Image analysis device

Abstract

The invention relates to an image analysis device, and aims to easily visualize features of an image representing the structure of a material. An image analysis device (4) is provided with: a conversion unit (41) that obtains a plurality of power spectrums by Fourier-transforming each of a plurality of original images indicating the structure of a material; an analysis unit (42) that performs principal component analysis on the plurality of power spectrums and acquires principal components and principal component scores; and a reconstruction unit (43) that reconstructs the image on the basis of the principal component and the principal component score, and outputs a feature image indicating features of the original image. A reconstruction unit (43) generates a difference image between the reference image and the changed image reconstructed by changing the principal component score, and outputs the generated difference image as a feature image. The principal component score to be changed may be determined using a learned regression classification model constructed by machine learning.

Description

Technical Field

[0001] The present invention relates to an image analyzing device. Background Art

[0002] Patent document 1 discloses a new substance search method, which uses machine learning to learn the relationship between structural information and physical property information of known substances, inputs target physical properties into the obtained learned model, and thereby determines at least one candidate substance.

[0003] Prior art literature

[0004] Patent Document 1: Japanese Patent Application Publication No. 2017-91526 Summary of the invention

[0005] Technical issues to be solved by the invention

[0006] The feature quantity is extracted from the image obtained by photographing the material, and it is determined whether it is an important feature quantity that affects the performance of the material, and it is used for the development of new materials, etc. At this time, it is desired to visualize which area in the image changes due to the change of the feature quantity.

[0007] The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to easily visualize the features of an image representing the structure of a material.

[0008] Technical solutions to solve problems

[0009] In order to solve the above-mentioned problems, the image analysis device of the present invention is characterized in that it comprises: a transformation unit, which performs Fourier transform on multiple original images representing the structure of a material to obtain multiple power spectra; an analysis unit, which performs principal component analysis on the multiple power spectra to obtain principal components and principal component scores; and a reconstruction unit, which reconstructs the image based on the principal components and the principal component scores, and outputs a feature image representing the characteristics of the original image, the reconstruction unit generates a differential image between a changed image reconstructed by changing the principal component scores and a reference image, and outputs the generated differential image as the feature image.

[0010] Effects of the Invention

[0011] According to the present invention, it is possible to easily visualize the features of an image representing the structure of a material. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a diagram showing the configuration of an information processing system including the image analysis device according to the first embodiment.

[0013] Figure 2 Yes Figure 1 A diagram for explaining the function of the image analysis device shown.

[0014] Figure 3 It means by Figure 1 Flowchart of processing performed by the image analysis device shown.

[0015] Figure 4 This is a diagram for explaining the image analysis device according to the second embodiment.

[0016] Figure 5 This is a diagram for explaining the image analysis device according to the third embodiment.

[0017] Figure 6 This is a diagram for explaining the image analysis device according to the fourth embodiment.

[0018] Description of symbols

[0019] 1 information processing system, 2 user terminal, 21 processing device, 22 display device, 3 server, 4 image analysis device, 40 processing device, 41 conversion unit, 42 analysis unit, 43 reconstruction unit, 45 storage device, N network DETAILED DESCRIPTION

[0020] Hereinafter, embodiments of the present invention will be described using the drawings. Configurations or functions denoted by the same reference numerals in each embodiment have the same configurations or functions in each embodiment unless otherwise specified, and description thereof will be omitted.

[0021] [First embodiment]

[0022] use Figure 1 to Figure 3 , the image analysis device 4 according to the first embodiment is described. Figure 1 It is a diagram showing the configuration of an information processing system 1 including an image analysis device 4 according to the first embodiment. Figure 2 Yes Figure 1 The figure explains the function of the image analysis device 4 shown.

[0023] The information processing system 1 is a system that analyzes an image input by a user and displays the analysis result to the user by an image analysis device 4. The information processing system 1 includes a user terminal 2 and a server 3 having the image analysis device 4. The user terminal 2 and the server 3 are connected via a network N so as to be able to communicate.

[0024] The user terminal 2 includes a processing device 21 and a display device 22. The processing device 21 includes a processor and a memory, and the processor executes a program to implement the functions of the user terminal 2. The display device 22 includes a display, and displays the processing result of the processing device 21.

[0025] The user terminal 2 receives an image input by the user and transmits it to the image analysis device 4 of the server 3. The user terminal 2 receives the analysis result of the image analysis device 4 from the server 3 and displays it to present it to the user.

[0026] The image analysis device 4 performs the following processing: extracting feature quantities from the image input by the user, and visualizing which area in the image changes due to the change of the feature quantity. The image (hereinafter also referred to as the "original image") that is the analysis object of the image analysis device 4 is, for example, an image representing the structure of the material. The image representing the structure of the material is, for example, an image obtained by photographing the material using a microscope, or an image obtained by simulating the photographed image. The material is, for example, a metal material, a resin material, or a coating material used in a vehicle. For example, the surface structure of a metal material varies depending on the type of metal combined and its composition. It is not easy for the user to grasp the difference in surface structure based on the image. Therefore, the image analysis device 4 extracts feature quantities from the multiple images input, and visualizes which area in the image changes due to the change of the feature quantity. Thus, the image analysis device 4 enables the user to easily grasp the important feature quantities that affect the performance of the material. The performance of the material is, for example, battery performance, magnetic performance, rigidity, thermoplasticity, tensile performance, or mechanical durability.

[0027] The image analysis device 4 includes a processing device 40 and a storage device 45. The storage device 45 is composed of a memory that stores the image received from the user terminal 2 and the analysis result of the image analysis device 4. The processing device 40 includes a processor and a memory, and each function is realized by the processor executing the program. As each function, the processing device 40 includes a conversion unit 41, an analysis unit 42, and a reconstruction unit 43.

[0028] The transformation unit 41 performs Fourier transformation on the multiple original images to obtain the amplitude spectrum and phase spectrum of each original image. The transformation unit 41 obtains the two-dimensional power spectrum by squaring the obtained amplitude spectrum. Figure 2 As shown in FIG. 1 , the two-dimensional power spectrum is represented in the frequency space as the center being the origin and the power spectrum extending from the center. The conversion unit 41 accumulates the two-dimensional power spectrum in the azimuth direction to convert it into one dimension and obtains the one-dimensional power spectrum. The one-dimensional power spectrum is represented by a curve graph with the horizontal axis being the frequency and the vertical axis being the power value (intensity). Figure 2 In , the one-dimensional power spectrum of the original image is represented as "Xorg". Get a one-dimensional power spectrum from an image.

[0029] Thus, the transform unit 41 performs Fourier transform on the plurality of original images to obtain a plurality of one-dimensional power spectra. The image analyzer 4 can reduce the dimension of the data to be analyzed by obtaining the periodic structural information of the size, shape and arrangement of the structures in the original images as one-dimensional power spectra.

[0030] The analysis unit 42 performs principal component analysis on the multiple one-dimensional power spectra obtained by the transformation unit 41 to obtain principal components and principal component scores. The principal component is represented, for example, as a vector having the first principal component PC1, the second principal component PC2, the third principal component PC3, ... as elements. The principal component represents a spectral component with a large variance in a data set of multiple one-dimensional power spectra. The principal component score represents the proportion of each principal component.

[0031] The reconstruction unit 43 reconstructs the image based on the principal components and principal component scores obtained by the analysis unit 42, and outputs a feature image representing the features of the original image. At this time, the reconstruction unit 43 generates a reconstructed image by changing the principal component scores (hereinafter also referred to as the "changed image"). The reconstruction unit 43 generates a difference image between the generated changed image and a reference image. The reconstruction unit 43 outputs the generated difference image as a feature image. The reference image is, for example, the original image. Alternatively, the reference image is an image reconstructed without changing the principal component scores (hereinafter also referred to as the "pre-change image"). The reconstruction unit 43 of this embodiment uses the pre-change image as the reference image.

[0032] The value of the changed principal component score is input by the user to the user terminal 2, and the user terminal 2 sends it to the server 3, and is set in the reconstruction unit 43. The reconstruction unit 43 changes the principal component score according to the set value. The reconstruction unit 43 reconstructs the one-dimensional power spectrum based on the principal component obtained by the analysis unit 42 and the changed principal component score. Figure 2 In , the one-dimensional power spectrum reconstructed by changing the principal component scores is represented as "Xrec_changed".

[0033] The reconstruction unit 43 calculates a first ratio (Xrec_changed / Xorg), which represents the ratio of the change of the one-dimensional power spectrum reconstructed by changing the principal component score relative to the one-dimensional power spectrum of the original image. The reconstruction unit 43 modulates the amplitude spectrum of the original image based on the calculated first ratio. For example, the reconstruction unit 43 creates a mask that modulates the two-dimensional power spectrum of the original image according to the calculated first ratio. The reconstruction unit 43 modulates the amplitude spectrum of the original image according to the first ratio by multiplying the two-dimensional power spectrum by the created mask. The reconstruction unit 43 generates a changed image by performing an inverse Fourier transform on the amplitude spectrum and phase spectrum modulated based on the first ratio.

[0034] In addition, the reconstruction unit 43 reconstructs the one-dimensional power spectrum based on the principal component obtained by the analysis unit 42 and the principal component score before the change. Figure 2 In , the one-dimensional power spectrum reconstructed without changing the principal component scores is represented as "Xrec".

[0035] The reconstruction unit 43 calculates a second ratio (Xrec / Xorg) indicating a ratio of change in the one-dimensional power spectrum after reconstruction without changing the principal component score relative to the one-dimensional power spectrum of the original image. The reconstruction unit 43 modulates the amplitude spectrum of the original image based on the calculated second ratio. The reconstruction unit 43 generates an image before the change by performing an inverse Fourier transform on the amplitude spectrum and phase spectrum modulated based on the second ratio.

[0036] The reconstruction unit 43 generates a difference image between the generated image after the change and the image before the change. The reconstruction unit 43 outputs the generated difference image as a feature image representing the feature of the original image. The feature image output from the reconstruction unit 43 is sent to the user terminal 2 and displayed on the display device 22.

[0037] Figure 3 It means by Figure 1 Flowchart of the processing performed by the image analysis device 4 shown.

[0038] In step S1 , the processing device 40 performs Fourier transform on a plurality of original images respectively to obtain an amplitude spectrum and a phase spectrum of each original image.

[0039] In step S2 , the processing device 40 calculates a one-dimensional power spectrum of each original image based on each acquired amplitude spectrum.

[0040] In step S3 , the processing device 40 performs principal component analysis on the calculated plurality of one-dimensional power spectra to obtain principal component vectors and principal component scores.

[0041] In step S4, the processing device 40 reconstructs the one-dimensional power spectrum based on the changed value of the principal component score. The processing device 40 reconstructs the one-dimensional power spectrum based on the unchanged value of the principal component score.

[0042] In step S5, the processing device 40 calculates a first ratio, which represents the ratio of the one-dimensional power spectrum reconstructed by changing the principal component score to the one-dimensional power spectrum of the original image. The processing device 40 calculates a second ratio, which represents the ratio of the one-dimensional power spectrum reconstructed without changing the principal component score to the one-dimensional power spectrum of the original image.

[0043] In step S6, the processing device 40 modulates the amplitude spectrum of the original image based on the calculated first ratio. The processing device 40 modulates the amplitude spectrum of the original image based on the calculated second ratio.

[0044] In step S7, the processing device 40 performs inverse Fourier transform on the amplitude spectrum and phase spectrum modulated at the first ratio to generate a post-transformation image. The processing device 40 performs inverse Fourier transform on the amplitude spectrum and phase spectrum modulated at the second ratio to generate a pre-transformation image.

[0045] In step S8, the processing device 40 calculates the difference between the image after the change and the image before the change, and generates a difference image. The processing device 40 outputs the generated difference image as a feature image representing the feature of the original image and stores it in the storage device 45. The processing device 40 outputs the feature image stored in the storage device 45 to the communication device of the server 3. The communication device of the server 3 sends the difference image output from the processing device 40 to the user terminal 2. The user terminal 2 displays the difference image sent from the server 3 on the display device 22.

[0046] As described above, the image analysis device 4 of the first embodiment includes: a transformation unit 41, which performs Fourier transformation on a plurality of original images representing the structure of the material, respectively, to obtain a plurality of power spectra (one-dimensional power spectra); an analysis unit 42, which performs principal component analysis on the plurality of power spectra, to obtain principal components and principal component scores; and a reconstruction unit 43, which reconstructs the image based on the principal components and principal component scores, and outputs a feature image representing the features of the original image. The reconstruction unit 43 generates a difference image between a changed image reconstructed by changing the principal component scores and a reference image, and outputs the generated difference image as a feature image.

[0047] Thus, the image analysis device 4 obtains the periodic structure information of the original image as a power spectrum, thereby reducing the dimension of the data to be analyzed. The image analysis device 4 performs principal component analysis on multiple power spectra, thereby being able to relatively easily extract the feature quantity of the original image as a principal component score. The image analysis device 4 generates and outputs a differential image between the changed image reconstructed by changing the principal component score and the reference image, thereby being able to directly reflect the changes in the image caused by the changes in the principal component score on the image for expression. Therefore, the image analysis device 4 can easily visualize which area in the image has changed due to the change in the feature quantity, and can make it easy for the user to intuitively grasp it. Thus, the image analysis device 4 can easily visualize the features of the image representing the structure of the material.

[0048] Furthermore, in the image analysis device 4 according to the first embodiment, the reconstruction unit 43 generates a pre-change image reconstructed without changing the principal component score as a reference image, and generates and outputs a difference image.

[0049] Thus, the image analysis device 4 compares the reconstructed images (the image after the change and the image before the change) to generate a difference image, thereby preventing the error generated during the reconstruction from being reflected in the difference image. The image analysis device 4 can easily and accurately visualize which area in the image has changed due to the change in the feature amount. Thus, the image analysis device 4 can easily and accurately visualize the features of the image representing the structure of the material.

[0050] Furthermore, in the image analysis device 4 of the first embodiment, the transformation unit 41 performs Fourier transformation on the original image to obtain an amplitude spectrum and a phase spectrum, and calculates a power spectrum based on the obtained amplitude spectrum. The reconstruction unit 43 calculates a first ratio, and modulates the amplitude spectrum of the original image based on the calculated first ratio, and the first ratio represents the ratio of the power spectrum reconstructed by changing the principal component score relative to the power spectrum of the original image. The reconstruction unit 43 generates a changed image based on the amplitude spectrum and the phase spectrum modulated based on the first ratio.

[0051] Thus, the image analysis device 4 can reconstruct the image after the change and the image before the change without losing the phase information of the original image. The image analysis device 4 can easily establish a correspondence between the area in the image after the change that has changed due to the change in the principal component score and the area in the image before the change. The image analysis device 4 can more easily and accurately visualize which area in the image has changed due to the change in the feature quantity. Thus, the image analysis device 4 can more easily and accurately visualize the features of the image representing the structure of the material.

[0052] Furthermore, in the image analysis device 4 of the first embodiment, the reconstruction unit 43 calculates a second ratio, and modulates the amplitude spectrum of the original image based on the calculated second ratio, wherein the second ratio indicates a ratio of change of the power spectrum after reconstruction relative to the power spectrum of the original image without changing the principal component scores. The reconstruction unit 43 generates an image before the change based on the amplitude spectrum and phase spectrum modulated based on the second ratio.

[0053] Thus, the image analysis device 4 can easily and accurately establish correspondence between the area in the image after the change that has changed due to the change in the principal component score and the area in the image before the change. The image analysis device 4 can more easily and more accurately visualize which area in the image has changed due to the change in the feature quantity. Thus, the image analysis device 4 can more easily and more accurately visualize the features of the image representing the structure of the material.

[0054] [Second embodiment]

[0055] use Figure 4Next, an image analysis device 4 according to the second embodiment will be described. In the image analysis device 4 according to the second embodiment, descriptions of the same configurations and operations as those of the first embodiment will be omitted. Figure 4 This is a diagram for explaining the image analysis device 4 according to the second embodiment.

[0056] Preferably, when the one-dimensional power spectrum of the original image has peaks in multiple frequency ranges, it is possible to visualize which region in the image the peaks in each frequency range represent. In the image analysis device 4 of the second embodiment, a differential image is generated and outputted only in a frequency range specified by a user.

[0057] Specifically, if Figure 4 As shown in the reference numeral 101, when the principal component score after the user changes is set, as shown in Figure 4 As shown in the reference numeral 102, the reconstruction unit 43 of the second embodiment reconstructs the one-dimensional power spectrum after changing the principal component score and the one-dimensional power spectrum without changing the principal component score. Figure 4 As shown in the reference numeral 103, the reconstruction unit 43 calculates a differential power spectrum, which represents the difference between the one-dimensional power spectrum reconstructed by changing the principal component scores and the one-dimensional power spectrum reconstructed without changing the principal component scores. For example, the reconstruction unit 43 calculates the differential power spectrum by subtracting each power value of the one-dimensional power spectrum reconstructed without changing the principal component scores from each power value of the one-dimensional power spectrum reconstructed by changing the principal component scores.

[0058] And, if Figure 4 As shown in the reference numerals 104 and 105, when the frequency range is specified by the user, Figure 4 As shown in the reference numeral 106 of , the reconstruction unit 43 extracts the differential power spectrum of the specified frequency range from the calculated differential power spectrum. At this time, the reconstruction unit 43 sets the portion outside the specified frequency range of each power value of the one-dimensional power spectrum after the principal component score is changed to the same value as each power value of the one-dimensional power spectrum without changing the principal component score. Thus, the reconstruction unit 43 can extract the differential power spectrum of the specified frequency range.

[0059] And, if Figure 4 As shown in the reference numeral 107, the reconstruction unit 43 generates and outputs a differential image of a specified frequency range based on the extracted differential power spectrum. Figure 4 Reference numeral 108 denotes a differential image of the entire frequency range.

[0060] Thus, the image analysis device 4 of the second embodiment can generate a differential image limited to a specified frequency range even when the one-dimensional power spectrum of the original image has peaks in multiple frequency ranges. The image analysis device 4 of the second embodiment can visualize which region in the image the peaks of each of the multiple frequency ranges represent. Thus, the image analysis device 4 of the second embodiment can visualize the features of the image representing the structure of the material in more detail.

[0061] [Third embodiment]

[0062] use Figure 5 An image analysis device 4 according to a third embodiment will be described. In the image analysis device 4 according to the third embodiment, descriptions of the same configurations and operations as those of the first to second embodiments will be omitted. Figure 5 This is a diagram for explaining the image analysis device 4 according to the third embodiment.

[0063] The image analysis device 4 according to the third embodiment superimposes a difference image obtained by converting each power value of the difference power spectrum into an absolute value on an original image and outputs the resultant image.

[0064] Specifically, if Figure 5 As shown in the reference numeral 201, the reconstruction unit 43 of the third embodiment calculates the differential power spectrum by the same method as the reconstruction unit 43 of the second embodiment to generate a differential image. Figure 5 As shown in the reference numeral 202, the reconstruction unit 43 converts each power value of the differential power spectrum into an absolute value, and generates a differential image after the absolute value conversion. Figure 5 The absolute value difference image shown in the reference numeral 202 is Figure 5 The original image shown in the reference numeral 203 overlaps. And, as Figure 5 As shown in reference numeral 204 , the reconstruction unit 43 outputs a composite image obtained by superimposing the absolute valued differential image and the original image.

[0065] Thus, the image analysis device 4 of the third embodiment makes the difference image overlap with the original image, and thus can make it easier for the user to understand and visualize which area in the image has changed due to the change in the feature amount. The image analysis device 4 of the third embodiment generates a difference image after the absolute value is converted, and thus can make it easier for the user to understand and visualize the area in the image that has undergone a large change due to the change in the feature amount. Thus, the image analysis device 4 of the third embodiment can make it easier for the user to understand and visualize the features of the image representing the structure of the material.

[0066] [Fourth embodiment]

[0067] use Figure 6 Next, an image analysis device 4 according to a fourth embodiment will be described. In the image analysis device 4 according to the fourth embodiment, descriptions of the same configurations and operations as those of the first to third embodiments will be omitted. Figure 6 This is a diagram for explaining the image analysis device 4 according to the fourth embodiment.

[0068] The image analysis device 4 according to the fourth embodiment generates and outputs a difference image by separating a frequency range in which each power value of the difference power spectrum indicates a positive value from a frequency range in which each power value indicates a negative value.

[0069] Specifically, if Figure 6 As shown in the reference numeral 301, the reconstruction unit 43 of the fourth embodiment calculates the differential power spectrum by the same method as the reconstruction unit 43 of the second embodiment. In addition, the reconstruction unit 43 determines the frequency range (fpos) in which each power value of the differential power spectrum represents a positive value, and the frequency range (fneg) in which each power value represents a negative value.

[0070] And, if Figure 6 As shown in the reference numeral 302 of FIG. 1 , the reconstruction unit 43 sets each power value of the frequency range (fneg) representing a negative value to the same value as each power value of the one-dimensional power spectrum without changing the principal component score. Thus, the reconstruction unit 43 extracts the differential power spectrum of the frequency range representing a positive value. Similarly, Figure 6 As shown in the reference numeral 303, the reconstruction unit 43 sets each power value of the frequency range (fpos) representing a positive value to the same value as each power value of the one-dimensional power spectrum without changing the principal component score. Thus, the reconstruction unit 43 extracts the differential power spectrum of the frequency range representing a negative value.

[0071] And, if Figure 6 As shown in the reference numeral 304 of FIG. 1 , the reconstruction unit 43 generates and outputs a differential image based on the differential power spectrum representing the frequency range of positive values. Figure 6 As shown in the reference numeral 305, the reconstruction unit 43 generates and outputs a differential image based on the differential power spectrum of the frequency range representing negative values. In this way, the reconstruction unit 43 generates and outputs a differential image of the frequency range representing positive values ​​and a differential image of the frequency range representing negative values ​​separately.

[0072] Thus, the image analysis device 4 of the fourth embodiment can generate a differential image by dividing the frequency range where the power value of the one-dimensional power spectrum increases and the frequency range where the power value decreases due to the change of the principal component score. The image analysis device 4 of the fourth embodiment divides and visualizes the characteristics of the region in the image represented by the frequency range that has a positive correlation with the change of the feature amount and the frequency range that has a negative correlation with the change of the feature amount. Thus, the image analysis device 4 of the fourth embodiment can visualize the characteristics of the image representing the structure of the material in more detail.

[0073] In addition, the user terminal 2 may receive a plurality of material performance values ​​(material characteristics) input by the user and respectively associated with a plurality of original images, and transmit the plurality of material performance values ​​to the image analysis device 4 of the server 3. The storage device 45 of the image analysis device 4 stores the plurality of original images and the plurality of material performance values ​​received from the user terminal 2 in association with each other.

[0074] The processing device 40 of the image analysis device 4 may also include a determination unit that determines the principal component score of the important feature quantity that affects the performance of the material based on the result of the principal component analysis of the analysis unit 42. The determination result of the determination unit is sent to the user terminal 2 and displayed on the display device 22. The user can grasp the principal component score of the important feature quantity that affects the performance of the material. The user can change the value of the principal component score of the important feature quantity to input to the user terminal 2. The value of the principal component score input by the user is sent to the server 3 and is set in the reconstruction unit 43.

[0075] The reconstruction unit 43 changes the principal component score determined by the determination unit to be an important feature quantity, and reconstructs the one-dimensional power spectrum based on the changed principal component score to generate a changed image.

[0076] The determination unit has a regression classification model that uses the principal component scores obtained by the analysis unit 42 as explanatory variables and each material performance value as a target variable. The regression classification model is constructed by machine learning. The regression classification model is constructed using a well-known model. For example, the regression classification model is constructed using a linear model such as Lasso or Ridge, or a decision tree model such as Random Forest. The determination unit calculates the contribution of the explanatory variable to the prediction of the target variable from the learned regression classification model. The contribution is calculated using a well-known method. For example, when the regression classification model is a linear model, the contribution is calculated based on the regression coefficient. For example, when the regression classification model is a decision tree model, the contribution is calculated based on Gini impurity or cross entropy.

[0077] The image analysis device 4 is provided with the aforementioned determination unit, and can limit the principal component scores of the object to be changed to the principal component scores of the important characteristic quantities that affect the performance of the material. Therefore, the image analysis device 4 can visualize which region in the image changes due to the change of the important characteristic quantities that affect the performance of the material. Thus, the image analysis device 4 can visualize the features of the image representing the structure of the material, and can visualize the correlation between the material performance and the characteristic quantities of the image.

[0078] The embodiments of the present invention are described in detail above, but the present invention is not limited to the above embodiments, and various changes can be made within the scope of the spirit of the present invention. The present invention can add the structure of a certain embodiment to the structure of other embodiments, replace the structure of a certain embodiment with other embodiments, and / or delete a part of the structure of a certain embodiment.

Claims

1. An image analysis device, characterized in that: have: a transforming unit that performs Fourier transform on a plurality of original images representing the structure of the material to obtain a plurality of power spectra; an analyzing unit that performs principal component analysis on the plurality of power spectra to obtain principal components and principal component scores; as well as a reconstruction unit that reconstructs the image based on the principal component and the principal component score, and outputs a feature image representing the feature of the original image, The reconstruction unit generates a difference image between a modified image reconstructed by changing the principal component score and a reference image, and outputs the generated difference image as the feature image.

2. The image analysis device according to claim 1, characterized in that The reconstruction unit generates a pre-change image reconstructed without changing the principal component score as the reference image, and generates and outputs the difference image.

3. The image analysis device according to claim 2, characterized in that The transform unit performs the Fourier transform on the original image to obtain an amplitude spectrum and a phase spectrum, and calculates the power spectrum based on the obtained amplitude spectrum. The reconstruction unit, calculating a first ratio, the first ratio indicating a ratio of a change in the power spectrum reconstructed by changing the principal component score relative to the power spectrum of the original image, Based on the calculated first ratio, the amplitude spectrum of the original image is modulated, The changed image is generated based on the amplitude spectrum and the phase spectrum modulated based on the first ratio.

4. The image analysis device according to claim 3, characterized in that The reconstruction unit, calculating a second ratio, the second ratio indicating a ratio of a change in the power spectrum of the original image after the reconstructed power spectrum without changing the principal component score, Based on the calculated second ratio, the amplitude spectrum of the original image is modulated, The pre-change image is generated based on the amplitude spectrum and the phase spectrum modulated based on the second ratio.

5. The image analysis device according to claim 2, characterized in that The reconstruction unit, calculating a differential power spectrum, the differential power spectrum indicating the difference between the power spectrum reconstructed by changing the principal component score and the power spectrum reconstructed by not changing the principal component score, extracting the differential power spectrum of a specified frequency range from the calculated differential power spectrum, Based on the extracted differential power spectrum, the differential image of the designated frequency range is generated and output.

6. The image analysis device according to claim 2, characterized in that The reconstruction unit, calculating a differential power spectrum, the differential power spectrum indicating the difference between the power spectrum reconstructed by changing the principal component score and the power spectrum reconstructed by not changing the principal component score, determining a frequency range in which each power value of the calculated differential power spectrum represents a positive value and a frequency range in which each power value represents a negative value, The difference image in the frequency range representing the positive value and the difference image in the frequency range representing the negative value are divided and generated, and the generated difference images are output.

Images