Analysis device, analysis system, analysis method, and program

Abstract

The objective of the present invention is to accurately calculate an amount of particulate matter without being affected by changes in the amount of light emitted onto the particulate matter. An analysis device (100) comprises: a holding member (1) that holds particulate matter (FP); a light source (3) that emits light toward an imaging target area (VA) including a holding area of the holding member (1) in which the particulate matter (FP) is held; a two-dimensional sensor (4) that images the imaging target area (VA); and a calculating unit (5) that calculates particle amount-related information relating to the amount of particulate matter (FP) on the basis of analysis image data acquired by employing the two-dimensional sensor (4) to image the imaging target area (VA) for use in analyzing the particulate matter (FP), and an impact resulting from a change in the amount of light from the light source (3) when the analysis image data were acquired, relative to a reference amount of light.

Description

Analytical device, analytical system, analytical method, and program

[0001] The present invention relates to an analytical device for analyzing particles, an analytical system, a particle analysis method, and a program for causing a computer to execute the method.

[0002] In recent years, particulate matter (e.g., PM2.5) contained in the atmosphere has become a major environmental problem. Methods and devices for analyzing particulate matter have been developed for the purpose of understanding the properties and sources of particulate matter.

[0003] For example, it is known to analyze the amount of particulate matter (collection amount, mass concentration, etc.) trapped by a collection filter. One method for analyzing the amount of particulate matter involves acquiring an image of colored particulate matter trapped by a collection filter and calculating the amount of colored particulate matter based on the brightness of pixels included in the acquired image (see, for example, Patent Document 1).

[0004] Japanese Patent Application Laid-Open No. 2019-20411

[0005] When acquiring the image, light is irradiated from a light source onto the area of ​​the collection filter where particulate matter is trapped. In this case, if the amount of light irradiated from the light source changes due to various factors (e.g., temperature, aging, etc.), images with different brightness may be acquired even if the amount of particulate matter trapped on the collection filter is the same. As a result, when the amount of particulate matter is calculated based on the image, the calculated amount may differ from the actual amount of particulate matter.

[0006] The object of the present invention is to accurately calculate the amount of particulate matter when calculating the amount of particulate matter based on an image of the particulate matter, without being affected by changes in the amount of light irradiated onto the particulate matter.

[0007] Below, several aspects will be described as means for solving the problems. These aspects can be combined as desired. An analysis device according to one aspect of the present invention includes a holding member, a light source, a two-dimensional sensor, and a calculation unit. The holding member holds particles. The light source irradiates light onto a target area of ​​the holding member, including a holding area where the particles are held. The two-dimensional sensor photographs the target area. The calculation unit calculates particle amount-related information regarding the amount of particles held in the holding member based on analysis image data acquired by photographing the target area with the two-dimensional sensor for use in particle analysis, and on the effect of changes in the light source light intensity when the analysis image data was acquired relative to a reference light intensity.

[0008] In the above-described analysis device, analytical image data of a photographed area including a holding area where particles are held is acquired for use in particle analysis, and information relating to the amount of particles (particle amount-related information) is calculated based on the acquired analytical image data and the effect of changes in the light intensity of the light source at the time the analytical image data was acquired relative to a reference light intensity. This allows accurate particle amount-related information to be calculated that is not affected by changes in the light intensity of the light source.

[0009] In the above-described analysis device, the influence may be a nonlinear change in pixel luminance of the analysis image data with respect to a change in the light intensity of the light source. This makes it possible to calculate accurate particle amount-related information that is not affected by changes in the light intensity of the light source, even when analysis pixel data having a complex characteristic in which pixel luminance changes nonlinearly with respect to a change in the light intensity of the light source is used.

[0010] In the above-described analyzing device, the calculation unit may correct the luminance of the analytical image data based on a change in the color tone of the photographed area illuminated by the light from the light source when the light intensity of the light source changes relative to a reference light intensity, and calculate the particle amount-related information based on the corrected analytical image data. This allows the luminance of the analytical image data to be corrected more accurately, thereby allowing the particle amount-related information to be calculated more accurately based on the corrected analytical image data.

[0011] In the above-described analysis device, the calculation unit may calculate the particle amount-related information based on first corrected image data in which the luminance of analytical image data acquired by capturing an image of the target area using a two-dimensional sensor when the light source light intensity is a first light intensity is corrected to the luminance when the light source light intensity is a second light intensity, which is a reference light intensity different from the first light intensity, thereby eliminating the above-described influence. In this way, even if the analytical image data is acquired when the light source light intensity has changed from the reference light intensity (i.e., the second light intensity), the luminance of the analytical image data can be corrected to generate image data (i.e., first corrected image data) assuming that the light source light intensity is the reference light intensity. Then, by calculating the particle amount-related information based on the first corrected image data, accurate particle amount-related information that is not affected by changes in light intensity can be calculated.

[0012] In the analysis device, the calculation unit may generate the first corrected image data by reducing the luminance correction range of pixels having low luminance in the analysis image data and increasing the luminance correction range of pixels having high luminance, thereby allowing the analysis image data to be corrected more accurately by reflecting the tendency of the luminance change range in response to changes in the light intensity of the light source.

[0013] In the above-described analysis device, the calculation unit may generate converted data by converting the luminance of each pixel of the analysis image data to a hue based on a first relational expression expressing the relationship between the luminance of a pixel of the first correction image data acquired by capturing an image of the target area with the two-dimensional sensor when the light source has a first light intensity and the hue of a position in the target area corresponding to that pixel, and may generate the first corrected image data by converting the hue of each pixel of the converted data to the luminance when the light source has a second light intensity based on a second relational expression expressing the relationship between the luminance of a pixel of the second correction image data acquired by capturing an image of the target area with the two-dimensional sensor when the light source has a second light intensity. This allows for more accurate correction of the luminance of the analysis image data.

[0014] In the above-described analysis device, the first relational expression and the second relational expression may respectively represent the luminance of pixels in the first correction image data and the second correction image data as polynomials of second order or higher of the hue of the portion of the subject area corresponding to the pixel. This allows the first relational expression and the second relational expression to more accurately represent the relationship between the luminance and hue of the pixel of the analysis image data. As a result, the luminance of the analysis image data can be corrected more accurately.

[0015] In the above-described analyzing device, the calculation unit may calculate pre-correction particle amount-related information based on the analytical image data, and correct the pre-correction particle amount-related information based on the amount of change when the light amount of the light source changes from a reference light amount to the light amount when the analytical image data was acquired. In this way, even if the analytical image data is acquired when the light amount of the light source has changed from the reference light amount and the pre-correction particle amount-related information is calculated based on the analytical image data, accurate particle amount-related information that is not affected by the change in the light amount of the light source can be calculated.

[0016] In the above-described analysis device, the calculation unit may calculate the particle amount-related information using a trained model that has learned the influence of changes in the light intensity of the light source relative to a reference light intensity, thereby enabling calculation of accurate particle amount-related information that is not affected by changes in the light intensity of the light source.

[0017] In the above-described analysis device, the calculation unit may input the analytical image data to a trained model that has learned the above-described influences using training data, and the trained model receives image data acquired by capturing an image of the target area using a two-dimensional sensor and outputs information regarding the amount of particles held in the holding member when the image data was captured, thereby outputting the particle amount-related information from the trained model. This allows accurate particle amount-related information that is not affected by changes in the light intensity of the light source to be calculated using the analytical image data as is.

[0018] In the above-described analysis device, the calculation unit may input the analysis image data to a trained model that has learned the above-described influence using training data, and the trained model takes image data acquired by photographing the target area with a two-dimensional sensor as input and generates corrected image data by correcting the luminance of the image data to the luminance when the light source has a reference light intensity as output. The calculation unit then outputs second corrected image data from the trained model, in which the luminance of the analysis image data has been corrected to the luminance when the light source has a reference light intensity, and calculates particle quantity-related information based on the second corrected image data.

[0019] This allows the second corrected image data, which is not affected by changes in the light intensity of the light source, to be generated directly from the analytical image data. Furthermore, by calculating particle amount-related information based on this second corrected image data, accurate particle amount-related information, which is not affected by changes in the light intensity of the light source, can be calculated.

[0020] In the above-described analysis device, the imaging target area may include the above-described retention area and a non-retention area where no particles are retained. In this case, the calculation unit may input the third representative luminance and the fourth representative luminance of the analysis image data to a trained model that has learned the above-described influence using teacher data that receives the first representative luminance and the second representative luminance as input and outputs the first corrected representative luminance and the second corrected representative luminance, thereby outputting the third corrected representative luminance and the fourth corrected representative luminance from the trained model, and calculate the particle amount-related information based on the third corrected representative luminance and the fourth corrected representative luminance.

[0021] The first representative luminance is the representative luminance of pixels corresponding to the non-retention area of ​​image data acquired by capturing an image of the target area with a two-dimensional sensor. The second representative luminance is the representative luminance of pixels corresponding to the retention area of ​​this image data. The first corrected representative luminance is the representative luminance of pixels corresponding to the non-retention area when the luminance of this image data is corrected to the luminance when the light intensity of the light source is a reference light intensity. The second corrected representative luminance is the representative luminance of pixels corresponding to the retention area when the luminance of the image data is corrected to the luminance when the light intensity of the light source is a reference light intensity. The third representative luminance is the representative luminance of pixels corresponding to the non-retention area of ​​the analysis image data. The fourth representative luminance is the representative luminance of pixels corresponding to the retention area of ​​the analysis image data. The third corrected representative luminance is the representative luminance of pixels corresponding to the non-retention area when the luminance of the analysis image data is corrected to the luminance when the light intensity of the light source is a reference light intensity. The fourth corrected representative luminance is the representative luminance of pixels corresponding to the retention area when the luminance of the analysis image data is corrected to the luminance when the light intensity of the light source is a reference light intensity.

[0022] This makes it possible to train a trained model using a small amount of information to calculate accurate particle quantity-related information that is not affected by changes in the light intensity of the light source.

[0023] An analysis system according to another aspect of the present invention includes an analysis device and a correction device. The analysis device includes a holding member, a light source, a two-dimensional sensor, and a calculation unit. The holding member holds particles. The light source irradiates a target area of ​​the holding member, including a holding region where the particles are held, with light. The two-dimensional sensor photographs the target area. The calculation unit calculates particle amount-related information regarding the amount of particles held in the holding member based on analysis image data acquired by photographing the target area with the two-dimensional sensor for use in particle analysis, and on the effect of changes in the light intensity of the light source when the analysis image data was acquired relative to a reference light intensity. The correction device generates information regarding the effect of changes in the light intensity of the light source when the analysis image data was acquired relative to a reference light intensity, and transmits the information to the analysis device.

[0024] In the above analysis system, the analysis device acquires analytical image data of the imaging target area, including the holding area where particles are held, for use in particle analysis, and calculates information related to the amount of particles (particle amount-related information) based on the acquired analytical image data and the effect of changes in the light intensity of the light source at the time the analytical image data was acquired relative to a reference light intensity. This allows accurate particle amount-related information to be calculated that is not affected by changes in the light intensity of the light source.

[0025] An analysis method according to a further aspect of the present invention comprises the following steps: a step of irradiating a target area of ​​a holding member that holds particles with light from a light source, the target area including a holding region where particles are held; a step of photographing the target area and acquiring analytical image data to be used in particle analysis; and a step of calculating particle amount-related information regarding the amount of particles held in the holding member based on the analytical image data and the effect of changes in the light amount of the light source when the analytical image data was acquired relative to a reference light amount.

[0026] In the above analysis method, analytical image data of a photographed area including a holding area where particles are held is acquired for use in particle analysis, and information on the amount of particles (particle amount-related information) is calculated based on the acquired analytical image data and the effect of changes in the light intensity of the light source at the time the analytical image data was acquired relative to a reference light intensity. This allows accurate particle amount-related information to be calculated that is not affected by changes in the light intensity of the light source.

[0027] A program according to a further aspect of the present invention is a program for causing a computer to execute the above-described analysis method.

[0028] When acquiring analytical image data used for particle analysis, accurate particle amount-related information can be calculated that is not affected by changes in the amount of light from the light source that irradiates the imaging target area with light.

[0029] A diagram showing the configuration of an analysis device. A diagram showing an example of image data of a photographing target area. A diagram showing the configuration of a calculation unit. A diagram showing an example of light intensity-luminance characteristics. A diagram showing an example of hue-luminance characteristics. A diagram showing a schematic diagram of a method for calculating hue-luminance characteristics for a specific light intensity. A flowchart showing the analysis operation of particulate matter in the first embodiment. A flowchart showing a method for correcting the luminance of image data for analysis. A diagram showing an example of a luminance histogram. A flowchart showing a method for analyzing particulate matter in the second embodiment. A diagram showing the configuration of an analysis system of the third embodiment.

[0030] 1. First Embodiment (1) Overview of the Analysis Device The analysis device 100 performs analysis on the amount of particles (referred to as particulate matter FP) held on the holding member 1. The analysis on the amount of particles is performed based on image data acquired from an area (referred to as an imaging target area VA) including an area (referred to as a holding area) where the particulate matter FP is held on the holding member 1.

[0031] The analysis target of the analysis device 100 is particulate matter (FP) contained in a sample gas Gs. The sample gas Gs is a gas containing particulate matter (FP), such as the atmosphere or gases generated in various combustion processes. Examples of various combustion processes include combustion processes in thermal power plants, steel plants, incinerators, and coal combustion processes. In such combustion processes, for example, unburned matter in ash, fly ash, and the like are generated as particulate matter (FP).

[0032] Other examples of particulate matter FP to be analyzed include dust generated from various transportation devices (such as automobiles or ships) (dust from brakes, tires, internal combustion engines, steam engines, or exhaust gas purification devices and motors), dust generated by natural disasters such as volcanic eruptions (e.g., volcanic ash), and dust generated during mining development.

[0033] (2) Configuration of the Analysis Device The configuration of the analysis device 100 will be described below with reference to Fig. 1. Fig. 1 is a diagram showing the configuration of the analysis device. The analysis device 100 includes a holding member 1, a collection unit 2, a light source 3, a two-dimensional sensor 4, and a calculation unit 5.

[0034] The holding member 1 is a member that captures particulate matter FP contained in the sample gas Gs and retains the particulate matter FP. The holding member 1 is a collection filter formed by laminating a collection layer made of a porous fluororesin material having pores capable of capturing particulate matter FP on a reinforcing layer made of a nonwoven fabric of a polymer material (such as polyethylene). With the above configuration, the holding member 1 allows gas to flow through the holding member 1 in the thickness direction, while at the same time improving its strength. Furthermore, the holding member 1 can be made less likely to become charged. As the holding member 1, other filters such as a single-layer glass filter or a single-layer fluororesin material filter can also be used.

[0035] The holding member 1 has a white color. When particulate matter FP is held by the holding member 1, the area where the particulate matter FP is held changes to the color of the particulate matter FP. This color becomes darker as the amount of particulate matter FP held increases.

[0036] The analysis device 100 includes a moving unit 11. The moving unit 11 moves the holding member 1 in the longitudinal direction. The moving unit 11 has a take-up reel 11a and a supply reel 11b. The take-up reel 11a is connected to one end of the holding member 1 in the longitudinal direction. The take-up reel 11a is rotatable in a predetermined direction by, for example, a motor. The supply reel 11b is connected to the other end of the holding member 1. The supply reel 11b rotates in accordance with the movement of the holding member 1.

[0037] With this configuration, the holding member 1 can be fed from the feed reel 11b and wound onto the take-up reel 11a by rotating the take-up reel 11a in the moving unit 11. In other words, the holding member 1 can be moved in the length direction of the holding member 1 (the direction indicated by the thick arrow in FIG. 1) by rotating the take-up reel 11a.

[0038] The mechanism for moving the holding member 1 is not limited to the above mechanism. For example, a pin may be provided in the width direction of the holding member 1, and the direction of movement of the holding member 1 may be changed before and after passing through the pin. Also, a tension controller may be provided in the holding member 1 to adjust the tension.

[0039] The collection unit 2 is provided to correspond to a first position P1 in the longitudinal direction of the holding member 1. The collection unit 2 has a suction pump 21, an outlet 23, and an intake port 25. The collection unit 2 discharges the sample gas Gs from the outlet 23 toward the first position P1 of the holding member 1 by generating a suction force at the intake port 25 using the suction pump 21, and causes the particulate matter FP contained in the sample gas Gs to be collected by the holding member 1. As such, the first position P1 is the position in the longitudinal direction of the holding member 1 where the particulate matter FP is collected, and therefore can also be referred to as the "collection position." Furthermore, the area on the surface of the holding member 1 where the particulate matter FP collected by the collection unit 2 is held is referred to as the "holding area."

[0040] The light source 3 emits light onto the imaging area VA, from which image data can be captured by the two-dimensional sensor 4. The light source 3 is a surface-mounted white LED. Multiple light sources 3 are arranged to uniformly illuminate the imaging area VA. As shown in FIG. 1 , a view of the light source 3 from the holding member 1 side (the view surrounded by the dashed line in FIG. 1 ), the multiple light sources 3 are arranged at predetermined intervals on the substrate on an ellipse whose major axis is parallel to the longitudinal direction of the holding member 1. This allows light to be uniformly illuminated over a wide area in the longitudinal direction of the holding member 1. The size of the major and / or minor axes of the ellipse on which the multiple light sources 3 are arranged can be determined appropriately depending on the ratio of the length and width of the imaging area VA. Furthermore, the multiple light sources 3 can also be arranged on the circumference of any shape other than an ellipse, depending on the shape of the imaging area VA.

[0041] By making the light source 3 a surface-mounted type, the distance between the light source 3 and the surface of the holding member 1 can be made as large as possible, and light of an appropriate intensity can be uniformly irradiated over a wide area of ​​the surface of the holding member 1 without having to place a light-dispersing component such as a diffuser immediately in front of the light source 3.

[0042] The substrate on which the plurality of light sources 3 are arranged is provided with an opening that is smaller than the ellipse formed by the plurality of light sources 3. The two-dimensional sensor 4 is provided so that the light receiving surface protrudes from the opening.

[0043] By providing the multiple light sources 3 with the above configuration, it is possible to brighten the image capture area VA and obtain clearer image data. Furthermore, by irradiating the image capture area VA with a uniform amount of light, it is possible to avoid obtaining image data that is affected by the distribution of the amount of light emitted from the light sources 3.

[0044] The two-dimensional sensor 4 is, for example, a CCD image sensor in which charge-coupled devices (CCDs) are arranged in a two-dimensional array, or a CMOS image sensor in which light-receiving elements are arranged in an array. The two-dimensional sensor 4 may also include an optical element such as a wide-angle lens, so that it can acquire image data of a wide area of ​​the holding member 1.

[0045] The two-dimensional sensor 4 is disposed at a predetermined height from the surface of the holding member 1. The two-dimensional sensor 4 is also provided at a second position P2 in the longitudinal direction of the holding member 1 so that its light-receiving surface faces the surface of the holding member 1 on the side where the particulate matter FP is collected. The second position P2 can also be called the "imaging position" because it is the position to which the holding area moves in order for the two-dimensional sensor 4 to capture image data including the holding area where the particulate matter FP is held on the holding member 1. The second position P2 is, for example, midway between the first position P1 and a third position P3 where elemental analysis is performed.

[0046] By arranging the two-dimensional sensor 4 as described above, the two-dimensional sensor 4 can acquire image data of an imaging target area VA including a second position P2 and a third position P3, as shown in Fig. 2. Fig. 2 is a diagram showing an example of image data of an imaging target area. Note that the imaging target area VA also includes images of an elemental analysis unit 7 (described below), such as an X-ray source 71, which is present near the third position P3; however, for ease of understanding, the images of the elemental analysis unit 7 (X-ray source 71, etc.) are omitted from Fig. 2.

[0047] As described above, by including in the imaging target area VA not only the second position P2 where the holding area is located but also the third position P3 where elemental analysis is performed, the state of the holding member 1 located at the third position P3 can also be monitored. For example, before the start of elemental analysis, image data of the imaging target area VA can be displayed on the display 55 of the calculation unit 5 to check (monitor) whether there is a foreign object in the holding member 1 located at the third position P3 (i.e., the portion of the holding member 1 that is irradiated with X-rays), whether there is an abnormality in the elemental analysis unit 7, or other such error, and an error can also be reported, for example.

[0048] The imaging target area VA can also be expanded by increasing the number of pixels of the two-dimensional sensor 4 and / or by raising the position of the light receiving surface of the two-dimensional sensor 4 as high as possible. The imaging target area VA may also be expanded so that it includes the first position P1 (collection position).

[0049] The two-dimensional sensor 4 acquires image data of the imaging target area VA when the holding area of ​​the particulate matter FP moves to the second position P2. Thereafter, as shown in Fig. 2, partial image data IM to be used for analyzing the particulate matter FP is cut out from the image data of the imaging target area VA acquired by the two-dimensional sensor 4 as two-dimensional image data including the holding area. Alternatively, during analysis, image data of the second position P2 and a predetermined area therearound may be used as the partial image data IM from the image data of the imaging target area VA.

[0050] 2, the partial image data IM includes first region image data Im1 and second region image data Im2. The first region image data Im1 is image data corresponding to a retention region where particulate matter FP is retained. The second region image data Im2 is image data corresponding to a region (referred to as a non-retention region) that exists around the retention region and is included in the imaging target region VA and where particulate matter FP is not retained.

[0051] Furthermore, the two-dimensional sensor 4 may be capable of acquiring the partial image data IM in, for example, about 1 / 60 seconds (corresponding to the display length of one frame of a video), which allows the two-dimensional sensor 4 to simultaneously and quickly acquire the first region image data Im1 corresponding to the retention region where the particulate matter FP is retained and the second region image data Im2 corresponding to the non-retention region where the particulate matter FP is not retained.

[0052] The analysis device 100 may be provided with a plate-like member 41. The plate-like member 41 is a member having reflective properties that provide background compensation for image data. Specifically, the plate-like member 41 preferably has surface reflective properties that are substantially the same as those of the holding member 1. Here, "surface reflective properties" refers to comprehensive reflective properties that include reflectance for specularly reflected light and reflectance that includes the effects of diffuse reflection and the like. The thickness of the plate-like member 41 can be any thickness, such as a thick plate or a thin film, as long as it does not allow excessive light to pass through.

[0053] When two-dimensional image data of the holding area and / or non-holding area is acquired by the two-dimensional sensor 4, the plate-like member 41 is disposed on the opposite side of the surface of the holding member 1 that faces the two-dimensional sensor 4 in the imaging target area VA. By disposing the plate-like member 41, the pixels included in the partial image data IM can be made to stand out more.

[0054] Furthermore, for example, when acquiring color image data using a two-dimensional sensor 4, by making the surface reflection characteristics of the plate-like member 41 and the surface reflection characteristics of the holding member 1 identical, the color image data can be converted into image data with an accurate brightness distribution.

[0055] As described above, the plate-shaped member 41 can prevent an object present on the back side of the holding member 1 from becoming the background of the partial image data IM. Furthermore, the plate-shaped member 41 can equalize the amount of light in the imaging target area VA by making the surface reflection characteristics of the holding member 1 the same. Therefore, the plate-shaped member 41 can also be called a "background compensation unit" that equalizes the amount of light from the background of the partial image data IM.

[0056] 3, the calculation unit 5 is a computer system having, for example, a CPU 51, a memory unit 53 which is part or all of the memory area of ​​a storage device such as RAM or ROM, a display 55 (e.g., a liquid crystal display, an organic EL display, etc.), and various interfaces 57 (e.g., an I / O port, a communication interface, etc.) which input and output data and signals and convert signals between each part of the analysis device 100. Fig. 3 is a diagram showing the configuration of the calculation unit.

[0057] The CPU 51 executes various controls and information processing of the analysis device 100. Specifically, the CPU 51 analyzes the particulate matter FP using the partial image data IM. More specifically, the CPU 51 calculates particle amount-related information relating to the amount of the particulate matter FP based on the partial image data IM (referred to as analytical image data) acquired for the analysis of the particulate matter FP. The particle amount-related information can be, for example, mass concentration.

[0058] Furthermore, the CPU 51 calculates the particle amount-related information taking into account the influence of changes in the light intensity of the light source 3 relative to a reference light intensity when the analytical image data was acquired. In this embodiment, the CPU 51 takes into account the influence of changes in the light intensity of the light source 3 relative to a reference light intensity when the analytical image data was acquired, correcting the luminance of the acquired analytical image data to the luminance assumed when the analytical image data was acquired when the light intensity of the light source 3 was a reference light intensity, thereby generating first corrected image data from which the above influence has been removed. The CPU 51 analyzes the particulate matter FP using this first corrected image data. As a result, even if the light intensity output from the light source 3 changes due to aging or the like, the analysis results of the particulate matter FP will not be affected.

[0059] The CPU 51 executes various controls and information processing according to programs stored in the storage unit 53 and executable by the CPU 51. Note that some of the various controls and information processing in the analysis device 100 may be realized by hardware.

[0060] The storage unit 53 stores various settings of the analysis device 100 and information (e.g., parameters) required for control and information processing of the analysis device 100. The storage unit 53 stores information related to the influence of changes in the light intensity of the light source 3 when the analysis image data was acquired relative to a reference light intensity. In this embodiment, the information stored is information for correcting the luminance of the analysis image data. Specifically, the storage unit 53 stores light intensity-luminance characteristic association information I1 and color tone-luminance characteristic association information I2. The storage unit 53 may also store the reference light intensity.

[0061] The light intensity-brightness characteristic association information I1 is information relating to the light intensity-brightness characteristic. The light intensity-brightness characteristic represents the relationship between the amount of light irradiated onto a component having a specific color and the brightness (median) of the pixels included in the partial image data IM when the component is irradiated with different amounts of light and the partial image data IM is acquired.

[0062] The above "color tone" represents how the holding member 1 appears to the two-dimensional sensor 4 (color intensity). In other words, the "color tone" corresponds to the amount of particulate matter FP held by the holding member 1. The "color tone" can be, for example, light transmittance or reflectance. For example, if the "color tone" is light transmittance, and the particulate matter FP is black (or a color close to that), it indicates the degree of blackness of the holding member 1. For example, the color tone (light transmittance) of a holding member 1 that is truly black is 0%. On the other hand, the color tone (light transmittance) of a holding member 1 that is white (the original color of the holding member 1) is 100%.

[0063] In the light amount-brightness characteristics, the "light amount" is defined as the brightness (median) of image data obtained by irradiating light onto the white holding member 1. By defining the "light amount" as above, the light amount can be determined from the image data, eliminating the need to directly measure the light amount of the light source 3.

[0064] The light amount-brightness characteristic has a characteristic in which "brightness" changes nonlinearly with respect to "light amount", as shown in FIG. 4, for example. FIG. 4 is a diagram showing an example of the light amount-brightness characteristic. For example, when the light amount is x and the brightness is y, the light amount-brightness characteristic of the holding member 1 having a specific color is expressed as y=a １ *x２ +a ２ *x+a ３ (a １ , a ２ , a ３ The light intensity-brightness characteristic can be expressed as a quadratic equation with x (x: constant). The light intensity-brightness characteristic may also be expressed as another nonlinear function of the light intensity (x) (for example, a polynomial of second order or higher). The light intensity-brightness characteristic shown in FIG. 4 is a characteristic when the color tone is expressed as transmittance.

[0065] As shown in Fig. 4, the light intensity-brightness characteristic differs depending on the hue of the holding member 1. Therefore, the light intensity-brightness characteristic is calculated for each of the multiple hue of the holding member 1. As the hue of the holding member 1 differs, the coefficient a １ , a ２ , a ３ is different.

[0066] The light quantity-brightness characteristic relation information I1 may store the above quadratic equation calculated for each different color tone, or may store the coefficients a of the quadratic equation calculated for each different color tone. １ , a ２ , a ３ The information storage device may store the following information.

[0067] The hue-brightness characteristic association information I2 is information relating to the hue-brightness characteristic. The hue-brightness characteristic represents the relationship between the brightness of each pixel of the partial image data IM and the hue of the position in the photographed area VA corresponding to each pixel of the partial image data IM when the partial image data IM is acquired by irradiating the holding member 1 with a certain amount of light.

[0068] The hue-brightness characteristic has a characteristic in which "brightness" changes nonlinearly with respect to "hue," as shown in Fig. 5, for example. Fig. 5 is a diagram showing an example of the hue-brightness characteristic. The hue-brightness characteristic shown in Fig. 5 is a characteristic when hue is expressed as transmittance.

[0069] When a specific amount of light is irradiated onto the holding member 1, the color-brightness characteristic is expressed as follows: Y=b １ *X ２ +b ２ *X+b ３ (b １ , b２ , b ３ The hue-brightness characteristic may be expressed as a quadratic function of hue (X) (for example, a polynomial of second order or higher).

[0070] The color-brightness characteristic differs depending on the amount of light irradiated onto the holding member 1 (i.e., the amount of light from the light source 3). Therefore, the color-brightness characteristic is calculated for each of the multiple amounts of light irradiated onto the holding member 1. As the amount of light irradiated onto the holding member 1 differs, the coefficient b １ , b ２ , b ３ is different.

[0071] The color-brightness characteristic relation information I2 may store the above quadratic equation calculated for each different amount of light, or may store the coefficient b １ , b ２ , b ３ The information storage device may store the following information.

[0072] Returning to the description of the analysis device 100 using Fig. 1 , the analysis device 100 may include other analysis means for analyzing the particulate matter FP. Specifically, the analysis device 100 may include a collection amount measurement unit 6 and an elemental analysis unit 7.

[0073] The collected amount measurement unit 6 has a β-ray source 61 and a β-ray detector 63. The β-ray source 61 is provided inside the outlet 23 and irradiates β-rays toward the first position P1. The β-ray detector 63 is provided inside the suction port 25 so as to face the β-ray source 61 and measures the intensity of β-rays that have passed through the particulate matter FP held by the holding member 1. The calculation unit 5 calculates data related to the collected amount of particulate matter FP (mass concentration of particulate matter FP) based on the intensity of β-rays measured by the β-ray detector 63.

[0074] The elemental analysis unit 7 is provided at the third position P3. The elemental analysis unit 7 has an X-ray source 71 and a detector 73. The X-ray source 71 irradiates X-rays onto the particulate matter FP present at the third position P3. The detector 73 detects fluorescent X-rays generated from the particulate matter FP irradiated with X-rays by the X-ray source 71. The calculation unit 5 performs elemental analysis of the particulate matter FP based on the fluorescent X-rays detected by the detector 73. Specifically, the calculation unit 5 identifies elements contained in the particulate matter FP and calculates the content of the identified elements.

[0075] (3) Method for Analyzing Particulate Matter Using an Analytical Device (3-1) Method for Calculating Light Intensity-Brightness Characteristics Hereinafter, an analysis operation of particulate matter FP using the analytical device 100 having the above configuration will be described. First, a method for calculating the light intensity-brightness characteristics (light intensity-brightness characteristic related information I1) as information relating to the influence of changes in the light intensity of the light source 3 relative to the reference light intensity when the analytical image data was acquired will be described. The light intensity-brightness characteristics are calculated as follows. The calculation and storage of the light intensity-brightness characteristics can be performed, for example, when the analytical device 100 is operated for the first time, when the light source 3 is replaced, or when the analytical device 100 is calibrated.

[0076] First, multiple optical filters with different known color hues (e.g., transmittance, reflectance) are prepared. Next, an optical filter with a specific color hue is placed on a holding member 1 that does not hold particulate matter FP, etc., and a specific amount of light is irradiated from the light source 3 to obtain partial image data IM. This process is repeated while changing the amount of light output from the light source 3. As a result, multiple partial image data IM are obtained by irradiating different amounts of light onto the holding member 1 with a specific color hue. At this time, the light amount of the light source 3 is converted into the luminance (median) of pixels that are white in the partial image data IM (pixels corresponding to white areas of the holding member 1). The luminance of the white pixels will be referred to as white luminance.

[0077] For each of the multiple partial image data IM acquired with different light intensities, the luminance (median) of the pixel in the partial image data IM corresponding to the position having the specific color tone is calculated, and the light intensity at the time of acquisition of the partial image data IM is correlated with the calculated luminance. As a result, multiple coordinate points (i.e., multiple combinations of (light intensity (white luminance), luminance)) in the light intensity-luminance coordinate system are generated for the specific color tone, as shown in Figure 4.

[0078] Then, for the multiple coordinate points generated as described above, y = a １ *x ２ +a ２ *x+a ３ (x: light amount (white brightness), y: brightness, a １ , a ２ , a ３ : constant) and the coefficient a １ , a ２ , a ３ By calculating specific values ​​of the above, it is possible to calculate the light intensity-brightness characteristics for a specific hue. By performing the above method of calculating the light intensity-brightness characteristics for each of multiple hue, it is possible to calculate the light intensity-brightness characteristics for any multiple hue.

[0079] After calculating the light intensity-brightness characteristics for multiple shades of color, the quadratic equation calculated for each shade of color is stored in association with the shade in the light intensity-brightness characteristics association information I1. Alternatively, the coefficients of the quadratic equation calculated for each shade of color may be stored in association with the shade in the light intensity-brightness characteristics association information I1.

[0080] (3-2) Method for Calculating Hue-Luminance Characteristics Next, a method for calculating the hue-luminance characteristics (hue-luminance characteristics related information I2) as information regarding the influence of changes in the light intensity of the light source 3 relative to the reference light intensity when the analysis image data was acquired will be described. The hue-luminance characteristics can be calculated using the light intensity-luminance characteristics (light intensity-luminance characteristics related information I1) acquired as described above. Specifically, the calculation is performed as follows. The hue-luminance characteristics can be calculated and stored, similar to the light intensity-luminance characteristics, for example, when the analysis device 100 is operated for the first time, when the light source 3 is replaced, or when the analysis device 100 is calibrated.

[0081] First, the color tone-brightness characteristic is obtained when a specific light intensity (white luminance) is irradiated onto the holding member 1. Specifically, as shown in Fig. 6, for the light intensity-brightness characteristic calculated for each of a plurality of color tones (T1, T2, T3), the luminance is identified when the light intensity (white luminance) is fixed to a specific value (P in Fig. 5). In the example shown in Fig. 6, Lu1, Lu2, and Lu3 are identified as the luminance when the light intensity (white luminance) value is P for each of the plurality of color tones. Fig. 6 is a diagram schematically illustrating a method for calculating the color tone-brightness characteristic for a specific light intensity.

[0082] Next, the luminances (Lu1, Lu2, Lu3) identified above are associated with the hue (T1, T2, T3) at the time the luminances were identified. This generates multiple coordinate points in the hue-luminance coordinate system (i.e., multiple combinations of (hue, luminance)) for a specific amount of light (white luminance). Specifically, three coordinate points are generated: (T1, Lu1), (T2, Lu2), and (T3, Lu3).

[0083] Although the example of generating three coordinate points has been described above, this is merely an example, and more coordinate points can be generated. Specifically, by calculating the light intensity-luminance characteristics for more known hue colors, it is possible to generate more coordinate points in the hue-luminance coordinate system for a specific light intensity (white luminance) using a method similar to that described above. As a result, it is possible to calculate more accurate hue-luminance characteristics.

[0084] Then, for the multiple coordinate points generated as described above, Y=b １ *X ２ +b ２ *X+b ３ (X: color, Y: brightness, b １ , b ２ , b ３ : constant) and the coefficient b １ , b ２ , b ３ By calculating the specific numerical value of , it is possible to calculate the hue-luminance characteristics for a specific amount of light (white luminance). By performing the above-mentioned method for calculating the hue-luminance characteristics for each of a plurality of amounts of light (white luminance), it is possible to calculate the hue-luminance characteristics for any plurality of amounts of light (white luminance).

[0085] After calculating the hue-luminance characteristics for a plurality of light amounts (white luminance), the quadratic equation calculated for each light amount (white luminance) is stored in association with each light amount (white luminance) in the hue-luminance characteristics association information I2. Alternatively, the coefficients of the quadratic equation calculated for each light amount (white luminance) may be stored in association with each light amount (white luminance) in the hue-luminance characteristics association information I2.

[0086] (3-3) Analyzing Operation of Particulate Matter Next, a method for analyzing particulate matter FP using the analyzing device 100 will be described with reference to FIG. 7. FIG. 7 is a flowchart showing the analyzing operation of particulate matter. First, the collector 2 causes the holding member 1 to hold the particulate matter FP (step S1). This operation continues for a predetermined time (for example, one hour). After the particulate matter FP has been held in the holding member 1, the moving unit 11 moves the holding region of the holding member 1 in which the particulate matter FP is held from a first position P1 to a second position P2.

[0087] After moving the holding member 1, light is irradiated from the light source 3 toward the imaging target area VA of the holding member 1, and the imaging target area VA is imaged by the two-dimensional sensor 4 (step S2). As a result, image data including partial image data IM is obtained, as shown in FIG. 2. The obtained image data is output to the calculation unit 5. The calculation unit 5 cuts out the partial image data IM from the image data obtained by the two-dimensional sensor 4 as image data for analysis. Specifically, the calculation unit 5 extracts, from the image data, partial image data IM including an image of the holding area of ​​the particulate matter FP and an image of the surrounding area where the particulate matter FP is not held (non-holding area), as image data for analysis.

[0088] After acquiring the analytical image data, the calculation unit 5 corrects the luminance of the acquired analytical image data (step S3). The calculation unit 5 executes the operation shown in the flowchart of FIG. 8 to correct the luminance of the analytical image data (i.e., the analytical image data acquired in step S2) acquired when the light intensity of the light source 3 was a first light intensity (i.e., the light intensity of the light source 3 when step S2 was executed) to the luminance when the light intensity of the light source 3 was a second light intensity (i.e., a reference light intensity) different from the first light intensity, thereby generating first-corrected image data. The second light intensity, which is the reference light intensity, can be, for example, the maximum light intensity that can be output from the light source 3. FIG. 8 is a flowchart showing a method for correcting the luminance of analytical image data.

[0089] The calculation unit 5 corrects the luminance of the analysis image data based on the change in the hue of the photographed area VA illuminated by the light from the light source 3 when the light intensity of the light source 3 changes relative to the reference light intensity. Specifically, the calculation unit 5 first converts the luminance of each pixel of the analysis image data acquired in step S2 into a hue (step S31). Specifically, the calculation unit 5 converts the luminance of each pixel of the analysis image data acquired in step S2 into a hue based on a first relational expression that expresses the relationship between the luminance of a pixel of the first correction image data when the light intensity of the light source 3 is a first light intensity and the hue of the position corresponding to that pixel in the photographed area VA, i.e., the hue-luminance characteristic when the light intensity of the light source 3 is the light intensity at the time of execution of step S2.

[0090] The calculation unit 5 calculates information on the hue-luminance characteristics at the light intensity of the light source 3 when step S2 is executed (a polynomial (Y=b １１ *X ２ +b ２１ *X+b ３１ ), the above coefficient b １１ , b ２１ , b ３１ ) and uses this information to convert the luminance of the analytical image data into a hue to generate converted data. The light intensity of the light source 3 when performing step S2 can be the white luminance of the white portion (the portion where no particulate matter FP is held) of the analytical image data acquired in step S2. The converted data may be, for example, image data including pixels and hue assigned to the pixels, or may be numerical data stored in association with pixel coordinates and hue assigned to the coordinates.

[0091] Specifically, a polynomial (Y=b) representing the color-brightness characteristics when the light source 3 is in the light intensity state at the time of execution of step S2 is used. １１ *X ２ +b ２１ *X+b ３１ The luminance of each pixel in the analytical image data obtained in step S2 is substituted for "Y" in the above equation, and the quadratic equation obtained is solved for "X," and the solution (X value) is taken as the hue of that pixel. By performing this process for all pixels contained in the analytical image data, converted image data can be obtained in which the luminance of the pixels in the analytical image data has been converted into hue.

[0092] Next, the calculation unit 5 converts the hue of each pixel of the converted data generated in step S31 into luminance when the light amount of the light source 3 is set to a second light amount (a reference light amount (e.g., a maximum light amount)) to generate first corrected image data (step S32). Specifically, the calculation unit 5 converts the hue of each pixel of the converted data calculated in step S31 into luminance based on a second relational expression that expresses the relationship between the luminance of a pixel of the second correction image data when the light amount of the light source 3 is the second light amount (i.e., the reference light amount (maximum light amount)) and the hue of the position corresponding to that pixel in the shooting area VA, i.e., the hue-luminance characteristic when the light amount of the light source 3 is the reference light amount (maximum light amount).

[0093] The calculation unit 5 selects information on the color-brightness characteristic when the light amount of the light source 3 is the second light amount (reference light amount) from the color-brightness characteristic association information I2 stored in the storage unit 53 (a polynomial (Y= ｂ１２ *X ２ + ｂ２２ *X+b ３２ ), the above coefficient b １２ , b ２２ , b ３２ ) is selected, and this information is used to convert the hue of each pixel of the converted image data obtained in step S31 into luminance to generate first corrected image data.

[0094] Specifically, a polynomial (Y=b １２ *X ２ +b ２２ *X+b ３２ ) is substituted with the hue of each pixel of the converted data obtained in step S31, and "Y" calculated by substituting the hue for "X" is set to the luminance of that pixel. By performing this process for all pixels included in the converted data, it is possible to obtain first corrected image data (i.e., new image data for analysis) in which the luminance of each pixel is set to the luminance when the amount of light from the light source 3 is the reference amount of light. The first corrected image data obtained in this way has been generated by removing the influence of changes in the amount of light from the light source 3 relative to the reference amount of light when the image data for analysis was obtained, and is not affected by changes in the amount of light from the light source 3.

[0095] The luminance of the analytical image data acquired in step S2 has a characteristic of changing nonlinearly with changes in the light intensity of the light source 3. In this embodiment, the luminance of such analytical image data is corrected based on the change in hue of the photographed area VA illuminated by the light of the light source 3 when the light intensity of the light source 3 changes relative to a reference light intensity. Specifically, the luminance of the analytical image data is corrected using a hue-luminance characteristic, which is a nonlinear function. That is, the effect of changes in the light intensity of the light source 3 relative to a reference light intensity when the analytical image data is acquired is that the luminance of the pixels of the partial image data IM (analysis image data) acquired by the two-dimensional sensor 4 changes nonlinearly with changes in the light intensity of the light source 3. Based on the characteristic of changing nonlinearly with changes in the light intensity of the light source 3, the luminance of the analytical image data is corrected using nonlinear correction to remove this effect, thereby generating the first corrected image data. As a result, even if the luminance of the pixels of the analytical image data has a complex characteristic of changing nonlinearly with changes in the light intensity of the light source 3, the luminance of the analytical image data can be more accurately corrected to generate the first corrected image data.

[0096] 4, in the analysis image data obtained by analysis device 100, small values ​​of luminance (i.e., luminance when the hue is small) have a small rate of change (slope of the light intensity-luminance characteristic) relative to a change in the light intensity of light source 3, while large values ​​of luminance (i.e., luminance when the hue is large) have a large rate of change relative to a change in the light intensity of light source 3. Because the light intensity-luminance characteristic has such characteristics, the hue-luminance characteristic calculated using the light intensity-luminance characteristic also reflects these characteristics.

[0097] Therefore, when the luminance of the analytical image data is corrected using the above-mentioned hue-luminance characteristics, the luminance correction range for pixels having low luminance among the pixels included in the analytical image data is small, and the luminance correction range for pixels having high luminance is large. In this way, by correcting the luminance of the analytical image data using the hue-luminance characteristics, it is possible to perform appropriate luminance correction that reflects the luminance characteristics of the analytical image data.

[0098] Returning to the explanation of the analysis operation of the particulate matter FP, after the first corrected image data that is not affected by the change in the light intensity of the light source 3 is generated as described above, the particulate matter FP held by the holding member 1 is analyzed using this first corrected image data (step S4).

[0099] The calculation unit 5 generates a "brightness histogram" as shown in FIG. 9 from the first corrected image data generated by correcting the brightness of the analysis image data, and calculates particle amount-related information related to the amount of particulate matter FP from the generated brightness histogram. The brightness histogram is data that associates the brightness that each pixel of the first corrected image data may have with the number of pixels having a specific brightness included in the first corrected image data. The brightness histogram can be generated, for example, by scanning the pixels included in the first corrected image data, counting the number of pixels having a specific brightness, and associating the brightness with the number of pixels having that brightness. FIG. 9 is a diagram showing an example of a brightness histogram.

[0100] As shown in Fig. 9, in the brightness histogram generated from the first corrected image data, peaks of the number of pixels are observed at brightness Br1, which is a low brightness, and brightness Br2, which is a high brightness. Specifically, a peak PE1 at brightness Br1 and a peak PE2 at brightness Br2 are observed. The peak PE1 at low brightness Br1 corresponds to the brightness distribution of pixels included in the first region image data Im1 (i.e., image data corresponding to a retention region in which particulate matter is retained) of the first corrected image data. On the other hand, the high brightness Br2 corresponds to the brightness distribution of pixels included in the second region image data Im2 (i.e., image data corresponding to a non-retention region in which particulate matter is not retained).

[0101] In the brightness histogram, the difference ΔBr between the brightness Br1 at which peak PE1 exists and the brightness Br2 at which peak PE2 exists, i.e., the brightness difference between the brightness of the first region image data Im1 and the brightness of the second region image data Im2, is related to the amount of particulate matter FP held in the holding member 1. Therefore, the calculation unit 5 calculates the difference ΔBr between the brightness Br1 and the brightness Br2 in the brightness histogram, and can calculate particle amount-related information regarding the amount of particulate matter FP from the calculated difference ΔBr.

[0102] Alternatively, the calculation unit 5 can calculate the particle amount-related information based only on the luminance of pixels included in the first region image data Im1. Specifically, the calculation unit 5 can calculate the particle amount-related information using a calibration curve that represents the relationship between the luminance of pixels included in the first region image data Im1 and the amount of particulate matter FP. In this case, the calibration curve is stored in the storage unit 53.

[0103] In analyzing the particulate matter FP, while the particulate matter FP is held in the holding member 1 in step S1, the calculation unit 5 passes beta rays from the beta ray source 61 through the particulate matter FP held at the first position P1, detects the beta rays after passing through the particulate matter FP with the beta ray detector 63, and can calculate the amount of particulate matter FP held in the holding member 1 based on the intensity of the beta rays detected by the beta ray detector 63.

[0104] In addition, after acquiring the image data for analysis, the calculation unit 5 moves the holding member 1 using the moving unit 11 to move the holding area in which the particulate matter FP is held to the third position P3, irradiates X-rays from the X-ray source 71 onto the holding area at the third position P3, detects fluorescent X-rays generated from the holding area using the detector 73, and can identify the elements contained in the particulate matter FP and calculate the content of the identified elements based on the fluorescent X-rays detected by the detector 73.

[0105] In the analysis device 100 of the first embodiment, taking into consideration the possibility that the light intensity of the light source 3 when the analysis image data was acquired in step S2 has changed from the reference light intensity, the calculation unit 5 corrects the luminance of the analysis image data acquired in step S2 to the luminance when the light intensity of the light source 3 is set to the reference light intensity, thereby eliminating the influence of the change in light intensity and generating first corrected image data, and then calculates the particle quantity-related information using the first corrected image data.

[0106] In this way, in the analysis device 100, even if the light intensity of the light source changes from the reference light intensity when acquiring the analysis image data, the luminance of the acquired analysis image data is corrected to remove the effect of the change in light intensity, and first-corrected image data is generated assuming that the light intensity of the light source 3 is the reference light intensity. In other words, first-corrected image data that is not affected by the change in light intensity of the light source 3 is acquired. Then, this first-corrected image data is used to calculate the particle amount-related information. Because the first-corrected image data used to calculate the particle amount-related information is not affected by the change in light intensity, the particle amount-related information calculated based on the first-corrected image data is also accurate and not affected by the change in light intensity.

[0107] 2. Second embodiment (1) Overview of the second embodiment In the first embodiment described above, the particulate matter FP held in the holding member 1 is photographed by the two-dimensional sensor 4 to obtain analytical image data, the luminance of the obtained analytical image data is corrected to remove the influence of changes in light quantity to generate first corrected image data, and particle amount-related information is calculated based on this first corrected image data.

[0108] The method for calculating the particle amount-related information is not limited to this. In the second embodiment, analytical image data of the particulate matter FP held in the holding member 1 is acquired, particle amount-related information (referred to as pre-correction particle amount-related information) is calculated based on the analytical image data, and the calculated pre-correction particle amount-related information is corrected based on the amount of change when the light intensity of the light source 3 changes from the reference light intensity to the light intensity at the time the analytical image data was acquired.

[0109] In the second embodiment, only the method of calculating particle amount-related information is different from that in the first embodiment, and the configuration of the analysis device 100 and the functions of each component are the same as those in the first embodiment. Therefore, a description of the configuration of the analysis device 100 will be omitted here.

[0110] (2) Method for Analyzing Particulate Matter in the Second Embodiment A method for analyzing particulate matter FP in the second embodiment will now be described with reference to FIG. 10. FIG. 10 is a flowchart showing the method for analyzing particulate matter in the second embodiment. First, the particulate matter FP is held on the holding member 1 by the collection unit 2 (step S101), light is irradiated from the light source 3 toward the imaging target area VA of the holding member 1, and the imaging target area VA is imaged by the two-dimensional sensor 4 to obtain image data for analysis (step S102). The operations of steps S101 to S102 above are the same as steps S1 to S2 in the first embodiment, and therefore will not be described in detail here.

[0111] After acquiring the analytical image data, the calculation unit 5 calculates pre-correction particle amount-related information related to the amount of particulate matter FP held in the holding member 1 based on the acquired analytical image data (step S103). That is, the pre-correction particle amount-related information is calculated using analytical image data whose brightness has not been corrected. The calculation unit 5 calculates the pre-correction particle amount-related information using a brightness histogram generated from the analytical image data acquired in step S102, or calculates the pre-correction particle amount-related information based on the brightness of pixels included in the first region image data Im1 of the analytical image data acquired in step S102.

[0112] After calculating the pre-correction particle amount-related information, the calculation unit 5 corrects the pre-correction particle amount-related information calculated in step S103 (step S104). The calculation unit 5 corrects the pre-correction particle amount-related information calculated based on the analysis image data acquired in step S102 based on the amount of change when the light intensity of the light source 3 changes from the second light intensity (i.e., the reference light intensity) to the first light intensity (i.e., the light intensity when step S102 is performed).

[0113] Specifically, when the pre-correction particle amount-related information is proportional to the light amount, the calculation unit 5 corrects the pre-correction particle amount-related information calculated in step S103 by calculating the product of the pre-correction particle amount-related information calculated in step S103 and the ratio of the light amount during execution of step S102 to the reference light amount. More specifically, when the pre-correction particle amount-related information calculated in step S103 is A, the corrected particle amount-related information is A', the light amount during execution of S102 is P1, and the reference light amount is P2, the relationship A:A'=P1:P2 holds, and the calculation unit 5 can calculate the corrected particle amount-related information from the equation A'=A*(P2 / P1), which rewrites this relationship for A'.

[0114] The calculation unit 5 uses the corrected particle amount related information calculated as described above as the final analysis result regarding the amount of particulate matter FP.

[0115] As described above, in the analysis device 100 of the second embodiment, even if analytical image data is acquired when the light intensity of the light source 3 changes from the reference light intensity and particle amount-related information (pre-correction particle amount-related information) is calculated based on the analytical image data, this pre-correction particle amount-related information is corrected based on the amount of change when the light intensity of the light source 3 changes from the reference light intensity. In other words, in the second embodiment, accurate particle amount-related information that is not affected by changes in the light intensity of the light source 3 is calculated. As a result, the particulate matter FP can be analyzed more accurately.

[0116] Furthermore, the analysis device 100 of the second embodiment can calculate particle amount-related information that is not affected by changes in light intensity without correcting the luminance of the analysis image data, thereby reducing the information processing load on the analysis device 100.

[0117] In addition, in cases where the particle amount-related information is not proportional to the amount of light, for example, a correction formula for correcting the particle amount-related information may be calculated in advance as a function of the amount of light before and after the change in the amount of light, and the particle amount-related information may be corrected using this correction formula.

[0118] 3. Third Embodiment In the above-described analysis device 100, information relating to the influence of changes in the light intensity of the light source 3 relative to a reference light intensity when image data (partial image data IM) is acquired by capturing an image of the target area VA with the two-dimensional sensor 4, i.e., correction formulas for correcting the luminance of the analytical image data (color-luminance characteristics, light intensity-luminance characteristics) and correction formulas for correcting the particle amount-related information, are calculated within the analysis device 100. However, the present invention is not limited to this. In the third embodiment, information relating to the influence of changes in the light intensity of the light source 3 relative to a reference light intensity when the analytical image data is acquired (the above-described correction formulas) is calculated by a device separate from the analysis device 100, and the information calculated by this device is received by the analysis device 100 and stored in the storage unit 53.

[0119] Specifically, as shown in Fig. 11, an analysis system 300 of the third embodiment includes the above-described analysis device 100 and a correction device 200. Fig. 11 is a diagram showing the configuration of the analysis system of the third embodiment.

[0120] Correction device 200 is a computer system configured with memory devices such as a CPU, RAM, and ROM, and various interfaces (such as communication interfaces). When analysis device 100 corrects the luminance of analysis image data, correction device 200 calculates a first correction formula (i.e., color tone-luminance characteristics) for correcting the luminance of analysis image data acquired when the light intensity of light source 3 is a first light intensity (the light intensity when step S2 is executed) to the luminance when the light intensity of the light source is a second light intensity (a reference light intensity) different from the first light intensity, and transmits this to analysis device 100.

[0121] In the above case, correction device 200 may transmit the first correction formula itself to analysis device 100, or may transmit only the parameters (coefficients of the polynomial) included in the first correction formula to analysis device 100. In the latter case, analysis device 100 stores the polynomial that constitutes the first correction formula, and can generate the first correction formula by substituting the coefficients calculated by correction device 200 for the coefficients of this polynomial.

[0122] Furthermore, when the analysis device 100 corrects the particle amount-related information, the correction device 200 calculates a second correction formula for correcting the uncorrected particle amount-related information, which is calculated based on the analysis image data whose luminance has not been corrected, based on the amount of change when the light intensity of the light source 3 changes from the second light intensity (reference light intensity) to the first light intensity (light intensity when step S102 is executed), and transmits this to the analysis device 100.

[0123] In this way, by calculating the first correction formula or the second correction formula by a device (correction device 200) separate from the analysis device 100, the information processing load on the analysis device 100 can be reduced.

[0124] 4. Fourth Embodiment In the above-described first and second embodiments, the correction formulas for correcting the analytical image data (light intensity-luminance characteristic related information I1, color tone-luminance characteristic related information I2) and the correction formulas for correcting the pre-correction particle amount related information are calculated by calculations performed by the calculation unit 5 of the analysis device 100 or the correction device 200. However, the present invention is not limited to this, and these correction formulas may be calculated by machine learning, neural networks, deep learning, or the like.

[0125] For example, the equation y=a expressing the light quantity-luminance characteristic １ *x ２ +a ２ *x+a ３ (x: light amount (white brightness), y: brightness, a １ , a ２ , a ３ : constant) coefficient a １ , a ２ , a ３ , and / or the color-brightness characteristic formula Y=b １ *X ２ +b ２ *X+b ３ (X: color, Y: brightness, b １ , b ２ , b ３ : constant) coefficient b １ , b ２ , b ３ can be calculated using machine learning, neural networks, deep learning, etc.

[0126] 5. Fifth Embodiment The calculation unit 5 of the analysis device 100 may calculate particle amount-related information using a trained model in which artificial intelligence such as machine learning, a neural network, or deep learning is trained using training data that represents the effect of changes in the light intensity of the light source 3 relative to a reference light intensity when the target area VA is photographed with the two-dimensional sensor 4 to acquire image data (partial image data IM). Note that this trained model may be generated by the correction device 200 and transmitted to the analysis device 100, or may be generated by the calculation unit 5. Examples of methods for calculating particle amount-related information using a trained model include the following.

[0127] (1) Method 1 for calculating particle amount-related information using a trained model In the first method, image data (partial image data IM) acquired by photographing the photographing target area VA with the two-dimensional sensor 4 is input, and information related to the amount of particulate matter FP held in the holding member 1 when the image data was acquired is output. By inputting the analysis image data into a trained model trained with training data, particle amount-related information can be output from the trained model. Specifically, the particle amount-related information can be calculated as follows.

[0128] First, a plurality of partial image data IM are acquired by changing the light intensity of the light source 3 for the holding member 1 holding a predetermined amount of particulate matter FP. This is repeated by changing the amount of particulate matter FP held by the holding member 1.

[0129] Next, the plurality of partial image data IM acquired as described above are used as input, and artificial intelligence such as machine learning, neural networks, and deep learning is trained using training data that outputs information regarding the amount of particulate matter FP held in the holding member 1 when the partial image data IM was acquired, to generate a trained model. At this time, the artificial intelligence is trained using training data acquired by changing the amount of light. Therefore, in this embodiment, it can be said that the artificial intelligence is trained based on the effect of changes in the amount of light relative to a reference amount of light, and a trained model is obtained.

[0130] Thereafter, the calculation unit 5 inputs the analytical image data acquired during analysis execution into the trained model generated as described above, and can output the amount of particulate matter FP at the time the analytical image data was acquired as particle quantity-related information from this trained model. As described above, the trained model was obtained by training artificial intelligence based on the influence of changes in light intensity relative to a reference light intensity. In other words, the trained model has learned the influence of changes in light intensity relative to a reference light intensity. Therefore, it can be said that the calculation unit 5, which calculates particle quantity-related information using this trained model, calculates the particle quantity-related information based on the analytical image data and the influence of changes in light intensity relative to a reference light intensity.

[0131] According to this method, accurate particle amount-related information that is not affected by changes in the light intensity of the light source 3 can be calculated by directly using the analytical image data acquired by the two-dimensional sensor 4 during analysis.

[0132] (2) Method 2 for calculating particle quantity-related information using a trained model In a second method, analytical image data is input to a trained model trained with training data that inputs image data (partial image data IM) acquired by photographing the photographing target area VA with the two-dimensional sensor 4 and outputs corrected image data generated by correcting the luminance of the image data to the luminance when the light intensity of the light source 3 is a reference light intensity. Second corrected image data in which the luminance of the analytical image data has been corrected to the luminance when the light intensity of the light source 3 is a reference light intensity is output from the trained model, and particle quantity-related information can be calculated based on this second corrected image data. Specifically, the particle quantity-related information can be calculated as follows.

[0133] First, a plurality of partial image data IM are acquired by varying the light intensity of the light source 3 for the holding member 1 holding an arbitrary amount of particulate matter FP. Then, the luminance of each acquired partial image data IM is corrected to the luminance when the light intensity of the light source 3 is the reference light intensity, and corrected image data is generated. This luminance correction can be performed, for example, by the method described in the first embodiment.

[0134] Next, the multiple partial image data IM obtained as described above are used as input, and the corrected image data generated for the partial image data IM is used as output as training data to train artificial intelligence such as machine learning, neural networks, and deep learning, thereby generating a trained model.

[0135] Thereafter, the calculation unit 5 inputs the analysis image data acquired during analysis execution into the trained model generated as described above, and can output from this trained model second corrected image data in which the luminance of the analysis image data has been corrected to the luminance when the light intensity of the light source 3 is the reference light intensity. Based on this second corrected image data (for example, using the luminance histogram of the second corrected image data), the calculation unit 5 can calculate particle amount-related information of the particulate matter FP when the analysis image data during analysis execution was acquired.

[0136] According to this method, second corrected image data that is not affected by changes in the light intensity of the light source 3 can be generated directly from the analytical image data acquired by the two-dimensional sensor 4 during analysis. Furthermore, by calculating particle amount-related information based on this second corrected image data, accurate particle amount-related information that is not affected by changes in the light intensity of the light source 3 can be calculated.

[0137] (3) Method 3 for Calculating Particle Quantity-Related Information Using a Trained Model In a third method, a trained model trained using teacher data is input, which inputs a first representative luminance of pixels corresponding to the non-retained region and a second representative luminance of pixels corresponding to the retained region in image data (partial image data IM) acquired by photographing the photographing target region VA with the two-dimensional sensor 4, and outputs a first corrected representative luminance of pixels corresponding to the non-retained region and a second corrected representative luminance of pixels corresponding to the retained region when the luminance of the pixels of the partial image data IM is corrected to a luminance when the light intensity of the light source 3 is a reference light intensity. The trained model then outputs the third corrected representative luminance of pixels corresponding to the non-retained region and the fourth corrected representative luminance of pixels corresponding to the retained region when the luminance of the pixels of the analytical image data is corrected to a luminance when the light intensity of the light source 3 is a reference light intensity. Then, particle quantity-related information can be calculated based on the third corrected representative luminance and the fourth corrected representative luminance. Specifically, the particle amount related information can be calculated as follows.

[0138] First, a holding member 1 holding an arbitrary amount of particulate matter FP is photographed by the two-dimensional sensor 4 while changing the light intensity of the light source 3, thereby acquiring a plurality of partial image data IM. From each of the plurality of partial image data IM, a representative brightness (first representative brightness) of the pixels of the portion corresponding to the non-holding region where the particulate matter FP is not held (second region image data Im2) and a representative brightness (second representative brightness) of the pixels of the portion corresponding to the holding region where the particulate matter FP is held (first region image data Im1) are extracted. Here, the representative brightness can be, for example, the brightness value possessed by the largest number of pixels among the plurality of pixels included in the above region image data. Alternatively, the representative brightness can be the average value of the brightness values ​​possessed by the plurality of pixels included in the above region image data.

[0139] In addition, the brightness of each acquired partial image data IM is corrected to the brightness when the light intensity of the light source 3 is a reference light intensity to generate corrected image data, and the representative brightness of the pixels corresponding to the non-retained area of ​​this corrected image data (representative brightness after the first correction) and the representative brightness of the pixels corresponding to the retained area (representative brightness after the second correction) are extracted.

[0140] Next, the first representative luminance and the second representative luminance extracted as described above are used as input, and artificial intelligence such as machine learning, neural networks, and deep learning is trained using training data that outputs the first corrected representative luminance and the second corrected representative luminance, to generate a trained model.

[0141] Thereafter, the calculation unit 5 extracts the representative luminance (third representative luminance) of pixels corresponding to the non-retained regions and the representative luminance (fourth representative luminance) of pixels corresponding to the retained regions of the analysis image data acquired during analysis execution, and inputs the extracted third representative luminance and fourth representative luminance into the trained model. This allows the calculation unit 5 to output from this trained model the representative luminance (third corrected representative luminance) of pixels corresponding to the non-retained regions and the representative luminance (fourth corrected representative luminance) of pixels corresponding to the retained regions of the corrected image data obtained by correcting the luminance of the analysis image data acquired during analysis execution to the luminance when the light intensity of the light source 3 is a reference light intensity.

[0142] The calculation unit 5 can calculate particle amount-related information of the particulate matter FP when the analysis image data is acquired during analysis based on the third corrected representative brightness and the fourth corrected representative brightness output from the trained model (for example, using a calibration curve showing the relationship between the representative brightness of pixels corresponding to the non-retention area and the representative brightness of pixels corresponding to the retention area and the amount of particulate matter FP retained).

[0143] According to this method, compared to using all of the analytical image data acquired by the two-dimensional sensor 4 when performing the analysis, a trained model can be trained using a smaller amount of information to calculate accurate particle quantity-related information that is not affected by changes in the light intensity of the light source.

[0144] 6. Other Embodiments Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and various modifications are possible without departing from the spirit of the invention. In particular, the multiple embodiments and modifications described in this specification can be arbitrarily combined as needed. (A) The analysis device 100 may be capable of executing any of the analysis methods of the first to fifth embodiments. Furthermore, it may be possible to switch between the analysis methods to be executed.

[0145] (B) In addition to the above-described correction of the brightness of the analytical image data and the correction of the particle amount-related information, background correction may be performed on the analytical image data. The background correction corrects the phenomenon in which the brightness (background) of the second region image data Im2 (image data corresponding to the non-retention region where no particulate matter FP is retained) of the analytical image data varies depending on the location on the retaining member 1. This allows the particle amount-related information to be calculated more accurately using the analytical image data.

[0146] (C) A sensor capable of acquiring a color image may be used as the two-dimensional sensor 4. This makes it possible to acquire information about the hue of image data for analysis and the output difference between each color.

[0147] The color of particulate matter FP varies depending on its type. For example, if the main component of particulate matter FP is yellow sand, the analytical image data will be yellowish. If the main component is iron oxide, copper, or brown carbon, the analytical image data will be reddish or brownish. Furthermore, if the main component is ash or cement generated by open burning, the analytical image data will be white or grayish. Therefore, if a color image can be obtained as analytical image data, data on the components contained in the particulate matter FP can be obtained based on the hue of the analytical image data, etc.

[0148] (D) The collection unit 2 may be omitted from the analysis device 100. In this case, the particulate matter FP may be held in the holding member 1 in advance.

[0149] (E) The analysis results of particulate matter FP, information calculated during the analysis process, brightness-corrected filter images, etc., i.e., particle quantity-related information, corrected particle quantity-related information, first corrected image data, or second corrected image data, may be displayed on the display 55 of the calculation unit 5.

[0150] 7. Features of the Embodiments The above embodiments can also be described as follows. (1) An analysis device (e.g., analysis device 100) includes a holding member (e.g., holding member 1), a light source (e.g., light source 3), a two-dimensional sensor (e.g., two-dimensional sensor 4), and a calculation unit (e.g., calculation unit 5). The holding member holds particles (e.g., particulate matter FP). The light source irradiates light onto an imaging target area (e.g., imaging target area VA) of the holding member that includes a holding area where the particles are held. The two-dimensional sensor images the imaging target area. The calculation unit calculates particle amount-related information regarding the amount of particles held in the holding member based on analysis image data (e.g., analysis image data) acquired by imaging the imaging target area with the two-dimensional sensor for use in particle analysis and the effect of changes in the light source light intensity when the analysis image data was acquired relative to a reference light intensity.

[0151] In the above-described analysis device, analytical image data of a photographed area including a holding area where particles are held is acquired for use in particle analysis, and information relating to the amount of particles (particle amount-related information) is calculated based on the acquired analytical image data and the effect of changes in the light intensity of the light source at the time the analytical image data was acquired relative to a reference light intensity. This allows accurate particle amount-related information to be calculated that is not affected by changes in the light intensity of the light source.

[0152] (2) In the analysis device of (1), the influence may be a nonlinear change in pixel brightness of the analysis image data with respect to a change in the light intensity of the light source. This makes it possible to calculate accurate particle amount-related information that is not affected by changes in the light intensity of the light source, even when analysis pixel data having a complex characteristic in which pixel brightness changes nonlinearly with respect to a change in the light intensity of the light source is used.

[0153] (3) In the analysis device of (1) or (2), the calculation unit may correct the luminance of the analytical image data based on a change in color of the photographed area illuminated by the light from the light source when the light source light intensity changes relative to a reference light intensity, and calculate the particle amount-related information based on the corrected analytical image data. This allows the luminance of the analytical image data to be corrected more accurately, and therefore the particle amount-related information to be calculated more accurately based on the corrected analytical image data.

[0154] (4) In any of the analysis devices (1) to (3) above, the calculation unit may calculate the particle amount-related information based on first corrected image data in which the influence of the first light amount change is eliminated by correcting the luminance of the analysis image data acquired by capturing an image of the target area using the two-dimensional sensor when the light amount of the light source is a first light amount to the luminance when the light amount of the light source is a second light amount, which is a reference light amount different from the first light amount. This allows the calculation of accurate particle amount-related information that is not affected by changes in light amount, even if the analysis image data is acquired when the light amount of the light source has changed from the reference light amount (i.e., the second light amount).

[0155] (5) In the analysis device of (4), the calculation unit may generate the first corrected image data by reducing the luminance correction range of pixels having low luminance in the analysis image data and increasing the luminance correction range of pixels having high luminance. This allows the analysis image data to be corrected more accurately by reflecting the tendency of the luminance change range in response to changes in the light intensity of the light source.

[0156] (6) In the analysis device of (4) or (5) above, the calculation unit may generate converted data by converting the luminance of each pixel of the analysis image data to a hue based on a first relational expression (e.g., a hue-luminance characteristic at a first light intensity) that represents the relationship between the luminance of a pixel of the first correction image data acquired by capturing an image of the target area with a two-dimensional sensor when the light source has a first light intensity and the hue of a position corresponding to that pixel in the image of the target area; and may generate the first corrected image data by converting the hue of each pixel of the converted data to a luminance when the light source has a second light intensity based on a second relational expression (e.g., a hue-luminance characteristic at a second light intensity) that represents the relationship between the luminance of a pixel of the second correction image data acquired by capturing an image of the target area with a two-dimensional sensor when the light source has a second light intensity and the hue of a position corresponding to that pixel in the image of the target area. This allows for more accurate correction of the luminance of the analysis image data.

[0157] (7) In the analysis device of (6), the first relational expression and the second relational expression may respectively represent the luminance of pixels in the first correction image data and the second correction image data as polynomials of second order or higher of the hue of the portion of the subject area corresponding to the pixel. This allows the first relational expression and the second relational expression to more accurately represent the relationship between the luminance and hue of the pixel of the analysis image data. As a result, the luminance of the analysis image data can be corrected more accurately.

[0158] (8) In the analysis device of (1) or (2) above, the calculation unit may calculate pre-correction particle amount-related information based on the analysis image data, and correct the pre-correction particle amount-related information based on the amount of change when the light source light intensity changes from a reference light intensity to the light intensity when the analysis image data was acquired, thereby calculating the particle amount-related information. In this way, even if the analysis image data is acquired when the light source light intensity has changed from the reference light intensity, and the pre-correction particle amount-related information is calculated based on the analysis image data, accurate particle amount-related information that is not affected by the change in the light source light intensity can be calculated.

[0159] (9) In any of the analytical devices described in (1) to (8), the calculation unit may calculate the particle amount-related information using a trained model that has learned the influence of changes in the light intensity of the light source relative to a reference light intensity, thereby enabling calculation of accurate particle amount-related information that is not affected by changes in the light intensity of the light source.

[0160] (10) In the analysis device of (9) above, the calculation unit may input the analytical image data to a trained model that has learned the above-mentioned influences using training data, and the trained model receives image data acquired by capturing an image of the target area using a two-dimensional sensor and outputs information related to the amount of particles held by the holding member when the image data was captured. This allows the analytical image data to be used directly to calculate accurate particle amount-related information that is not affected by changes in the light intensity of the light source.

[0161] (11) In the analysis device of (9) above, the calculation unit may input the analysis image data to a trained model that has learned the above-mentioned influence using training data, and the trained model receives image data acquired by photographing the target area using a two-dimensional sensor and outputs corrected image data generated by correcting the luminance of the image data to the luminance when the light source has a standard light intensity. The calculation unit then outputs second corrected image data from the trained model, in which the luminance of the analysis image data has been corrected to the luminance when the light source has a standard light intensity, and calculates particle quantity-related information based on the second corrected image data.

[0162] This allows the second corrected image data, which is not affected by changes in the light intensity of the light source, to be generated directly from the analytical image data. Furthermore, by calculating particle amount-related information based on this second corrected image data, accurate particle amount-related information, which is not affected by changes in the light intensity of the light source, can be calculated.

[0163] (12) In the analysis device of (9), the imaging target area may have the retention area and a non-retention area where no particles are retained. In this case, the calculation unit may input the third representative luminance and the fourth representative luminance of the analysis image data to a trained model that has learned the above influence using teacher data that receives the first representative luminance and the second representative luminance as input and outputs the first corrected representative luminance and the second corrected representative luminance, thereby outputting the third corrected representative luminance and the fourth corrected representative luminance from the trained model, and calculate the particle amount-related information based on the third corrected representative luminance and the fourth corrected representative luminance.

[0164] The first representative luminance is the representative luminance of pixels corresponding to the non-retained area of ​​image data acquired by capturing an image of the target area with a two-dimensional sensor. The second representative luminance is the representative luminance of pixels corresponding to the retained area of ​​this image data. The first corrected representative luminance is the representative luminance of pixels corresponding to the non-retained area when the luminance of this image data is corrected to the luminance when the light intensity of the light source is a reference light intensity. The second corrected representative luminance is the representative luminance of pixels corresponding to the retained area when the luminance of this image data is corrected to the luminance when the light intensity of the light source is a reference light intensity. The third representative luminance is the representative luminance of pixels corresponding to the non-retained area of ​​the analysis image data. The fourth representative luminance is the representative luminance of pixels corresponding to the retained area of ​​the analysis image data. The third corrected representative luminance is the representative luminance of pixels corresponding to the non-retained area when the luminance of the analysis image data is corrected to the luminance when the light intensity of the light source is a reference light intensity. The fourth corrected representative luminance is the representative luminance of pixels corresponding to the retained area when the luminance of the analysis image data is corrected to the luminance when the light intensity of the light source is a reference light intensity.

[0165] This makes it possible to train a trained model using a small amount of information to calculate accurate particle quantity-related information that is not affected by changes in the light intensity of the light source.

[0166] (13) An analysis system (e.g., analysis system 300) includes the above-described analysis device and a correction device (e.g., correction device 200). The analysis device includes a holding member, a light source, a two-dimensional sensor, and a calculation unit. The holding member holds particles. The light source irradiates light onto an imaging target area including a holding region of the holding member where the particles are held. The two-dimensional sensor images the imaging target area. The calculation unit calculates particle amount-related information regarding the amount of particles held in the holding member based on analysis image data acquired by imaging the imaging target area with the two-dimensional sensor for use in particle analysis, and the effect of changes in the light source light intensity when the analysis image data was acquired relative to a reference light intensity.

[0167] The correction device generates information regarding the influence of a change in the light intensity of the light source relative to a reference light intensity when the analysis image data was acquired, and transmits the information to the analysis device.

[0168] In the above analysis system, the analysis device acquires analytical image data of the imaging target area, including the holding area where the particles are held, for use in particle analysis, and calculates information related to the amount of particles (particle amount-related information) based on the acquired analytical image data and the effect of changes in the light intensity of the light source at the time the analytical image data was acquired relative to a reference light intensity. This allows accurate particle amount-related information to be calculated that is not affected by changes in the light intensity of the light source.

[0169] (14) An analysis method includes the following steps: a step of irradiating a photographing target area including a holding area where particles are held in a holding member that holds particles with light from a light source; a step of photographing the photographing target area and acquiring analytical image data to be used for particle analysis; and a step of calculating particle amount-related information regarding the amount of particles held in the holding member based on the analytical image data and the effect of changes in the light amount of the light source at the time the analytical image data was acquired relative to a reference light amount.

[0170] In the above analysis method, analytical image data of a photographed area including a holding area where particles are held is acquired for use in particle analysis, and information on the amount of particles (particle amount-related information) is calculated based on the acquired analytical image data and the effect of changes in the light intensity of the light source at the time the analytical image data was acquired relative to a reference light intensity. This allows accurate particle amount-related information to be calculated that is not affected by changes in the light intensity of the light source.

[0171] (15) The program is a program for causing a computer to execute the above-described analysis method.

[0172] The present invention can be widely applied to analytical devices that analyze particulate matter.

[0173] 300: Analysis system 200: Correction device 100: Analysis device 1: Holding member 11: Moving unit 11a: Take-up reel 11b: Delivery reel 2: Collection unit 21: Suction pump 23: Discharge port 25: Suction port 3: Light source 4: Two-dimensional sensor 41: Plate-shaped member 5: Calculation unit 51: CPU 53: Storage unit I1: Light quantity-brightness characteristic related information I2: Color-brightness characteristic related information 55: Display 57: Interface 6: Collection amount measurement unit 61: Beta ray source 63: Beta ray detector 7: Elemental analysis unit 71: X-ray source 73: Detector FP: Particulate matter Gs: Sample gas IM: Partial image data VA: Imaged area

Claims

1. An analytical apparatus comprising: a holding member that holds particles; a light source that irradiates light onto an area to be photographed including a holding area of ​​the holding member in which the particles are held; a two-dimensional sensor that photographs the area to be photographed; and a calculation unit that calculates particle amount-related information regarding the amount of particles held in the holding member based on analysis image data obtained by photographing the area to be photographed with the two-dimensional sensor for use in analyzing the particles, and an effect of the change in the amount of light of the light source when the analysis image data was obtained relative to a reference light amount.

2. The analysis device according to claim 1, wherein the effect is that the luminance of the pixels of the analysis image data changes nonlinearly with respect to the change in the amount of light from the light source.

3. The analysis device of claim 1 or 2, wherein the calculation unit corrects the luminance of the analysis image data based on a change in color of the photographed area irradiated with light from the light source when the amount of light from the light source changes relative to a reference amount of light, and calculates the particle amount-related information based on the corrected analysis image data.

4. An analytical device as described in any one of claims 1 to 3, wherein the calculation unit calculates the particle amount-related information based on first corrected image data in which the influence is removed by correcting the luminance of the analytical image data obtained by photographing the target area with the two-dimensional sensor when the light amount of the light source is a first light amount to the luminance when the light amount of the light source is a second light amount, which is the reference light amount different from the first light amount.

5. The analysis device described in claim 4, wherein the calculation unit reduces the luminance correction range of pixels having low luminance in the analysis image data and increases the luminance correction range of pixels having high luminance to eliminate the influence and generate the first corrected image data.

6. The analysis device described in claim 4 or 5, wherein the calculation unit generates converted data in which the luminance of each pixel of the analysis image data is converted to a hue based on a first relational equation expressing the relationship between the luminance of a pixel of first correction image data obtained by photographing the target area with the two-dimensional sensor when the amount of light of the light source is the first light amount and the hue of a position in the target area corresponding to the pixel, and generates the first corrected image data by converting the hue of each pixel of the converted data to the luminance when the amount of light of the light source is the second light amount based on a second relational equation expressing the relationship between the luminance of a pixel of second correction image data obtained by photographing the target area with the two-dimensional sensor when the amount of light of the light source is the second light amount and the hue of a position in the target area corresponding to the pixel.

7. The analysis device described in claim 6, wherein the first relational equation and the second relational equation respectively express the brightness of pixels of the first correction image data and the second correction image data as polynomials of second or higher order of the color tone of the parts of the subject area corresponding to the pixels.

8. The analytical device according to claim 1 or 2, wherein the calculation unit calculates pre-correction particle amount-related information based on the analysis image data, and calculates the particle amount-related information by correcting the pre-correction particle amount-related information based on an amount of change when the light amount of the light source changes from the reference light amount to the light amount when the analysis image data was acquired.

9. An analytical device according to any one of claims 1 to 8, wherein the calculation unit calculates the particle amount-related information using a trained model that has learned the effect of changes in the light amount of the light source relative to a reference light amount.

10. The analytical device described in claim 9, wherein the calculation unit inputs the image data for analysis to the trained model that has learned the influence using teacher data that inputs image data acquired by photographing the target area with the two-dimensional sensor and outputs information regarding the amount of particles held by the holding member when the image data was acquired, thereby outputting the particle amount-related information from the trained model.

11. The analytical device described in claim 9, wherein the calculation unit inputs the image data for analysis to the trained model that has learned the influence using teacher data in which image data acquired by photographing the target area with the two-dimensional sensor is input and corrected image data generated by correcting the luminance of the image data to the luminance when the light amount of the light source is a reference light amount is output, and second corrected image data in which the luminance of the image data for analysis is corrected to the luminance when the light amount of the light source is a reference light amount is output from the trained model, and the particle quantity related information is calculated based on the second corrected image data.

12. The photographing target area has the retained area and a non-retained area in which the particles are not retained, and the calculation unit inputs a first representative luminance of a pixel corresponding to the non-retained area and a second representative luminance of a pixel corresponding to the retained area of ​​image data acquired by photographing the photographing target area with the two-dimensional sensor, and outputs a first corrected representative luminance of a pixel corresponding to the non-retained area and a second corrected representative luminance of a pixel corresponding to the retained area when the luminance of the pixel of the image data is corrected to a luminance when the amount of light of the light source is the reference amount of light, and inputs a third representative luminance of a pixel corresponding to the non-retained area of ​​the image data for analysis and a fourth representative luminance of a pixel corresponding to the retained area to the trained model that has learned the influence using teacher data that outputs the third corrected representative luminance of a pixel corresponding to the non-retained area and the fourth corrected representative luminance of a pixel corresponding to the retained area from the trained model when the luminance of the pixel of the image data for analysis is corrected to a luminance when the amount of light of the light source is the reference amount of light, The analysis device according to claim 9 , wherein the particle amount related information is calculated based on the third corrected representative luminance and the fourth corrected representative luminance.

13. An analysis system comprising: an analysis device comprising: a holding member for holding particles; a light source for irradiating light onto a region to be photographed including a holding region of the holding member in which the particles are held; a two-dimensional sensor for photographing the region to be photographed; and a calculation unit for calculating particle amount-related information regarding the amount of particles held in the holding member based on analysis image data obtained by photographing the region to be photographed with the two-dimensional sensor for use in analyzing the particles and an effect of changing the amount of light of the light source when the analysis image data was obtained relative to a reference light amount; and a correction device for generating information regarding the effect of changing the amount of light of the light source when the analysis image data was obtained relative to a reference light amount and transmitting the information to the analysis device.

14. An analysis method comprising the steps of: irradiating light from a light source onto a target area including a holding area in a holding member that holds particles, the target area being photographed to obtain analysis image data for use in analyzing the particles; and calculating particle amount-related information regarding the amount of the particles held in the holding member based on the analysis image data and the effect of changes in the amount of light of the light source at the time the analysis image data was obtained relative to a reference light amount.

15. A program for causing a computer to execute an analysis method comprising the steps of: irradiating light from a light source onto a target area including a holding area in a holding member that holds particles, the target area being photographed to obtain analysis image data to be used in analyzing the particles; and calculating particle amount-related information regarding the amount of the particles held in the holding member based on the analysis image data and the effect of changes in the amount of light of the light source at the time the analysis image data was obtained relative to a reference light amount.

Images