

[0031] Therefore, in this embodiment, in step S6, by mathematically removing "noise caused by the grid shape" from the relative frequency distribution obtained in step S5, the relative frequency of the fiber portion of each class is used as the length, and the orientation angle θ is used. r We find the "tilting elliptic function with argument [value]".

[0032] Specifically, as shown in Figure 5(a), θ is defined as the number of classes n from the image. n The initial histogram F is the frequency distribution of n Extract the initial histogram F. n This includes "noise caused by grid shape" A n Since it is known that it contains F n Break it down as follows: F n =B n +A n ...(9)

[0033] B n A is the tilt elliptic function which is the fiber information we want to analyze, and n This is "noise caused by the grid shape". In other words, the relative frequency distribution of the obtained orientation angles is "noise caused by the grid shape" A n And, "The relative frequency of the fiber portion of each class is defined as the length, and the orientation angle θ" r Function of a tilted ellipse with argument B n Assume that it is a function of the sum of with . Therefore, F n From A n By removing the elliptic function B, n We seek.

[0034] In step S7, the index calculation unit 16 of the evaluation device 1 calculates an index for evaluating the fiber orientation of the test product from the tilt elliptic function obtained in step S6. In this embodiment, the index for evaluating the fiber orientation of the test product is the orientation strength k. Step S7 corresponds to step (F) described in the claims.

[0035] Specifically, the sloping elliptic function B n If the curve perfectly coincides with the inclined ellipse, the following equation holds true.

number

number

number

[0036] Figure 7(a) is a histogram showing the relative frequency distribution after noise removal from the relative frequency distribution of fiber orientation angles for each grid in the binarized image shown in Figure 3(b), and Figure 7(b) shows the polar coordinate distribution and tilted ellipse obtained by converting the relative frequency of each class in the said relative frequency distribution to planar coordinates with the fiber orientation angle as the argument. The orientation intensity k calculated by the method of this embodiment is 1.92. In this embodiment, the error range of the polar coordinate distribution is also considerably small, indicating that the fiber orientation can be evaluated with high accuracy. On the other hand, in the conventional technique shown in Figures 5(a) and (b), the noise portion at -45 degrees overlapped, resulting in the calculation of an orientation intensity (2.40) that was higher than the actual value.

[0037] Note that a, b, θ r This does not affect the value of but the frequency distribution information of the obtained fibers B n =g(θ n ) If used as is, the noise portion will be removed, resulting in a function that does not sum to 1 (it is not a frequency distribution). Therefore, the obtained B n By normalizing the distribution as follows, the frequency distribution C of the true fiber angles in the original image can be obtained. n This can be determined.

number

[0038] In step S8, the linking unit 17 of the evaluation device 1 links the index for evaluating fiber orientation obtained in step S7 with the physical properties, appearance, and / or texture of the test product.

[0039] The test material covered by this invention is a solid or semi-solid substance containing at least enough fibers or fiber-like components to be imageable. Preferably, it is a food product containing edible fibers. The term "edible fibers" includes both soluble and insoluble fibers.

[0040] Here, the physical properties of the test material include hardness, fluidity, viscoelasticity, adhesion, cohesiveness, fracture strength (tensile strength), heat meltability, stringiness, shape retention, crystallinity, and uniformity (mixture), as well as their stability over time. The appearance of the test material includes visual elements created by the absorption and reflection of light, such as texture, feel, gloss, and color difference. This appearance includes not only the external appearance of the test material but also the appearance of its internal cross-section.

[0041] Furthermore, if the test product is food, its texture includes various sensations felt by the lips, teeth, tongue, palate, and throat from the moment the food is placed in the mouth until it is chewed and swallowed. These textures are not limited to but include chewiness (elasticity), crispness, brittleness, melt-in-the-mouth quality, tongue feel, smoothness, and throat feel. In particular, for solid and semi-solid foods, physical deliciousness (texture) tends to be emphasized and is an important element to be evaluated. Especially, the evaluation of how to bring the texture of artificial meat (alternative meat, cultured meat) and artificial seafood meat (alternative fish meat, cultured fish meat, genome-edited fish meat) closer to the texture of natural meat and seafood meat is important now and will become increasingly important in the future.

[0042] The correlation between the index for evaluating fiber orientation obtained in process S7 and the physical properties, appearance, and / or texture of the test product is carried out by matching the index evaluated for the test product with the physical properties, appearance, and / or texture measured separately for the same test product, and finding a certain correlation between the two. The measurement of the physical properties, appearance, and / or texture of the test product can be appropriately set based on the common technical knowledge of this industry, depending on the evaluation target and measurement target.

[0043] (Regarding the indicators) In the above embodiment, orientation strength k is calculated as an index for evaluating fiber orientation. Orientation strength is a value that ranges from 1 to ∞, and can express the strength of fiber orientation more intuitively. Furthermore, orientation flatness f may be determined as another index for evaluating the fiber orientation of the test product. Orientation flatness f has the following relationship with orientation strength k. f = 1 - 1 / k k = 1 / (1-f) Both orientation intensity k and orientation flattening f are the same indicators for comparison (the evaluation results will not be reversed).

[0044] Figure 9(a) shows the values ​​for the long side a, orientation intensity k, and orientation flatness f when the length b of the short side of the ellipse is set to 1, and Figure 9(b) is a graph showing the relationship between orientation intensity k and orientation flatness f. As shown by the filled area in Figure 9(a), a characteristic of the two indicators is that orientation intensity k is a value that is easy to evaluate when the long side is more than twice the length of the short side, while orientation flatness f is a value that is easy to evaluate when the long side is less than or equal to twice the length of the short side. Another way to look at it is that orientation intensity k can take on an infinite value, making it suitable for evaluating increasing functions (e.g., how many times larger it is when the initial value is 2), while orientation flatness f is a function that can take on a value of 0, making it suitable for evaluating decreasing functions (e.g., how much smaller it is when the initial value is 0.5).

[0045] Furthermore, the indices used to evaluate fiber orientation are not limited to orientation strength k and orientation flatness f; for example, orientation strength k' may also be used. Orientation strength k' is calculated as follows.

[0046] k'=(ba) / b=1-b / a Orientation intensity k' can take on infinity, just like orientation intensity k, but it becomes 0 when there is no orientation. Therefore, it can be said to be an indicator that more clearly represents the difference from the state with no orientation.

[0047] (Regarding the homogeneity of the analysis domain) As described above, in step S1, it is preferable for the imaging device 2 to image a region in the test product where the fiber orientation is uniform. For example, suppose the test product shown in Figure 8(a) is cut by a plane passing through the central axis, and the cross-section shown in Figure 8(b) is obtained. Here, in region R2 shown in Figure 8(c), the fiber orientation differs depending on the part, making it difficult to accurately evaluate the fiber orientation. Therefore, it is preferable to image region R1 where the fiber orientation is uniform overall.

[0048] Here, "region with homogeneous fiber orientation" means, as can be seen from the above, a region in which the fiber orientation (direction of fiber arrangement) is the same or consistent throughout. However, it is sufficient that the region can be evaluated for fiber orientation according to the present invention, and it does not need to be strictly homogeneous.

[0049] (Regarding the number of divided pixels) In the above embodiment, in step S3, the divided section 12 is divided into a 10-pixel × 10-pixel square grid, but the number of divided pixels is not limited to this, and even if the number of divided pixels is changed by ±1 pixel vertically and horizontally, the index (orientation intensity k, orientation angle θ) will still be the same. r It is sufficient if the value does not change much.

[0050] Figure 10(a) shows the number of divided pixels, the orientation intensity k, and the orientation angle θ. r Figure 10(b) shows the relationship between the number of divided pixels and the orientation intensity k. The total number of pixels in the image to be analyzed is 920 × 726 = 667,920 pixels. It can be seen that the orientation intensity k changes very little with increasing or decreasing the number of divided pixels when the number of divided pixels is in the range of 6 pixels × 6 pixels to 30 pixels × 30 pixels. Therefore, when evaluating the relative fiber orientation of multiple test samples, setting the number of divided pixels within the above range can reduce the influence of errors in the analysis conditions on the evaluation results.

[0051] (Regarding the threshold used in the binarization process) When a grayscale image is binarized into black and white, the appearance of the binarized image changes significantly depending on the threshold used in the binarization process, and this also changes the orientation intensity k. In the above embodiment, the threshold used in the binarization process is the brightness of the pixel with the lowest brightness ranking of 75% in the image before binarization. However, the threshold used in the binarization process is not particularly limited as long as it is a value that clearly defines the fiber region in the binarized image. For example, it is preferable that the threshold is the brightness of the pixel with the lowest brightness ranking of 55-80% in the image.

[0052] Figure 11 is a graph showing the relationship between the threshold value used in the binarization process of the image shown in Figure 3(a) and the orientation intensity k obtained by the method of the above embodiment. From this graph, it can be seen that the orientation intensity k does not change much when the threshold value is in the range of 55 to 80%.

[0053] (Regarding conventional fiber properties measurement techniques) This invention is one type of image analysis technology. On the other hand, there is a conventional "measurement technology" that directly measures fiber orientation. Since these are likely to yield the same fiber orientation results, a technical comparison will also be made with these technologies. However, as a fundamental premise, image analysis technology is significantly different from this invention in that it can perform analysis regardless of the measuring device.

[0054] One well-known technique for measuring fiber orientation is the use of second harmonic generation microscopy (SHG) (Francois Tiaho, Estimation of helical angles of myosin and collagen by second harmonic generation imaging microscopy, OPTICS EXPRESS, Vol.15, No.1 (2007), and Jessica C. Mansfield, Collagen reorganization in cartilage under strain probed by polarization sensitive second harmonic generation microscopy, JR Soc. Interface, 16 (2018)). SHG works by irradiating a sample with pulsed laser light at a certain deflection angle. The intensity of the reflected light changes depending on the regular orientation structure (fiber orientation), allowing the orientation of the fibers to be measured externally.

[0055] However, SHG has limitations, such as being able to measure only the portion close to the surface layer. Moreover, fiber properties measurement techniques using SHG are techniques that cannot be performed without the device in question, and are significantly different from image analysis techniques like the present invention, which can perform analysis regardless of the measuring device.

[0056] (Regarding 3D fiber orientation analysis) Although the above embodiment only analyzes two-dimensional images, it is theoretically possible to capture three-dimensional fiber orientation using a tilted ellipsoid. In that case, a three-dimensional cubic voxel is used, an approximation of the tilted ellipsoid is performed, and the noise caused by the cubic voxel is removed to obtain the orientation. If the lengths of the three semi-axes of the ellipsoid are a, b, and c (a ≤ b ≤ c), the three-dimensional orientation intensity k can be expressed as c / a. If c is sufficiently large compared to a and b, and the fibrous properties are such that a ≈ b, then if the two-dimensional image is cut by a plane passing through the line c, the results of the two-dimensional analysis will be equivalent to those of the three-dimensional analysis.

[0057] (Contribution to the Sustainable Development Goals (SDGs)) This invention can be used as a means to achieve the Sustainable Development Goals (SDGs). By utilizing this invention, it is possible to contribute to achieving the Sustainable Development Goals (SDGs) and their targets.

[0058] Specifically, the Sustainable Development Goals (SDGs) are as follows ("Transforming our world: the 2030 Agenda for Sustainable Development," Ministry of Foreign Affairs, Internet <URL: https: / / www.mofa.go.jp / mofaj / gaiko / oda / sdgs / pdf / 000101402.pdf>). Goal 1: End poverty in all its forms everywhere. Goal 2. End hunger, achieve food security and improved nutrition, and promote sustainable agriculture. Goal 3. Ensure healthy lives and promote well-being for all at all ages. Goal 4. Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all. Goal 5. Achieve gender equality and empower all women and girls. Goal 6. Ensure availability and sustainable management of water and sanitation for all. Goal 7. Ensure access to affordable, reliable, sustainable, and modern energy for all. Goal 8. Promote inclusive and sustainable economic growth, full and productive employment and decent work for all. Goal 9. Build resilient infrastructure, promote inclusive and sustainable industrialization and foster innovation. Goal 10. Reduce inequalities within and among countries. Goal 11. Make cities and human settlements inclusive, safe, resilient and sustainable. Goal 12. Ensure sustainable consumption and production patterns. Goal 13. Take urgent action to combat climate change and its impacts. Goal 14. Conserve and sustainably use the oceans, seas and marine resources for sustainable development. Goal 15. Protect, restore and promote the sustainable use of terrestrial ecosystems, sustainably manage forests, combat desertification, and halt and reverse land degradation and biodiversity loss. Goal 16. Promote peaceful and inclusive societies for sustainable development, provide access to justice for all, and establish effective, accountable and inclusive regimes at all levels. Construct a degree. Goal 17. Strengthen the means of implementation and revitalize the global partnership for sustainable development.

[0059] This invention can contribute, in particular, to achieving Goal 2 (end hunger, achieve food security and improved nutrition and promote sustainable agriculture) and Goal 3 (ensure healthy lives and promote well-being for all at all ages). Specifically, by evaluating the fiber orientation of food with high precision, it is possible to objectively quantify and analyze the various textures of food from the moment it is put in the mouth to chewing and swallowing. By measuring and evaluating the physical properties of food, it is possible to reduce the risk of aspiration and accidental ingestion for infants, the elderly, and people with disabilities, and to improve food safety. Furthermore, by analyzing the physical properties of food with high precision using objective numerical values ​​and indicators of the fiber orientation of food, it is possible to accurately grasp the quality of food, thereby reducing food loss, improving productivity, and contributing to the supply of safe and sufficient food for all. In addition, this invention can contribute, in particular, to achieving Goal 9 (build resilient infrastructure, promote inclusive and sustainable industrialization and foster innovation). In other words, by accurately evaluating the fiber orientation of food products, it is possible to promote technological development, research, and innovation, such as diversifying industries and creating added value in products, thereby improving resource utilization efficiency and expanding and improving the efficiency of industrial processes, and promoting innovation in a sustainable and resilient manner. Furthermore, the present invention can contribute in particular to achieving Goal 12 (Ensure sustainable consumption and production patterns). That is, by accurately evaluating the fiber orientation of food products, it is possible to promote technological development, research, and innovation, thereby achieving efficient use of resources, halving food waste, and reducing food losses in production and supply chains, such as post-harvest losses, thereby ensuring sustainable consumption and production patterns through sustainable development and development in harmony with nature. Moreover, the present invention can also contribute to achieving the targets associated with each of the goals outlined in "Transforming Our World: The 2030 Agenda for Sustainable Development." [Examples]

[0060] (Example 1: Verification using fibrous cheese) In Example 1, the orientation strength of fibrous cheese, whose properties change during storage, was quantified. Figure 12(a) is an image obtained by imaging a cross-section of fresh cheese immediately after production, and Figure 12(b) is an image obtained by imaging a cross-section of old cheese after 5 months of storage. In both images, the size of one pixel is 5 μm, and the number of pixels is 400 pixels × 400 pixels (analysis area fixed at 2 mm square).

[0061] These images were binarized, and the orientation intensity k was calculated using the method of the above embodiment. Figure 13 is a graph showing the relationship between the threshold used in the binarization process and the obtained orientation intensity k. In this graph, "New" represents the orientation intensity of fresh cheese shown in Figure 12(a), and "Old" represents the orientation intensity of old cheese shown in Figure 12(b).

[0062] As can be seen from the graph above, the decrease in fibrous properties over time is captured regardless of the threshold, but the following was found from the image analysis results. (1) When the threshold is around 0-55%, it captures black areas where there are no fibers, and therefore cannot be said to be accurately quantified. (2) When the threshold is around 55-80%, the protein fibers to be observed can be accurately quantified. (3) When the threshold is around 90-100%, only the high-brightness parts of the fibers are perceived as lines, resulting in a stronger detection of fiber orientation than is actually the case.

[0063] In Example 1, the protein fibers of the cheese were captured more clearly when the threshold was around 75%. In the image shown in Figure 12, the proteins to be evaluated are white (high brightness), and the structures to be excluded (fat and fat spots) are black. Since the protein content in the image of this example is approximately 25% (26% to be precise), setting the threshold to 75% is considered to be a suitable threshold for capturing the entire protein.

[0064] Therefore, the original images were binarized with a 75% threshold. Figures 14(a) and (b) are the images obtained by binarizing the images shown in Figures 12(a) and (b) with a 75% threshold, respectively. Calculating the orientation intensity and orientation angle from these images, fresh cheese was found to have an orientation intensity k = 3.23 and an orientation angle θ. r =70.7°, and the old cheese has an orientation intensity k=1.75 and an orientation angle θ r The result was 129.1°. This analysis result was consistent with the results of subjective evaluations of fiber properties, etc.

[0065] (Example 2: Verification using images without orientation) When the non-orientational image shown in Figure 15 is analyzed using the conventional method and the method of the present invention, the results are as follows.

[0066] When the image shown in Figure 15 is binarized and divided into multiple grids, the fiber orientation angle is determined for each grid, and a relative frequency distribution of the fiber orientation angles is created, the histogram shown in Figure 16(a) is obtained (conventional technique). In Figure 16(a), the "noise caused by the grid shape" is too large, making analysis using a sloping ellipse (an approximation assuming unimodality) impossible.

[0067] On the other hand, removing noise from the above relative frequency distribution yields the histogram shown in Figure 16(b) (invention). In Figure 16(b), the original angle frequency distribution is obtained (the frequency distribution is constant. Since n=20, F n (= constant at 0.05). In the histogram shown in Figure 16(b), the orientation angle θ r The values ​​are -15.3°, orientation intensity k=1.08, and noise parameter δ=0.144. As shown in Figure 17, the closer the orientation intensity is to 1, the closer the ellipse becomes to a circle, meaning there is no fiber orientation. In other words, it can be seen that the fiber orientation can be accurately evaluated using the method of the present invention.

[0068] (Example 3: Verification using weakly oriented images) When the image with very weak orientation shown in Figure 18 is analyzed using the conventional method and the method of the present invention, the results are as follows.

[0069] When the image shown in Figure 18 is binarized and divided into multiple grids, the fiber orientation angle is determined for each grid, and a relative frequency distribution of the fiber orientation angles is created, the histogram shown in Figure 19(a) is obtained (conventional technology). Furthermore, when noise is removed from this relative frequency distribution, the histogram shown in Figure 19(b) is obtained (invention). The orientation intensity k obtained from the relative frequency distribution shown in Figure 19(a) is 2.98, while the orientation intensity k obtained from the relative frequency distribution shown in Figure 19(b) is 1.82. Therefore, it can be seen that the original fiber orientation of an image can be accurately quantified by the method of the present invention. [Industrial applicability]

[0070] This invention can be used as a means to achieve the Sustainable Development Goals (SDGs), and by using this invention, it is possible to contribute to achieving the Sustainable Development Goals (SDGs) and their targets. In particular, this invention can contribute to achieving Goal 2 (End hunger, achieve food security and improved nutrition and promote sustainable agriculture) and Goal 3 (Ensure healthy lives and promote well-being for all at all ages) of the Sustainable Development Goals (SDGs). Specifically, by evaluating the fiber orientation of food with high precision, it is possible to objectively quantify and analyze various textures from putting food in the mouth to chewing and swallowing. By measuring and evaluating the physical properties of food, it is possible to reduce the risk of aspiration and accidental ingestion for infants, the elderly, and people with disabilities, and to improve food safety. Furthermore, by analyzing the physical properties of food with high precision using objective numerical values ​​and indicators of the fiber orientation of food, it is possible to accurately grasp the quality of food, thereby reducing food loss, improving productivity, and contributing to supplying safe and sufficient food to all people. Furthermore, the present invention can contribute to achieving Goal 9 of the Sustainable Development Goals (SDGs), in particular (build resilient infrastructure, promote inclusive and sustainable industrialization and foster innovation). Specifically, by accurately evaluating the fiber orientation of food, it is possible to promote technological development, research, and innovation such as diversification of industries and creation of added value in products, thereby improving resource utilization efficiency, expanding and improving the efficiency of industrial processes, and promoting sustainable and resilient innovation. The present invention can also contribute to achieving Goal 12 of the Sustainable Development Goals (SDGs), in particular (ensure sustainable consumption and production patterns). Specifically, by accurately evaluating the fiber orientation of food, it is possible to promote technological development, research, and innovation, thereby achieving efficient use of resources, halving food waste, and reducing food losses in production and supply chains such as post-harvest losses, thereby ensuring sustainable consumption and production patterns through sustainable development and development in harmony with nature. [Explanation of Symbols]

[0071] 1. Evaluation device 11 Image preparation section 12 Division 13 Angle calculation unit 14. Relative Frequency Distribution Creation Section 15. Sloping Elliptic Function Calculation Unit 16 Index calculation section 17. Stringing section P Evaluation Program f Orientation flatness k orientation strength k' Orientation strength θ r Orientation angle

Claims

1. A method for evaluating the fiber orientation of a test product having the following process: (A) A step of preparing images of the test product to be evaluated for fiber orientation, (B) A step of dividing the image prepared above into multiple grids, (C) A step of determining the fiber orientation angle for each of the grids divided above, (D) A step of creating a relative frequency distribution of the grid from the fiber orientation angles for each grid obtained above. (E) A step of determining a "noise function due to grid shape" that represents the relationship between the difference in transverse distance and the angle at that time for a straight line passing through the center point of each grid, and (F) A step of calculating an index for evaluating the fiber orientation of the test product by determining the parameters of the tilt elliptic function such that the sum of the squares of the values ​​obtained by removing from the relative frequency distribution obtained above the "tilt elliptic function with the relative frequency of the fiber portion of each class as the length and the orientation angle θr as the argument" and the "noise function due to the grid shape" is minimized.

2. The evaluation method according to claim 1, wherein the image prepared in step (A) is an image obtained by imaging a region in the test product in which the fiber orientation is uniform.

3. The evaluation method according to claim 1 or 2, wherein the image prepared in step (A) is an image obtained by binarizing an image obtained by imaging the test product.

4. The evaluation method according to claim 3, wherein the threshold used in the binarization process is the brightness of pixels in the image that are ranked from the lowest to 55-80% in brightness.

5. The evaluation method according to any one of claims 1 to 4, wherein the index for evaluating fiber orientation calculated in (F) is orientation strength k, orientation strength k' and / or orientation flatness f.

6. A method for evaluating the physical properties and / or texture of a test product, comprising a step of linking an index for evaluating fiber orientation obtained in step (F) of the evaluation method described in any one of claims 1 to 5 with the physical properties, appearance, and / or texture of the test product.

7. An evaluation apparatus for evaluating the fiber orientation of a test product having the following parts: (a) Image preparation unit for preparing images of the test product to be evaluated for fiber orientation, (b) A division unit that divides the image prepared above into a plurality of grids, (c) An angle calculation unit that determines the fiber orientation angle for each of the grids divided above, (d) A relative frequency distribution creation unit that creates a relative frequency distribution of the grids from the fiber orientation angles for each grid obtained above. (e) A tilt elliptic function calculation unit that calculates a "noise function due to grid shape" that represents the relationship between the difference in transverse distance and the angle at that time for a straight line passing through the center point of each grid, and (f) An index calculation unit that calculates an index for evaluating the fiber orientation of the test product by determining the parameters of the tilt elliptic function such that the sum of the squares of the values ​​obtained by removing from the relative frequency distribution obtained above the "tilt elliptic function with the relative frequency of the fiber portion of each class as the length and the orientation angle θr as the argument" and the "noise function due to the grid shape" is minimized.

8. The evaluation apparatus according to claim 7, wherein the image prepared by the image preparation unit is an image obtained by imaging a region in the test product in which the fiber orientation is uniform.

9. The evaluation apparatus according to claim 7 or 8, wherein the image prepared by the image preparation unit is an image obtained by binarizing an image obtained by imaging the test product.

10. The evaluation apparatus according to claim 9, wherein the threshold used in the binarization process is the brightness of pixels in the image that are ranked from the lowest brightness to 55-80%.

11. The evaluation apparatus according to any one of claims 7 to 10, wherein the index for evaluating fiber orientation calculated by the index calculation unit is orientation strength k, orientation strength k' and / or orientation flatness f.

12. An evaluation device for the physical properties and / or texture of a test product, comprising an index for evaluating fiber orientation obtained by the index calculation unit of the evaluation device according to any one of claims 7 to 11, and a linking unit for linking the physical properties, appearance, and / or texture of the test product.

13. An evaluation program that uses a computer to perform the following steps to assess the fiber orientation of a test sample: (A) A step of preparing images of the test product to be evaluated for fiber orientation, (B) A step of dividing the image prepared above into multiple grids, (C) A step of determining the angle of fiber orientation for each of the grids divided above, (D) A step of creating a relative frequency distribution of the grid from the fiber orientation angles for each grid obtained above. (E) A step of determining a "noise function due to grid shape" that represents the relationship between the difference in transverse distance and the angle at that time for a straight line passing through the center point of each grid, and (F) A step of calculating an index for evaluating the fiber orientation of the test product by determining the parameters of the tilt elliptic function such that the sum of the squares of the values ​​obtained by removing from the relative frequency distribution obtained above the "tilt elliptic function with the relative frequency of the fiber portion of each class as the length and the orientation angle θr as the argument" and the "noise function due to the grid shape" is minimized.

Images