Concentration distribution analysis system, information processing system, and concentration distribution analysis method

Abstract

To effectively specify the concentration distribution of gas to be observed even in a situation where the thickness of the gas to be observed is unknown.SOLUTION: A concentration distribution analysis system 100 is to specify the concentration distribution of gas to be observed present in an observation space 200, and the system is provided with: a plurality of observation cameras 11-m that is different from each other in the direction of an optical axis passing through the observation space 200; an observation data generation unit that acquires a plurality of observation data Pn corresponding to the directions different from each other relative to the observation space 200 according to results of imaging performed by the plurality of observation cameras 11-m; and an analysis processing unit that specifies the concentration distribution of the gas to be observed through inverse radon conversion on the plurality of observation data Pn.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a technique for analyzing the concentration of a gas (hereinafter referred to as an "observation target gas") present in a space to be observed (hereinafter referred to as an "observation space"). [Background technology]

[0002] Various analytical techniques have been proposed for optically analyzing the concentration of a target gas present in an observation space. For example, Patent Document 1 discloses a technique for estimating the presence or absence of a target gas by using a first filter whose transmission band includes the absorption line of the target gas and a second filter that does not transmit the absorption line. Furthermore, Patent Document 2 discloses a configuration for determining leakage of a target gas by extracting the difference between a reference image obtained by averaging multiple background images captured by an infrared camera and a measurement image captured by the infrared camera. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] U.S. Patent No. 5,306,913 [Patent Document 2] Patent No. 7289684 Summary of the Invention [Problem to be solved by the invention]

[0004] However, conventional analytical techniques can only measure the product of the concentration of a target gas and the thickness of the target gas. Therefore, when the thickness of the target gas is unknown, it is difficult to determine the concentration itself. In consideration of the above, one aspect of the present disclosure aims to effectively determine the concentration distribution of a target gas even when the thickness of the target gas is unknown. [Means for solving the problem]

[0005] In order to solve the above problems, a concentration distribution analysis system according to one aspect of the present disclosure is a concentration distribution analysis system that identifies the concentration distribution of an observation target gas present in an observation space, and includes a plurality of observation cameras with different directions of optical axes passing through the observation space, an observation data generation unit that acquires a plurality of observation data corresponding to different directions relative to the observation space based on the results of imaging by the plurality of observation cameras, and an analysis processing unit that identifies the concentration distribution of the observation target gas by performing an inverse Radon transform on the plurality of observation data.

[0006] An information processing system according to one aspect of the present disclosure is an information processing system that uses multiple observation cameras with different optical axis directions passing through the observation space to identify the concentration distribution of a target gas present in the observation space, and includes an observation data generation unit that acquires multiple observation data corresponding to different directions relative to the observation space based on the results of imaging by the multiple observation cameras, and an analysis processing unit that identifies the concentration distribution of the target gas by performing an inverse Radon transform on the multiple observation data.

[0007] A concentration distribution analysis method according to one aspect of the present disclosure is a concentration distribution analysis method implemented by a computer system that uses multiple observation cameras with different optical axis directions passing through the observation space to identify the concentration distribution of an observation target gas present in the observation space, and includes obtaining multiple observation data corresponding to different directions relative to the observation space based on the results of imaging by the multiple observation cameras, and identifying the concentration distribution of the observation target gas by performing an inverse Radon transform on the multiple observation data. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a block diagram illustrating a configuration of a concentration distribution analysis system according to a first embodiment of the present disclosure. [Figure 2] FIG. 2 is an explanatory diagram of the imaging range of each observation camera. [Figure 3] FIG. 2 is a block diagram illustrating an example of a functional configuration of an information processing system. [Figure 4] FIG. 1 is an explanatory diagram of observation data. [Figure 5] This is an explanatory diagram of the observation data corresponding to 0°. [Figure 6] This is an explanatory diagram of observation data corresponding to 45°. [Figure 7] This is an explanatory diagram of observation data corresponding to 90°. [Figure 8] This is an explanatory diagram of the observation data corresponding to 135°. [Figure 9] This is an explanatory diagram of observation data corresponding to 180°. [Figure 10] 1 is a flowchart illustrating a specific procedure for an inverse Radon transform. [Figure 11] FIG. 1 is an explanatory diagram of an inverse Radon transform. [Figure 12] 10 is a specific example of a concentration distribution. [Figure 13] 10 is a flowchart illustrating a specific procedure of a concentration distribution analysis process. [Figure 14] FIG. 10 is a block diagram illustrating the configuration of a concentration distribution analysis system according to a second embodiment. [Figure 15] FIG. 10 is an explanatory diagram of observation data corresponding to 45° in the second embodiment. [Figure 16] FIG. 10 is an explanatory diagram of observation data corresponding to 135° in the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] The embodiments for carrying out the present disclosure will be described with reference to the drawings. Note that the dimensions and scale of each element in each drawing may differ from those of the actual product. Furthermore, the embodiment described below is an exemplary embodiment that may be envisioned when carrying out the present disclosure. Therefore, the scope of the present disclosure is not limited to the embodiment exemplified below.

[0010] A: First embodiment 1 is a block diagram illustrating the configuration of a concentration distribution analysis system 100 according to a first embodiment of the present disclosure. The concentration distribution analysis system 100 is a measurement system for optically estimating the concentration distribution D of a gas to be observed (observation target gas). The observation target gas is a gas present in an observation space 200. The observation space 200 is a space within a flow path (e.g., a flue) through which the observation target gas generated in, for example, an industrial process or a chemical process flows.

[0011] As illustrated in Figure 1, the concentration distribution analysis system 100 is a computer system including an observation system 10 and an information processing system 20. The observation system 10 optically detects the state of a target gas. The information processing system 20 analyzes the results of detection by the observation system 10 to identify the concentration distribution D of the target gas within an observation space 200.

[0012] The observation system 10 comprises multiple observation cameras 11-1 to 11-3. Each observation camera 11-m (m = 1 to 3) is an imaging device that captures images of the observation space 200. Specifically, each observation camera 11-m comprises an optical system such as a photographing lens and an imaging element that receives incident light from the optical system. Each observation camera 11-m is capable of detecting incident light within a frequency band that includes the wavelength of the absorption line of the gas being observed. The optical axis Lm of each observation camera 11-m passes through the observation space 200. The direction of the optical axis Lm of each observation camera 11-m is different.

[0013] Mutually orthogonal X and Y axes are assumed to exist in observation space 200. An XY plane including the X and Y axes is, for example, a virtual plane parallel to the horizontal direction. Multiple observation cameras 11-1 to 11-3 are installed along the XY plane. Concentration distribution D identified by concentration distribution analysis system 100 is the distribution of the concentration of the observation target gas within the XY plane.

[0014] A reference point C is set on the XY plane. Reference point C is a specific point within observation space 200. For example, the center of observation space 200 is exemplified as reference point C. Multiple observation cameras 11-1 to 11-3 are installed on a virtual circle centered on reference point C. That is, the distance between each of observation cameras 11-1 to 11-3 and reference point C is equal. For example, the midpoint of the imaging surface of the imaging element mounted on each observation camera 11-m is located on the circumference centered on reference point C.

[0015] Specifically, observation cameras 11-m are installed at predetermined angles on a circle centered on reference point C. For example, three observation cameras 11-1 to 11-3 are installed at 90° intervals as shown in the example of FIG. 1. That is, observation camera 11-1 is installed in the range of negative numbers on the Y axis, and observation camera 11-3 is installed in the range of positive numbers on the Y axis. Furthermore, observation camera 11-2 is installed in the range of positive numbers on the X axis.

[0016] The optical axes Lm of the observation cameras 11-m intersect at reference point C. That is, the optical axes Lm of the observation cameras 11-m are located in the XY plane. As explained above, the multiple observation cameras 11-1 to 11-3 are installed on the XY plane. The plane on which the multiple observation cameras 11-1 to 11-3 are located (or the plane including each optical axis Lm) can also be interpreted as the XY plane. Furthermore, the space surrounded by the multiple observation cameras 11-1 to 11-3 in the flow path through which the target gas flows can also be interpreted as observation space 200.

[0017] FIG. 2 is an explanatory diagram of the range A that each observation camera 11-m can capture (hereinafter referred to as the "imaging range"). Each observation camera 11-m can capture an image of imaging range A at a predetermined angle including the optical axis Lm. In other words, imaging range A corresponds to the angle of view of observation camera 11-m. In the following explanation, for convenience, it is assumed that the angle of imaging range A is common to multiple observation cameras 11-1 to 11-3. However, the angle of imaging range A may differ for each observation camera 11-m.

[0018] The imaging range A includes range AC, range AR, and range AL. Range AC is a range of a predetermined angle located in the center of the imaging range A and including the optical axis Lm. Range AR is a range of a predetermined angle located to the right of range AC in the imaging range A. Range AL is a range of a predetermined angle located to the left of range AC in the imaging range A. For example, range AC is a range of 54°, and ranges AR and AL are each a range of 37°. Therefore, the angle of the imaging range A is 128°.

[0019] Each observation camera 11-m generates imaging data Gm representing the results of imaging the imaging range A. The imaging data Gm is data representing the intensity of incident light reaching each position on the imaging surface of the observation camera 11-m. Each observation camera 11-m is connected to the information processing system 20, for example, by wire or wirelessly. The imaging data Gm generated by each observation camera 11-m is supplied to the information processing system 20.

[0020] The information processing system 20 identifies the concentration distribution D of the target gas by analyzing the results of imaging by the multiple observation cameras 11-1 to 11-3. The information processing system 20 is realized by an information device such as a personal computer or a tablet terminal. Specifically, the information processing system 20 includes a control device 21, a storage device 22, a display device 23, and an operation device 24. The information processing system 20 may be realized as a single device, or may be realized as multiple devices configured separately from each other.

[0021] The control device 21 is composed of one or more processors that control each element of the information processing system 20. Specifically, the control device 21 is composed of one or more types of processors, such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), an FPGA (Field Programmable Gate Array), or an ASIC (Application Specific Integrated Circuit).

[0022] The storage device 22 is one or more memories that store programs executed by the control device 21 and data used by the control device 21. The storage device 22 is configured with a known recording medium such as a magnetic recording medium or a semiconductor recording medium. The storage device 22 may also be configured with a combination of multiple types of recording media.

[0023] The display device 23 displays an image under the control of the control device 21. Specifically, the display device 23 displays the concentration distribution D identified for the target gas to be observed. The operation device 24 is an input device that accepts operations from a user. Note that the display device 23 or the operation device 24, which are separate from the information processing system 20, may be connected to the information processing system 20 by wire or wirelessly.

[0024] 3 is a block diagram illustrating an example of the functional configuration of the information processing system 20. The control device 21 executes a program stored in the storage device 22 to realize multiple functions (an observation data generation unit 31, an analysis processing unit 32) for identifying the concentration distribution D of the observation target gas.

[0025] [Observation Data Generation Unit 31] Observation data generation unit 31 generates multiple observation data Pn (n=0°, 45°, 90°, 135°, 180°) based on the results of imaging by multiple observation cameras 11-1 to 11-3. Each observation data Pn corresponds to a different direction (i.e., angle n) with respect to observation space 200. Observation data generation unit 31 generates observation data Pn using imaging data Gm (G1 to G3) generated by each observation camera 11-m.

[0026] FIG. 4 is an explanatory diagram of observation data Pn. As illustrated in FIG. 4, a detection plane Q is assumed with respect to the observation space 200. The detection plane Q is a virtual plane set with respect to the observation space 200. The normal to the detection plane Q passes through the reference point C. The detection plane Q is rotatable with respect to the observation space 200 around an axis passing through the reference point C. The angle n is the angle of the normal to the detection plane Q with the reference point C as the center, and is defined with the Y-axis direction as 0°.

[0027] Observation data Pn is data representing the results of imaging detection plane Q at angle n relative to observation space 200 using observation system 10. Specifically, observation data Pn is composed of multiple observation values ​​p(r, n). Each observation value p(r, n) is the brightness value of the rth pixel among multiple pixels corresponding to different positions on detection plane Q at angle n. Observation data generation unit 31 uses imaging data Gm generated by each observation camera 11-m to generate multiple observation data Pn (P0, P45, P90, P135, P180) corresponding to different angles n. Note that the position (r, n) of the pixel corresponding to observation value p(r, n) corresponds to polar coordinates on the XY plane.

[0028] The size of the detection plane Q is set to a range onto which the entire observation space 200 can be projected. For example, when the angle n is 0° as shown in the example of FIG. 5, it is assumed that the width of the observation space 200 corresponds to 256 pixels. When the angle n is 45° as shown in the example of FIG. 4, a total of 362 (=256×√2) pixels are required to represent the entire observation space 200 using the observation data Pn. That is, the observation data Pn is composed of a total of 362 observation values ​​p(r,n). Specifically, the observation data Pn is composed of a total of 362 observation values ​​p(r,n), ranging from the observation value p(-53,n) corresponding to the left end of the detection plane Q to the observation value p(310,n) corresponding to the right end of the detection plane Q.

[0029] As shown in FIG. 4, the observation space 200 is partitioned into a plurality of regions U1 to U4. Each region Uk (k=1 to 4) is a square space when viewed from a direction perpendicular to the XY plane. The planar shape and size are common to the plurality of regions U1 to U4. The plurality of regions U1 to U4 are arranged in a matrix of 2 rows and 2 columns. Region U2 is located in the positive direction of the X axis when viewed from region U1, and region U3 is located in the negative direction of the Y axis when viewed from region U1. Region U4 is located in the negative direction of the Y axis when viewed from region U2 (the positive direction of the X axis when viewed from region U3).

[0030] 5 to 9 are explanatory diagrams relating to the generation of observation data Pn. Illustrated in Fig. 5 to 9 are explanatory diagrams of the angle of detection plane Q relative to observation space 200 and graphs of the absorbance of the target gas represented by observation data Pn. As described above, observation data Pn is data representing the observation values ​​p(r, n) of the target gas observed at different positions on detection plane Q. There is a correlation in which the higher the absorbance of the target gas, the smaller the observation value p(r, n).

[0031] In the following explanation, it is assumed for convenience that the concentration of the target gas in region U1 is 0.1, the concentration in region U2 is 0.2, the concentration in region U3 is 0.3, and the concentration in region U4 is 0.4. It is also assumed for convenience that the concentration of the target gas is uniform within each region Uk.

[0032] FIG. 5 is an explanatory diagram of observation data P0 when angle n is 0°. As illustrated in FIG. 5, observation data generation unit 31 uses the portion of imaging data G1 generated by observation camera 11-1 that corresponds to range AC of imaging range A as observation data P0. Ideally, observation system 10 would observe parallel light perpendicular to detection plane Q, but as illustrated in FIG. 5, observation camera 11-1 can actually detect incident light that converges from detection plane Q toward observation camera 11-3. Therefore, as can be seen from the graph in FIG. 5, the results of observations at positions on detection plane Q that are away from optical axis L1 of observation camera 11-1 contain errors from the true values ​​of the target gas.

[0033] Fig. 6 is an explanatory diagram of observation data P45 when angle n is 45°. As illustrated in Fig. 6, observation data generation unit 31 generates observation data P45 using imaging data G1 generated by observation camera 11-1 and imaging data G2 generated by observation camera 11-2.

[0034] Specifically, the observation data generation unit 31 employs a portion of the imaging data G1 corresponding to range AL as the observation value p(r, 45) corresponding to region QL (r = -53 to 127) in the left half of the detection plane Q in the observation data P45, and employs a portion of the imaging data G2 corresponding to range AR as the observation value p(r, 45) corresponding to region QR (r = 128 to 310) in the right half of the detection plane Q in the observation data P45. For example, the region QL in the observation data P45 is generated by extending the portion of the imaging data G1 corresponding to range AL, and the region QR in the observation data P45 is generated by extending the portion of the imaging data G2 corresponding to range AR. As explained above, the observation data P45 is generated by combining a portion of the imaging data G1 (region QL) and a portion of the imaging data G2 (region QR).

[0035] 7 is an explanatory diagram of observation data P90 when angle n is 90°. As illustrated in Fig. 7, observation data generation unit 31 uses, as observation data P90, a portion of imaging data G2 generated by observation camera 11-2 that corresponds to range AC of imaging range A.

[0036] FIG. 8 is an explanatory diagram of observation data P135 when angle n is 135°. As illustrated in FIG. 8, observation data generation unit 31 generates observation data P135 using imaging data G2 generated by observation camera 11-2 and imaging data G3 generated by observation camera 11-3. Specifically, observation data generation unit 31 employs a portion of imaging data G2 corresponding to range AL as observation value p(r, 135) corresponding to region QL (r = -53 to 127) in the left half of detection plane Q in observation data P135, and employs a portion of imaging data G3 corresponding to range AR as observation value p(r, 135) corresponding to region QR (r = 128 to 310) in the right half of detection plane Q in observation data P135. That is, observation data P135 is generated by combining a portion of imaging data G2 (range AL) and a portion of imaging data G3 (range AR).

[0037] Fig. 9 is an explanatory diagram of observation data P180 when angle n is 180°. As illustrated in Fig. 9, observation data generation unit 31 uses, as observation data P180, a portion of imaging data G3 generated by observation camera 11-3 that corresponds to range AC of imaging range A.

[0038] In the above description, for example, observation data P0 is data including the results of observation camera 11-1 capturing an image of area AC. On the other hand, observation data P45 is data including the results of observation camera 11-1 capturing an image of area AL. Also, for example, observation data P90 is data including the results of observation camera 11-2 capturing an image of area AC, and observation data P135 is data including the results of observation camera 11-2 capturing an image of area AL. That is, in the first embodiment, multiple pieces of observation data Pn are generated using different parts of imaging data Gm generated by observation camera 11-n.

[0039] First, we focus on an arbitrary first observation camera 11-m1 among the multiple observation cameras 11-m (m1 = 1 to 3). We also focus on a first range A1 and a second range A2 arbitrarily selected from the multiple ranges (AC, AR, AL) that make up the imaging range A. As can be seen from the above example, the multiple observation data Pn in the first embodiment include first observation data Pn1 containing the results of the first observation camera 11-m1 capturing the first range A1, and second observation data Pn2 containing the results of the first observation camera 11-m1 capturing the second range A2. Therefore, we can effectively identify the concentration distribution D of the target gas while limiting the total number of observation cameras 11-m capturing images of the observation space 200.

[0040] Furthermore, for example, observation data P45 is composed of a combination of the results of observation camera 11-1 capturing an image of range AL and the results of observation camera 11-2 capturing an image of range AR. Similarly, for example, observation data P135 is composed of a combination of the results of observation camera 11-2 capturing an image of range AL and the results of observation camera 11-3 capturing an image of range AR. In other words, a single piece of observation data Pn is generated using the results of multiple different observation cameras 11-n capturing images of different ranges within imaging range A.

[0041] Among the multiple observation cameras 11-m, attention is focused on the first observation camera 11-m1 and the second observation camera 11-m2 (m1, m2 = 1 to 3, m1 ≠ m2). As can be understood from the above example, the multiple observation data Pn in the first embodiment include observation data Pn that includes the results of the first observation camera 11-m1 capturing an image of the first range A1 and the results of the second observation camera 11-m2 capturing an image of the second range A2. Therefore, the concentration distribution D of the target gas can be effectively identified while limiting the total number of observation cameras 11-m capturing an image of the observation space 200.

[0042] [Analysis Processing Unit 32] 3 identifies the concentration distribution D of the target gas from the plurality of observation data Pn. Specifically, the analysis processing unit 32 identifies the concentration distribution D of the target gas by performing an inverse Radon transform on the plurality of observation data Pn. The inverse Radon transform is expressed by the following equation (1).

[0043]

number

[0044] The symbol d(x, y) in formula (1) is the density at the coordinates (x, y) on the XY plane. The function h(r, n) in formula (1) is a filter function, and for example, an impulse response is used. However, for simplicity in the first embodiment, the following formula (2) is used, in which the filter function h(r, n) is conveniently fixed to 1.

number

[0045] As can be seen from equation (2), the inverse Radon transform performed by the analysis processing unit 32 is a calculation process in which each observation value p(r, n) of the observation data Pn is integrated over all pixels (rmin to rmax) of the detection plane Q, and then integrated over all angles n (n = 0°, 45°, 90°, 135°, 180°).

[0046] Fig. 10 is a flowchart illustrating a specific procedure for the inverse Radon transform executed by the analysis processing unit 32. Fig. 11 is an explanatory diagram of the inverse Radon transform.

[0047] First, as illustrated in FIG. 11, analysis processing unit 32 arranges multiple pieces of observation data Pn (P0, P45, P90, P135, P180) corresponding to different angles n on the XY plane (S31). Analysis processing unit 32 also reserves storage capacity in storage device 22 for storing processing matrix E (S32). Processing matrix E is a square matrix with a size corresponding to the number of horizontal pixels of observation camera 11-n. For example, assuming the number of horizontal pixels of observation camera 11-n is 256, as illustrated in FIG. 11, storage capacity is reserved for processing matrix E of 256 rows by 256 columns. As illustrated in FIG. 11, each element e(x, y) of processing matrix E corresponds to the XY plane.

[0048] The analysis processing unit 32 selects one of the multiple observation data Pn corresponding to different angles n (hereinafter referred to as "selected observation data Pn") (S33). The analysis processing unit 32 sequentially selects each of the multiple elements e(x, y) of the processing matrix E, and updates the numerical value of the element e(x, y) stored in the storage device 22 according to the observation value p(r, n) of the selected observation data Pn (S34).

[0049] Specifically, the analysis processing unit 32 sequentially selects the processing matrix E row by row, and when any row is selected, sequentially selects the element e(x, y) corresponding to that row. The analysis processing unit 32 adds the observation value p(r, n) that is closest to the selected element e(x, y) in the XY plane among the multiple observation values ​​p(r, n) of the selected observation data Pn to that element e(x, y).

[0050] As mentioned above, the position (r, n) of the pixel corresponding to the observed value p(r, n) of the observation data Pn corresponds to polar coordinates on the XY plane. Therefore, the distance δ between the coordinates (x, y) corresponding to the element e(x, y) of the processing matrix E and the pixel (r, n) corresponding to the observed value p(r, n) is expressed by the following formula (3).

number

[0051] Of the multiple observation values ​​p(r,n) of the selected observation data Pn, the observation value p(r,n) corresponding to the pixel for which the distance δ in equation (3) is the smallest is added to the element e(x,y) stored in the storage device 22. The above process is performed for all elements e(x,y) of the processing matrix E. The update of the element e(x,y) by adding the observation value p(r,n) corresponds to the process of integrating each observation value p(r,n) of the observation data Pn over all pixels (rmin to rmax) of the detection plane Q in equation (2) above.

[0052] The analysis processing unit 32 determines whether the processing matrix E has been updated (S34) for all observation data Pn (S35). If there is unselected observation data Pn (S35: NO), the analysis processing unit 32 selects observation data Pn other than the currently selected observation data Pn as new selected observation data Pn (S33), and updates the processing matrix E using the selected observation data Pn (S34). Updating the processing matrix E for multiple observation data Pn corresponds to the process of integrating each observation value p(r,n) of the observation data Pn over all angles n (n=0°, 45°, 90°, 135°, 180°) in equation (2) above.

[0053] When the processing matrix E has been updated for all of the observation data Pn (S35: YES), the analysis processing unit 32 ends the processing (inverse Radon transform) of Fig. 10. The processing matrix E at the time of completing the processing of Fig. 10 is determined as the concentration distribution D of the observation target gas. That is, the element e(x, y) in the updated processing matrix E corresponds to the concentration d(x, y) in the concentration distribution D of Equation (1).

[0054] FIG. 12 is a specific example of concentration distribution D. Image 1 in FIG. 12 is an image in which each region Uk in the observation space 200 is illustrated with a gradation corresponding to the concentration of the gas being observed. Image 2 in FIG. 12 is concentration distribution D calculated by inverse Radon transform of multiple observation data Pn. Image 3 in FIG. 12 is the result of processing the concentration distribution D in image 2.

[0055] The processing is, for example, a process of replacing each density d(x, y) in each of the multiple regions U1 to U4 in the observation space 200 with a representative value of the density d(x, y) in the region Uk. The representative value of the density d(x, y) is, for example, the average value of all the density d(x, y) in the region Uk. Therefore, the density d(x, y) is set to a uniform value within the region Uk.

[0056] The intensity I(L) of light propagating through the target gas is expressed by the following equation (4) according to the Beer-Lambert law:

number

[0057] The above constraints pose a problem when detecting potentially toxic or flammable gases. For example, consider a gas for which the exposure limit for living organisms is 100 ppm. Let's assume that the product of the gas concentration and the thickness of the target gas is calculated to be 100 ppm m using equation (4). In this case, if the propagation length L is unknown, it is impossible to determine whether the measured value is, for example, 1000 ppm × 0.1 m or 10 ppm × 10 m. In other words, it is impossible to determine whether the target gas concentration is above or below the exposure limit.

[0058] In contrast to the conventional technology described above, in the first embodiment, multiple pieces of observation data Pn corresponding to different directions (angles n) relative to observation space 200 are acquired based on the imaging results of multiple observation cameras 11-1 to 11-3, and the concentration distribution D of the target gas is determined by performing an inverse Radon transform on the multiple pieces of observation data Pn. Therefore, as can be seen from the comparison of images 1 to 3, according to the first embodiment, the concentration distribution D of the target gas can be effectively determined even in situations where the thickness (propagation length L) of the target gas is unknown. Therefore, the first embodiment is particularly suitable for detecting target gases that are dangerous to living organisms, for example.

[0059] 13 is a flowchart illustrating a specific procedure of a process (hereinafter referred to as "concentration distribution analysis process") executed by the information processing system 20. For example, the concentration distribution analysis process is started in response to an instruction from a user via the operation device 24. The concentration distribution analysis process is an example of a "concentration distribution analysis method."

[0060] When the concentration distribution analysis process starts, control device 21 (observation data generation unit 31) acquires imaging data Gm generated by each observation camera 11-m of observation system 10 (S1). Control device 21 (observation data generation unit 31) uses multiple imaging data G1 to G3 to generate multiple observation data Pn (P0, P45, P90, P135, P180) corresponding to different directions (S2). The generation of observation data Pn is as described above with reference to Figures 5 to 9.

[0061] The control device 21 (analysis processing unit 32) identifies the concentration distribution D of the observation target gas by performing an inverse Radon transform on the multiple observation data Pn (S3). The specific steps of the inverse Radon transform (S31 to S35) are as described above with reference to FIGS. 10 and 11.

[0062] The control device 21 (analysis processing unit 32) performs a processing process on the concentration distribution D (S4). As described above, the processing process is a process of replacing the concentration d(x, y) in each region Uk with a representative value of the concentration d(x, y) within that region Uk. The control device 21 displays the concentration distribution D after the processing process on the display device 23 (S5). Note that the processing process may be omitted.

[0063] The control device 21 determines whether a predetermined termination condition is met (S6). The termination condition may be, for example, that an instruction to terminate the concentration distribution analysis process is issued via an operation on the operation device 24, or that a predetermined time has elapsed since the start of the concentration distribution analysis process. If the termination condition is not met (S6: NO), the control device 21 transitions the process to step S1. That is, until the termination condition is met, the control device 21 repeats the acquisition of imaging data Gm (S1), the generation of observation data Pn (S2), the identification of concentration distribution D (S3), the processing process (S4), and the display of the analysis results (S5). If the termination condition is met (S6: YES), the control device 21 terminates the concentration distribution analysis process.

[0064] B: Second embodiment A second embodiment of the present disclosure will be described. Note that, for elements in the following exemplary aspects that have the same functions as those in the first embodiment, the same reference numerals as those in the first embodiment will be used, and detailed descriptions of each will be omitted as appropriate.

[0065] 14 is a block diagram illustrating the configuration of a concentration distribution analysis system 100 according to the second embodiment. The concentration distribution analysis system 100 of the second embodiment differs from that of the first embodiment in the configuration of the observation system 10. The observation system 10 of the second embodiment includes observation cameras 11-4 and 11-5 in addition to observation cameras 11-1 to 11-3 similar to those of the first embodiment.

[0066] Multiple observation cameras 11-1 to 11-5 are installed at 45° intervals on a circle centered on reference point C. Specifically, observation camera 11-4 is installed between observation cameras 11-1 and 11-2, and observation camera 11-5 is installed between observation cameras 11-2 and 11-3. As in the first embodiment, the optical axes Lm of each observation camera 11-m (m = 1 to 5) intersect at reference point C. Each observation camera 11-m generates imaging data Gm representing the results of imaging imaging range A.

[0067] In the second embodiment, the method by which observation data generation unit 31 generates observation data P45 and observation data P135 differs from that in the first embodiment. Specifically, observation data P45 is generated using imaging data G4 generated by observation camera 11-4, and observation data P135 is generated using imaging data G5 generated by observation camera 11-5. The generation of the other observation data Pn (P0, P90, P180) is the same as in the first embodiment.

[0068] Fig. 15 is an explanatory diagram of observation data P45 in the second embodiment. Observation data generation unit 31 in the first embodiment described above generates observation data P45 using imaging data G1 generated by observation camera 11-1 and imaging data G2 generated by observation camera 11-2. As illustrated in Fig. 15, observation data generation unit 31 in the second embodiment generates observation data P45 using imaging data G4 generated by observation camera 11-4. Specifically, observation data generation unit 31 uses the portion of imaging data G4 generated by observation camera 11-4 that corresponds to range AC of imaging range A as observation data P45.

[0069] Fig. 16 is an explanatory diagram of observation data P135 in the second embodiment. Observation data generation unit 31 in the first embodiment described above generates observation data P135 using imaging data G2 generated by observation camera 11-2 and imaging data G3 generated by observation camera 11-3. As illustrated in Fig. 16, observation data generation unit 31 in the second embodiment generates observation data P135 using imaging data G5 generated by observation camera 11-5. Specifically, observation data generation unit 31 uses the portion of imaging data G5 generated by observation camera 11-5 that corresponds to range AC of imaging range A as observation data P135.

[0070] The second embodiment also achieves the same effects as the first embodiment. Furthermore, since the second embodiment uses five observation cameras 11-1 to 11-5, the concentration distribution D of the target gas can be determined with higher accuracy than the first embodiment. As can be seen from the example of the second embodiment, the configuration (FIGS. 6 and 8) in which a plurality of different imaging data Gm are used to generate observation data Pn corresponding to one angle n is not essential to the present disclosure.

[0071] C: Modified Example Specific modified embodiments that can be added to each of the embodiments exemplified above are exemplified below. Two or more embodiments arbitrarily selected from the following examples may be appropriately combined within the scope of not being mutually contradictory.

[0072] (1) The total number of observation cameras 11-m used to capture images of the observation space 200 is arbitrary. For example, the observation system 10 may be composed of four observation cameras 11-m or six or more observation cameras 11-m. Alternatively, the observation system 10 may be composed of two observation cameras 11-m. However, when three or more observation cameras 11-m are used to generate the multiple observation data Pn, the concentration distribution D of the observation target gas can be determined with higher accuracy than when two observation cameras 11-m are used.

[0073] (2) As described above, the functions of the information processing system 20 according to each of the above embodiments are realized through cooperation between one or more processors constituting the control device 21 and a program stored in the storage device 22. The programs exemplified above can be provided in a form stored on a computer-readable recording medium and installed on a computer. The recording medium is, for example, a non-transitory recording medium, such as an optical recording medium (optical disk) such as a CD-ROM, but also includes any known type of recording medium, such as a semiconductor recording medium or a magnetic recording medium. Note that a non-transitory recording medium includes any recording medium other than a transitory, propagating signal, and does not exclude volatile recording media. Furthermore, in a configuration in which a distribution device distributes a program via a communication network, the recording medium storing the program in the distribution device corresponds to the non-transitory recording medium described above.

[0074] (3) The term "nth" (n is a natural number) in this application is used only as a formal and convenient label to distinguish each element in the description and does not have any substantive meaning. Therefore, there is no room for restrictive interpretation of the position or order of each element based on the term "nth."

[0075] D: Notes From the above-described exemplary embodiments, the following configurations can be understood, for example.

[0076] A concentration distribution analysis system according to one aspect (aspect 1) of the present disclosure is a concentration distribution analysis system for determining the concentration distribution of a target gas present in an observation space, and includes: a plurality of observation cameras with different optical axis directions passing through the observation space; an observation data generation unit that acquires a plurality of observation data corresponding to different directions relative to the observation space based on the results of imaging by the plurality of observation cameras; and an analysis processing unit that determines the concentration distribution of the target gas by performing an inverse Radon transform on the plurality of observation data. In the above aspect, a plurality of observation data corresponding to different directions relative to the observation space is acquired based on the results of imaging by the plurality of observation cameras, and the concentration distribution of the target gas is determined by performing an inverse Radon transform on the plurality of observation data. Therefore, the concentration distribution of the target gas can be effectively determined even in a situation where the thickness of the target gas is unknown.

[0077] In a specific example (Aspect 2) of Aspect 1, the plurality of observation cameras are three or more observation cameras installed on a circle centered on a reference point within the observation space. In this aspect, three or more observation cameras are used to generate a plurality of observation data. Therefore, compared to an aspect in which two observation cameras are used to generate a plurality of observation data, the concentration distribution of the target gas can be determined with higher accuracy.

[0078] In a specific example (Aspect 3) of Aspect 1 or Aspect 2, each of the plurality of observation cameras captures an imaging range including a first range and a second range, and the plurality of observation data includes first observation data including the results of a first observation camera of the plurality of observation cameras capturing the first range, and second observation data including the results of the first observation camera capturing the second range. In the above aspect, the plurality of observation data used to identify the concentration distribution includes first observation data including the results of a first observation camera capturing the first range, and second observation data including the results of the first observation camera capturing the second range. Therefore, the concentration distribution of the target gas can be effectively identified while reducing the total number of observation cameras capturing the observation space.

[0079] In a specific example (Aspect 4) of Aspect 1 or Aspect 2, each of the plurality of observation cameras captures an image of an imaging range including a first range and a second range, and the plurality of observation data includes observation data including a result of imaging the first range by a first observation camera among the plurality of observation cameras and a result of imaging the second range by a second observation camera among the plurality of observation cameras that is different from the first observation camera. In the above aspect, the plurality of observation data used to identify the concentration distribution includes observation data including a result of imaging the first range by the first observation camera and a result of imaging the second range by the second observation camera. Therefore, the concentration distribution of the target gas can be effectively identified while reducing the total number of observation cameras capturing the observation space.

[0080] An information processing system according to one aspect (aspect 5) of the present disclosure is an information processing system that uses multiple observation cameras with different optical axis directions passing through the observation space to identify the concentration distribution of a target gas present in the observation space, and includes an observation data generation unit that acquires multiple observation data corresponding to different directions relative to the observation space based on the results of imaging by the multiple observation cameras, and an analysis processing unit that identifies the concentration distribution of the target gas by performing an inverse Radon transform on the multiple observation data.

[0081] A concentration distribution analysis method according to one aspect (aspect 6) of the present disclosure is a concentration distribution analysis method implemented by a computer system that uses multiple observation cameras with different optical axis directions passing through the observation space to identify the concentration distribution of a target gas present in the observation space, and includes obtaining multiple observation data corresponding to different directions relative to the observation space based on the results of imaging by the multiple observation cameras, and identifying the concentration distribution of the target gas by performing an inverse Radon transform on the multiple observation data. [Explanation of symbols]

[0082] 100...concentration distribution analysis system, 200...observation space, 10...observation system, 11-m (11-1, 11-2, 11-3, 11-4, 11-5)...observation camera, 20...information processing system, 21...control device, 22...storage device, 23...display device, 24...operation device, 31...observation data generation unit, 32...analysis processing unit.

Claims

1. A concentration distribution analysis system that identifies the concentration distribution of an observation target gas present in an observation space, a plurality of observation cameras whose optical axes pass through the observation space in different directions; an observation data generation unit that acquires a plurality of observation data corresponding to different directions with respect to the observation space in accordance with the results of imaging by the plurality of observation cameras; an analysis processing unit that identifies a concentration distribution of the target gas by performing an inverse Radon transform on the plurality of observation data; A concentration distribution analysis system comprising:

2. The plurality of observation cameras are three or more observation cameras installed on a circumference of a circle centered on a reference point in the observation space. The concentration distribution analysis system according to claim 1 .

3. each of the plurality of observation cameras captures an image of an image capturing range including a first range and a second range; The plurality of observation data First observation data including a result of an image of the first range captured by a first observation camera among the plurality of observation cameras; and second observation data including the result of the first observation camera capturing the second range.

3. The concentration distribution analysis system according to claim 1 or 2.

4. each of the plurality of observation cameras captures an image of an image capturing range including a first range and a second range; The plurality of observation data The observation data includes a result of imaging the first range by a first observation camera among the plurality of observation cameras and a result of imaging the second range by a second observation camera among the plurality of observation cameras that is different from the first observation camera.

3. The concentration distribution analysis system according to claim 1 or 2.

5. An information processing system that identifies a concentration distribution of an observation target gas present in an observation space by using a plurality of observation cameras having optical axes passing through the observation space in different directions, an observation data generation unit that acquires a plurality of observation data corresponding to different directions with respect to the observation space in accordance with the results of imaging by the plurality of observation cameras; an analysis processing unit that identifies a concentration distribution of the target gas by performing an inverse Radon transform on the plurality of observation data; An information processing system comprising:

6. Identifying the concentration distribution of the target gas present in the observation space by using a plurality of observation cameras each having an optical axis passing through the observation space in different directions. A concentration distribution analysis method implemented by a computer system, comprising: acquiring a plurality of observation data corresponding to different directions with respect to the observation space according to the results of imaging by the plurality of observation cameras; Identifying a concentration distribution of the target gas by performing an inverse Radon transform on the plurality of observation data. A concentration distribution analysis method including:

Images