Evaluation system, evaluation method, and evaluation program

Abstract

To provide an evaluation system for easily evaluating coating of an object.SOLUTION: An evaluating system includes at least one processor. The at least one processor acquires a target image showing a target having a substrate and a coating area on the substrate, inputs the target image to a trained model that estimates the coating area from an input image, identifies a coating area of the target, and calculates an evaluation value for the identified coating area.SELECTED DRAWING: Figure 1

Description

[[Technical Field]]

[0001] One aspect of the present disclosure relates to an evaluation system, an evaluation method, and an evaluation program. [[Background Art]]

[0002] Patent Document 1 describes a conductive particle shape evaluation apparatus that evaluates the surface shape of conductive particles having a plurality of conductive protrusions on the surface thereof. [[Prior Art Documents]] [[Patent Documents]]

[0003] [[Patent Document 1]] Japanese Unexamined Patent Publication No. 2016-061722 [[Summary of the Invention]] [[Problem to be Solved by the Invention]]

[0004] There is a demand for a mechanism for easily performing evaluation related to coating of an object. [[Means for Solving the Problem]]

[0005] An evaluation system according to one aspect of the present disclosure includes at least one processor. The at least one processor acquires a target image representing an object having a base material and a coating region on the base material, inputs the target image to a trained model that estimates a coating region from an input image, specifies the coating region of the object, and calculates an evaluation value related to the specified coating region.

[0006] An evaluation method relating to one aspect of this disclosure is performed by an evaluation system comprising at least one processor. This evaluation method includes the steps of: acquiring an object image showing an object having a substrate and a coating region on the substrate; inputting the object image into a trained model that estimates the coating region from an input image to identify the coating region of the object; and calculating an evaluation value relating to the identified coating region.

[0007] An evaluation program relating to one aspect of this disclosure involves causing a computer to perform the following steps: acquiring an image of an object having a substrate and a coating region on the substrate; inputting the image of the object into a trained model that estimates the coating region from the input image to identify the coating region of the object; and calculating an evaluation value for the identified coating region.

[0008] In this respect, the coated area of ​​the object is estimated from the target image by a trained model. This configuration makes it easy to identify the coated area and to calculate evaluation values ​​for that coated area. Therefore, the evaluation of the object's coating can be easily performed. [Effects of the Invention]

[0009] According to one aspect of this disclosure, the evaluation of the coating of the object can be easily performed. [Brief explanation of the drawing]

[0010] [Figure 1] This diagram shows the functional configuration of the evaluation system. [Figure 2] This flowchart shows an example of generating an image of an object. [Figure 3] This figure shows an example of image processing related to Figure 2. [Figure 4] This figure shows an example of image processing related to Figure 2. [Figure 5] This flowchart shows an example of generating training images. [Figure 6]This figure shows an example of image processing related to Figure 5. [Figure 7] This is a flowchart showing an example of generating a pre-trained model. [Figure 8] This flowchart shows an example of an evaluation of the coating of an object. [Figure 9] This figure shows an example of image processing related to Figure 8. [Modes for carrying out the invention]

[0011] The embodiments described herein will be described in detail below with reference to the attached drawings. In the description of the drawings, the same or equivalent elements are denoted by the same reference numerals, and redundant descriptions are omitted.

[0012] [System Overview] The evaluation system described herein is a computer system that performs an evaluation of the coating of an object captured in an image. The object has a substrate and a coating area. Coating refers to a state in which at least a part of the substrate is hidden by the coating area in the appearance of the object. The evaluation system performs an evaluation of the coating area, for example, an evaluation of the coating area's coverage of the substrate. Evaluation of the coating area refers to a process of quantitatively determining the coating area that covers the substrate. In one example, the evaluation system calculates an evaluation value, which is a quantitative index of the coating area, and outputs that evaluation value. For example, the evaluation value is a value relating to the coating area's coverage of the substrate.

[0013] The object refers to a solid being evaluated by the evaluation system. The substrate refers to the main component of the object. The coating area refers to the component located on the substrate. The object can have any shape, dimensions, and composition. For example, the object may be spherical, planar, or columnar, or it may have a more complex shape. The object may be visible or too small to be seen without a microscope. The shape and dimensions of the substrate can influence the shape and dimensions of the object. The coating area covers at least a portion of the surface of the substrate. The coating area may be formed by powdery or granular particles adhering to the substrate, or by applying a liquid coating agent or a metal-containing coating agent to the substrate. Multiple coating areas, separated from each other, may be provided on a single substrate. The coating area may be formed by multiple coating elements arranged on the substrate. Examples of coating elements include individual particles. Particles arranged on the substrate as coating elements may be perceived as protrusions. The coating area may be perceived as a raised convex portion above the surface of the substrate, or as a collection of densely packed protrusions. At least some of the components may differ between the substrate and the coating region, or all components may be the same. Both the substrate and the coating region may be organic compounds or inorganic compounds, or may contain both organic and inorganic compounds.

[0014] The object may be particulate matter. In one example, the particulate matter comprises a core particle and a plurality of fine particles arranged on the surface of the core particle. The diameter of the fine particles is smaller than the diameter of the core particle. The core particle is an example of a substrate, the individual fine particles are an example of a coating element, and the collection of multiple fine particles is an example of a coating area.

[0015] The evaluation system identifies a coating region of a target object using a trained model generated by machine learning. The trained model is a computational model that estimates the coating region of the target object from an image representing the target object. Machine learning refers to a method that autonomously discovers laws or rules by repeatedly performing learning based on given information. The evaluation system inputs an image of the target object into the trained model to identify the coating region, and executes evaluation related to the identified coating region. The evaluation system may generate the trained model by machine learning. Generation of the trained model corresponds to a learning phase, and identification of the coating region by the trained model corresponds to an operation phase. The learning model used in the evaluation system may be a model capable of performing instance segmentation or panoptic segmentation, and may be, for example, Mask R-CNN, DeepMask, FCIS (Fully Convolutional Instance-aware Semantic Segmentation), Panoptic Feature Pyramid Network, or UPSNet.

[0016] By introducing a trained model, it becomes possible to automatically and accurately identify the coating region. Depending on the properties of the target object and the coating region, it may be necessary for a person to visually identify the coating region. There is a conventional technique that converts an image of a target object into a binary image that distinguishes the coating region from other regions, and identifies the coating region based on the binary image. However, in this conventional technique, it is necessary for a person to visually set a binarization threshold for each individual image, which takes a long time for processing. Additionally, identification of the coating region may depend on human perception. Furthermore, it is difficult to accurately identify the coating region. In contrast to such conventional techniques, the evaluation system can automatically and accurately identify the coating region, so a user of the evaluation system can easily obtain an evaluation related to coating of the target object.

[0017] [System Configuration] FIG. 1 is a diagram showing a functional configuration of an evaluation system 10 according to an example. The evaluation system 10 includes a processor 101 as a hardware component. The processor 101 is, for example, a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or a GPU (Graphics Processing Unit). The evaluation system 10 further includes, as hardware components, a main storage device configured of RAM and ROM, an auxiliary storage device configured of flash memory, a hard disk, and the like, an input device such as a keyboard and a mouse, an output device such as a monitor and a speaker, and a communication module that performs data communication with an external device. Each functional module of the evaluation system 10 is implemented by the processor 101 executing a program stored in the auxiliary storage device.

[0018] An evaluation program for causing a computer to function as the evaluation system 10 includes program codes for implementing each functional module of the evaluation system 10. This evaluation program may be provided after being recorded in a non-transitory recording medium such as a CD-ROM, a DVD-ROM, or a semiconductor memory, for example. Alternatively, the evaluation program may be provided as a data signal superimposed on a carrier wave via a communication network. The provided evaluation program is stored in, for example, the auxiliary storage device.

[0019] The evaluation system 10 may be configured of one computer, or may be configured of an assembly of a plurality of computers, that is, a distributed system. Examples of computers used for the evaluation system 10 include various types of computers such as personal computers, workstations, tablet terminals, and smartphones. When a plurality of computers are used for the evaluation system 10, these computers are connected via a communication network such as the Internet or an intranet, whereby one logical evaluation system 10 is constructed. The evaluation system 10 may be implemented as a client-server system such as a cloud system, or may be implemented by a stand-alone computer.

[0020] In one example, the evaluation system 10 is connected to at least one external storage device via a communication network. The external storage device is a device or recording medium that stores various types of data used for processing in the evaluation system 10. The external storage device may be a component of the evaluation system 10, or it may be located in a separate computer system. The communication network may be the internet, an intranet, or a combination thereof. The communication network may be a wired network, a wireless network, or a combination thereof.

[0021] Figure 1 shows a first original image database 31, a training image database 32, and a second original image database 33 as examples of external storage. Both the first original image database 31 and the second original image database 33 are storage devices that store at least one original image representing at least one object. The training image database 32 is storage devices that store at least one training image used for machine learning. At least two of the first original image database 31, the training image database 32, and the second original image database 33 may be integrated into a single database.

[0022] In one example, the processor 101 functions as a preprocessor 11, a learning unit 12, and an evaluation unit 13. The preprocessor 11 is used in both the learning phase and the operation phase. The learning unit 12 corresponds to the learning phase, and the evaluation unit 13 corresponds to the operation phase.

[0023] The preprocessing unit 11 is a functional module that generates object images representing a single object. In the learning phase, the preprocessing unit 11 generates a sample image 42, which is an example of an object image, from a first original image 41 read from the first original image database 31. The sample image 42 is used to generate a training image 43. In the operation phase, the preprocessing unit 11 generates a target image 45, which is another example of an object image, from a second original image 44 read from the second original image database 33. The target image 45 is used for evaluation regarding the covering of the object.

[0024] The learning unit 12 is a functional module that generates a trained model 20. In one example, the learning unit 12 includes a training image generation unit 121 and a model generation unit 122. The training image generation unit 121 is a functional module that generates training images 43 based on sample images 42. The model generation unit 122 is a functional module that generates a trained model 20 by machine learning based on the training images 43.

[0025] The evaluation unit 13 is a functional module that performs an evaluation of the coating of an object. In one example, the evaluation unit 13 includes a coating identification unit 131 and a calculation unit 132. The coating identification unit 131 is a functional module that identifies the coating area of ​​an object from an image 45 using a trained model 20. The calculation unit 132 is a functional module that calculates an evaluation value based on the identified coating area.

[0026] [System operation] The following describes an example of processing by the evaluation system 10, as well as an example of the evaluation method related to this disclosure. In the following examples, particulate matter is shown as the target object, and individual fine particles on core particles are shown as coating elements.

[0027] (Generating an image of the target object) The generation of object images will be explained with reference to Figures 2 to 4. Figure 2 is a flowchart showing an example of this process as processing flow S1. Figures 3 and 4 are diagrams showing examples of image processing related to processing flow S1. The evaluation system 10 can execute processing flow S1 in both the learning phase and the operation phase.

[0028] In step S11, the preprocessing unit 11 acquires the original image. In one example, the original image showing particulate matter is an SEM image obtained by imaging multiple parts of particulate matter collected on a carbon tape using a scanning electron microscope (SEM). If auxiliary information such as text or scales is indicated in the original image, the preprocessing unit 11 may remove this auxiliary information by image processing such as cropping to obtain an original image that does not contain the auxiliary information. In the learning phase, the preprocessing unit 11 reads the first original image 41 from the first original image database 31, and in the operation phase, it reads the second original image 44 from the second original image database 33. In the following description of the processing flow S1, the first original image 41 and the second original image 44 are collectively referred to as the "original image".

[0029] In step S12, the preprocessor 11 converts the original image into a binary image. In one example, the preprocessor 11 blurs the original image by applying an averaging filter based on a predetermined kernel size. Subsequently, the preprocessor 11 performs a binarization process, such as Otsu's binarization, on the blurred original image to convert the brightness of the object region, which is the area of ​​the object, to 255 and the brightness of the background region to 0. As a result, a binary image is generated in which the object is represented in white and the background is represented in black. In this binary image, if black dots exist as noise within the white area, the preprocessor 11 removes these black dots. By applying an averaging filter to the original image before the binarization process, it is possible to prevent or suppress the situation in which a large number of black dots occur within the area of ​​the object. Figure 3 shows that the original image 201 is converted into a binary image 202.

[0030] In step S13, the preprocessor 11 identifies the center of each object by distance transformation. Distance transformation of a binary or grayscale image is a process that calculates the distance to the nearest pixel with a brightness of 0 for each pixel whose brightness is not 0, and generates a distance map showing the individual distances. The distance map shows the individual distances in grayscale such that brightness increases as the distance increases. For each white pixel in the object region, the preprocessor 11 calculates the distance to the nearest black pixel in the background region. Based on the distance of each white pixel, the preprocessor 11 generates a distance map corresponding to the binary image. Based on this distance map, the preprocessor 11 extracts regions where the brightness is above a predetermined threshold as the center of each object. In each center, the preprocessor 11 identifies the pixel with the highest pixel value as the center of the object. Figure 3 shows that the binary image 202 is transformed into a distance map 203. In the distance map 203, the individual centers with relatively high brightness correspond to the individual objects in the original image 201. In each center in the distance map 203, the pixel with the highest brightness indicates the center of the object.

[0031] In step S14, the preprocessing unit 11 generates a reference image showing the central part of each object. As described above, this central part is obtained from the distance map. Figure 3 shows that a reference image 204, in which the central part is shown in white and other areas in black, is generated based on the distance map 203. In one example, the preprocessing unit 11 displays this reference image on a monitor. The user can then verify the generation of the object image through this reference image.

[0032] In step S15, the preprocessing unit 11 selects objects whose entirety is visible in the original image. In one example, the preprocessing unit 11 determines the dimensions of an object based on a distance map, using the distance set at the pixel corresponding to the center of the object. The preprocessing unit 11 then assumes the shape of the object based on its center and dimensions. The preprocessing unit 11 selects objects whose entire shape is located within the original image as objects whose entirety is visible in the original image. For example, the preprocessing unit 11 obtains the distance set at the pixel corresponding to the center of the object as the radius of the object. The preprocessing unit 11 then assumes a virtual circle defined by its center and radius as the shape of the object. The preprocessing unit 11 selects objects whose entire virtual circle is located within the original image as objects whose entirety is visible in the original image. Image 205 in Figure 4 shows that seven particulate matter particles whose entirety is visible in the original image 201 have been selected based on the distance map 203.

[0033] In step S16, the preprocessing unit 11 extracts selected objects from the original image and generates object images. The preprocessing unit 11 generates an object image for each of the n selected objects, resulting in n object images. The image group 206 shown in Figure 4 is seven object images corresponding to seven objects selected from the original image 201. The object images are sample images 42 in the learning phase and target images 45 in the operation phase. In the learning phase, the preprocessing unit 11 generates one or more sample images 42 from one first original image 41. The preprocessing unit 11 may store the sample images 42 in a predetermined storage device, or it may output the sample images 42 to the learning unit 12 for the generation of a training image 43. In the operation phase, the preprocessing unit 11 generates one or more target images 45 from one second original image 44. The preprocessing unit 11 may store the target images 45 in a predetermined storage device, or it may output the target images 45 to the evaluation unit 13 for evaluation of the object's coverage.

[0034] The evaluation system 10 may execute the processing flow S1 multiple times or repeatedly in both the learning phase and the operation phase.

[0035] (Generating training images) The generation of training images will be explained with reference to Figures 5 and 6. Figure 5 is a flowchart showing an example of this process as processing flow S2. Figure 6 is a diagram showing an example of image processing related to processing flow S2. Processing flow S2 corresponds to the generation of training data used in the learning phase.

[0036] In step S21, the training image generation unit 121 displays the sample image 42 that it is about to process on the monitor. For example, the training image generation unit 121 may display a sample image 42 selected by the user, or it may display a sample image 42 input from the preprocessing unit 11.

[0037] In step S22, the training image generation unit 121 receives input for labels for the sample image 42. Labels are information treated as ground truth in machine learning. Labels for the sample image 42 indicate the coating region, for example, the region of each coating element. In one example, the training image generation unit 121 receives labels set based on user input.

[0038] In step S23, the training image generation unit 121 generates a training image 43 based on the sample image 42 and labels. For example, the training image generation unit 121 may generate the training image 43 by embedding labels in the sample image 42. The training image generation unit 121 stores the generated training image 43 in the training image database 32. Figure 6 shows the generation of a training image 220 based on the sample image 210. The training image 220 includes labels indicating the regions of each coating element that form the coating region of the object. Each label may be displayed using various representational methods such as color, pattern, or mark to distinguish each coating element. In the example shown as a training image 220, the fact that two or more coating elements have the same pattern does not mean that there is any relationship between the coating elements, but rather that the coating elements are identified individually.

[0039] The evaluation system 10 may execute the processing flow S2 multiple times or repeatedly. As the evaluation system 10 executes the processing flow S2 for each sample image 42, training images 43 are accumulated in the training image database 32.

[0040] (Generating a pre-trained model) The generation of a trained model will be explained with reference to Figure 7. Figure 7 is a flowchart showing an example of this process as processing flow S3. Processing flow S3 corresponds to the training phase.

[0041] In step S31, the model generation unit 122 obtains a training image 43 from the training image database 32.

[0042] In step S32, the model generation unit 122 performs learning based on the training images 43. For example, the model generation unit 122 performs learning using Mask R-CNN. In one example, the model generation unit 122 inputs the training images 43 into a machine learning model that includes a neural network and obtains the estimated result of the coating region output from the machine learning model. Based on the error between the estimated result and the labels of the training images 43, the model generation unit 122 updates the parameters in the machine learning model using a method such as backpropagation. For example, the model generation unit 122 updates the weights of the neural network.

[0043] In step S33, the model generation unit 122 determines whether or not to terminate machine learning. If the model generation unit 122 determines that the predetermined termination conditions are not met (NO in step S33), the process returns to step S31. In the iterative process, the model generation unit 122 obtains the next training image 43 in step S31 and performs learning based on that training image 43 in step S32. On the other hand, if the model generation unit 122 determines that the termination conditions are met (YES in step S33), the process proceeds to step S34. The termination conditions may be set based on the error, or based on the number of training images 43 to be processed, i.e., the number of learning iterations. Alternatively, the model generation unit 122 may evaluate the performance of the machine learning model using given validation data, and terminate machine learning if the evaluation meets a given criterion.

[0044] In step S34, the model generation unit 122 outputs the machine learning model after machine learning has been completed as the trained model 20. For example, the model generation unit 122 stores the trained model 20 in a predetermined storage device. This trained model 20 is then used by the evaluation unit 13.

[0045] (Evaluation of the covering of the object) The evaluation of the covering of the object will be explained with reference to Figures 8 and 9. Figure 8 is a flowchart showing an example of the process as processing flow S4. Figure 9 is a diagram showing an example of image processing related to processing flow S4. Processing flow S4 corresponds to the operation phase.

[0046] In step S41, the coating identification unit 131 acquires a target image. For example, the coating identification unit 131 may acquire a target image 45 selected by the user, or it may acquire a target image 45 input from the preprocessing unit 11.

[0047] In step S42, the coating identification unit 131 inputs the target image 45 into the trained model 20 to identify the coating region of the object indicated by the target image 45. The coating identification unit 131 obtains the estimation result from the trained model 20 to identify the coating region. For example, the coating identification unit 131 identifies multiple coating elements that form the coating region.

[0048] In step S43, the calculation unit 132 calculates evaluation values ​​for the identified coating area. The calculation unit 132 may also calculate evaluation values ​​for the coverage of the substrate by the identified coating area. An example of an evaluation value for coverage is the coverage rate, which indicates the percentage of the substrate covered by the coating area. The calculation unit 132 may calculate physical parameters such as area, height, and radius as evaluation values ​​for each of at least one coating element. The height of a coating element refers to the distance from the substrate surface to the top of the coating element. For coating elements located at the edge of an object, the calculation unit 132 may calculate the height based on the position on the substrate surface and the position of the top of the coating element. The calculation unit 132 may also calculate evaluation values ​​for shape, such as roundness, for each of at least one coating element. The calculation unit 132 may also calculate statistical values ​​for multiple coating elements, such as standard deviation, mean, and median, as evaluation values. For example, the calculation unit 132 can calculate statistical values ​​for various physical parameters such as area, roundness, height, and radius. If the coating element is a particle, the calculation unit 132 may calculate the coefficient of variation (CV) of the particle size as the evaluation value. The CV value is obtained by the following formula. The CV value is also an example of a statistical value. CV value = (standard deviation) / (median diameter)

[0049] As an example of calculating evaluation values, the calculation of coverage and related image processing will be explained with reference to Figure 9. Figure 9 shows the target image 250 and two auxiliary images 260 and 270. The target image 250 shows particulate matter as the target object, and the individual microparticles that are coating elements are represented by color or pattern. Auxiliary image 260 shows only the coating area, i.e., multiple microparticles. Auxiliary image 270 shows the method for calculating the coverage. Auxiliary image 270 includes a virtual circle 271 that shows the central region of the target object, which will be described later, and a virtual circle 272 that shows the outer edge of the substrate. Within the virtual circle 271, only multiple microparticles are represented, as in auxiliary image 260. Outside the virtual circle 271, particulate matter is represented in a form in which individual microparticles are identified, as in the target image 250. The virtual circle 272 may be omitted.

[0050] In one example, the calculation unit 132 sets a central region of the object to calculate the coverage rate. The central region is a part of the object shown by the image. For example, the calculation unit 132 sets the central region as an area within a predetermined radius from the center of the object. The calculation unit 132 calculates an evaluation value in the central region. For example, the calculation unit 132 calculates the total area of ​​the coating regions located within the central region, that is, the total area of ​​one or more coating elements located within the central region. The calculation unit 132 then calculates the coverage rate as the ratio of this total area to the area of ​​the central region. Each area can be specified by the number of pixels. Therefore, the calculation unit 132 may calculate the coverage rate as the ratio of the total number of pixels of the coating regions located within the central region to the number of pixels in the central region. The calculation unit 132 may calculate the area as an evaluation value for each of the coating elements located within the central region, or it may calculate an evaluation value related to the shape. The calculation unit 132 may calculate statistical values ​​related to multiple coating elements located within the central region as evaluation values. If the coating element is a particle, the calculation unit 132 may calculate the CV value within the central region as the evaluation value.

[0051] In step S44, the evaluation unit 13 outputs an evaluation value. The evaluation unit 13 may display the evaluation value on a monitor, store it in a predetermined storage device such as a database, or transmit it to another computer.

[0052] The evaluation unit 13 may display at least one of the target image 250 and auxiliary images 260 and 270. That is, the evaluation unit 13 may display at least an image showing the identified coating area. Through this image, the user can confirm the identification of the coating area and the basis for the evaluation value.

[0053] The evaluation system 10 may execute the processing flow S4 multiple times or repeatedly. For example, each time the user selects a target image 45, the evaluation system 10 executes the processing flow S4 in response to that selection. The evaluation unit 13 may calculate a statistical value of multiple evaluation values ​​corresponding to multiple objects as a further evaluation value.

[0054] [Differentiation] The technology relating to this disclosure has been described in detail above based on various examples. However, this disclosure is not limited to the embodiments described above. The technology relating to this disclosure can be modified in various ways without departing from its essence.

[0055] The evaluation system does not necessarily have to include either a learning unit or an evaluation unit. The trained model is portable between computer systems. Therefore, the learning unit may provide the trained model to other computer systems, and the evaluation unit may use the trained model provided by other computers.

[0056] The original images may be provided from devices other than the database, for example, directly from imaging devices such as SEMs and cameras.

[0057] The preprocessing unit may be used only in either the learning phase or the operation phase. Alternatively, the evaluation system may not include a preprocessing unit. The learning unit may acquire sample images from another computer system, and the evaluation unit may acquire target images from another computer system. Training images may be generated by another computer system, and therefore, the evaluation system may not include a training image generation unit.

[0058] At least one of the sample image, training image, and target image may be an image depicting multiple objects. Therefore, the learning unit may generate a trained model based on a sample image depicting multiple objects or a training image depicting multiple objects. This trained model is a computational model trained to identify the coating region for at least one of the one or more objects depicted in the target image. The evaluation unit may input a target image depicting one or more objects into the trained model and identify the coating region for at least one of the one or more objects. In this example, the evaluation unit calculates an evaluation value for the identified coating region for at least one object.

[0059] The processing steps of a method executed by at least one processor are not limited to the examples in the above embodiments. For example, some of the steps described above may be omitted, or each step may be performed in a different order. Also, any two or more of the steps described above may be combined, or some of the steps may be modified or deleted. Alternatively, other steps may be performed in addition to each of the above steps.

[0060] In comparing the relative magnitudes of two numerical values ​​in this disclosure, either of the two criteria, "greater than or equal to" and "greater than," may be used, or either of the two criteria, "less than or equal to" and "less than," may be used.

[0061] In this disclosure, the expression "at least one processor executes a first process, a second process, ... and the nth process," or a corresponding expression, refers to a concept that includes cases where the entity executing the n processes from the first process to the nth process changes midway through. In other words, this expression refers to a concept that includes both cases where all n processes are executed by the same processor and cases where the processor changes at an arbitrary rate for the n processes.

[0062] [Note] As can be seen from the various examples above, this disclosure includes the following aspects: (Note 1) Equipped with at least one processor, The aforementioned at least one processor, An image of an object having a substrate and a coating area on the substrate is obtained. The target image is input to a trained model that estimates the coating region from the input image to identify the coating region of the object. The evaluation value for the identified coating area is calculated. Evaluation system. (Note 2) The at least one processor calculates the evaluation value relating to the coating of the substrate by the identified coating region. The evaluation system described in Appendix 1. (Note 3) The at least one processor calculates a coating rate as the evaluation value, which indicates the proportion of the substrate that is covered by the identified coating area. The evaluation system described in Appendix 2. (Note 4) The aforementioned at least one processor, The target image is input to the trained model to identify a plurality of coating elements that form the coating region. The statistical values ​​relating to the plurality of coating elements are calculated as the evaluation values. The evaluation system described in Appendix 1. (Note 5) The aforementioned at least one processor, The central region of the object shown by the aforementioned target image is set, The evaluation value in the central region is calculated. The evaluation system described in any one of the appendices 1-4. (Note 6) The at least one processor sets the region within a predetermined radius from the center of the object as the central region. The evaluation system described in Appendix 5. (Note 7) The at least one processor displays an image showing at least the identified coating region. The evaluation system described in any one of the appendices 1 to 6. (Note 8) The aforementioned at least one processor, A source image showing at least one of the aforementioned objects is obtained, The original image is converted into a binary image, A distance transformation is performed on the binary image to determine the center and dimensions of each of the at least one object, Based on the center and dimensions of each of the at least one object, an object image is generated showing the entire object in the original image. The evaluation system described in any one of the appendices 1-7. (Note 9) The aforementioned object is particulate matter. The evaluation system described in any one of the appendices 1 to 8. (Note 10) The substrate is a core particle, The coating region is an aggregate of multiple fine particles. The evaluation system described in Appendix 9. (Note 11) An evaluation method performed by an evaluation system comprising at least one processor, A step of acquiring an object image showing an object having a substrate and a coating area on the substrate, The steps include: inputting the target image into a trained model that estimates the coating region from an input image to identify the coating region of the object; A step of calculating an evaluation value for the identified coating area, An evaluation method that includes this. (Note 12) A step of acquiring an object image showing an object having a substrate and a coating area on the substrate, The steps include: inputting the target image into a trained model that estimates the coating region from an input image to identify the coating region of the object; A step of calculating an evaluation value for the identified coating area, An evaluation program that causes a computer to execute a command.

[0063] According to appendices 1, 11, and 12, the pre-trained model estimates the coating region of an object from the target image. This configuration allows for easy identification of the coating region and simplifies the calculation of evaluation values ​​for that coating region. Therefore, evaluation of the object's coating can be easily performed.

[0064] According to Appendix 2, the evaluation value regarding the coating of the substrate by the coating area can be easily calculated.

[0065] According to Appendix 3, the coverage rate due to the coating area can be easily calculated.

[0066] According to Appendix 4, the trained model makes it easy to identify multiple coating elements, and also facilitates the calculation of statistical values ​​related to those coating elements.

[0067] According to Appendix 5, the evaluation value is calculated for the central region of the object. Depending on the three-dimensional shape of the object, the object may appear in the image as if the coating area is concentrated at its periphery. Therefore, it may be difficult to accurately calculate the evaluation value for the coating area. In contrast, the central region of the object appears in the image as if it were in its actual state. Therefore, by calculating the evaluation value for that central region, the evaluation of the object's coating can be performed more accurately.

[0068] According to Appendix 6, a circular central region is set, so the central region can be appropriately set regardless of the external shape of the object.

[0069] According to Appendix 7, the identified coating area is displayed, allowing the user to see the processing results from the trained model. The user can then refer to these results and gain a sense of satisfaction with the evaluation of the object's coating.

[0070] According to Appendix 8, the center and dimensions of individual objects are identified through conversion to a binary image and distance transformation. Based on this identification, an object image is obtained that shows the entire object in the original image. This method allows for the automatic acquisition of an appropriate object image for evaluating the coverage of an object.

[0071] According to Appendix 9, evaluation of the coating of particulate matter can be easily performed.

[0072] According to Appendix 10, the trained model makes it easy to identify clusters of multiple fine particles, and also facilitates the calculation of evaluation values ​​for those clusters. Therefore, evaluation of coatings can be easily performed for particulate matter containing fine particles as components of the coating region. [Explanation of Symbols]

[0073] 10...Evaluation system, 11...Preprocessing unit, 12...Learning unit, 13...Evaluation unit, 20...Trained model, 31...First original image database, 32...Teacher image database, 33...Second original image database, 41...First original image, 42...Sample image, 43...Teacher image, 44...Second original image, 45...Target image, 121...Teacher image generation unit, 122...Model generation unit, 131...Coating identification unit, 132...Calculation unit.

Claims

1. Equipped with at least one processor, The at least one processor, An image of an object having a substrate and a coating area on the substrate is obtained. The target image is input to a trained model that estimates the coating region from the input image to identify the coating region of the object. The central region of the object shown by the aforementioned target image is set, The evaluation value for the identified coating area in the central region is calculated. Evaluation system.

2. The at least one processor calculates the evaluation value relating to the coating of the substrate by the identified coating region. The evaluation system according to claim 1.

3. The at least one processor calculates a coating rate as the evaluation value, which indicates the proportion of the substrate that is covered by the identified coating area. The evaluation system according to claim 2.

4. The at least one processor, The target image is input to the trained model to identify a plurality of coating elements that form the coating region. The statistical values ​​relating to the plurality of coating elements are calculated as the evaluation values. The evaluation system according to claim 1.

5. The at least one processor sets the region within a predetermined radius from the center of the object as the central region. The evaluation system according to claim 1.

6. The at least one processor displays an image showing at least the identified coating region. The evaluation system according to any one of claims 1 to 5.

7. The at least one processor, A source image showing at least one of the aforementioned objects is obtained, The original image is converted into a binary image, A distance transformation is performed on the binary image to determine the center and dimensions of each of the at least one object, Based on the center and dimensions of each of the at least one object, an object image is generated showing the entire object in the original image. The evaluation system according to any one of claims 1 to 5.

8. The aforementioned object is particulate matter. The evaluation system according to any one of claims 1 to 5.

9. The substrate is a core particle, The coating region is an aggregate of multiple fine particles. The evaluation system according to claim 8.

10. An evaluation method performed by an evaluation system comprising at least one processor, A step of acquiring an object image showing an object having a substrate and a coating area on the substrate, The steps include: inputting the target image into a trained model that estimates the coating region from an input image to identify the coating region of the object; The steps include setting the central region of the object shown in the aforementioned target image, A step of calculating an evaluation value for the identified coating area in the central region, An evaluation method that includes this.

11. A step of acquiring an object image showing an object having a substrate and a coating area on the substrate, The steps include: inputting the target image into a trained model that estimates the coating region from an input image to identify the coating region of the object; The steps include setting the central region of the object shown in the aforementioned target image, A step of calculating an evaluation value for the identified coating area in the central region, An evaluation program that causes a computer to execute a command.

Images