Leakage source position calculation device, leakage source position calculation system, program, and method for calculating position of leakage source

Abstract

To provide a leakage source position calculation device which can acquire the directions of flows of a plurality of leaking fluids at different times and can estimate the positions of leakage by a method for determination which is different according to whether the direction of flow is substantially the same.SOLUTION: A leakage source position calculation device 3 includes: a leakage direction calculation unit 32 for acquiring the direction of leakage of at least two leaking fluids at different times; a leakage source candidate calculation unit 33 for acquiring a leakage source candidate position from each of the directions of leakage at different times; and a leakage source determination unit 34 for determining the position of a leakage source by a method for determination according to whether an angle formed by the plurality of leakage directions at different times is smaller than a threshold value or is at least the threshold value.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a leakage source position calculation device, a leakage source position calculation system, a program, and a leakage source position calculation method for determining the position of a leakage source such as gas from time-series images. [Background technology]

[0002] The following techniques have been proposed as techniques for determining the location of the leakage source of a leaking fluid such as gas. For example, Patent Document 1 proposes a technique for tracing the direction in which the fluid moves for each pixel of a fluid image and calculating the leakage source location of the leaking fluid. Patent Document 2 proposes a technique for determining multiple movement trajectories of the leaking fluid at the same time and calculating the leakage source location of the leaking fluid from the intersection of the trajectories. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. WO2020 / 110410 [Patent Document 2] International Publication No. WO2016 / 208317 Summary of the Invention [Problem to be solved by the invention]

[0004] In the method of Patent Document 1, if the leaking fluid or the leak source is hidden by an obstruction, the leak source cannot be traced back to the source. In this case, the location where the leaking fluid begins to be visible from behind the obstruction is assumed to be the leak source, resulting in a large error between the actual leak source position and the estimated leak source position.

[0005] In addition, in the method of Patent Document 2, when the angle between the multiple movement directions of the leaking fluid in the same direction is small, the intersection point is often connected at a position far away from the actual leakage source position, which increases the error between the actual leakage source position and the estimated leakage source position, making it difficult to estimate with high accuracy.

[0006] The present invention has been made to solve such problems, and aims to provide a leakage source position calculation device, leakage source position calculation system, program, and leakage source position calculation method that acquires the flow directions of multiple leakage fluids at different times and enables the leakage source position to be estimated using different determination methods depending on whether the flow directions are approximately the same. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, the present invention provides a leakage source position calculation device that includes a leakage direction acquisition unit that acquires leakage directions of at least two or more leaking fluids at different times, a candidate leakage source position acquisition unit that acquires candidate leakage source positions from each of the multiple leakage directions at different times, and a leakage source determination unit that determines the leakage source position using a determination method depending on whether the angle formed by the multiple leakage directions at different times is less than a threshold value or greater than a threshold value.

[0008] The present invention also provides a leak source position calculation system that includes an image acquisition unit that acquires images of a subject including a monitored object, and a leak source position calculation device that determines the leak source position from time-series images acquired by the image acquisition unit, wherein the leak source position calculation device includes a leak direction acquisition unit that acquires leak directions of at least two or more leaking fluids at different times, a candidate leak source position acquisition unit that acquires candidate leak source positions from the multiple leak directions at different times, and a leak source determination unit that determines the leak source position using a determination method depending on whether the angle formed by the multiple leak directions at different times is less than a threshold value or greater than a threshold value.

[0009] Furthermore, the present invention is a program executed by a calculation processing unit of a leakage source position calculation device that determines a leakage source position from time-series images, the program causing the calculation processing unit to execute the steps of acquiring leakage directions of at least two or more leaking fluids at different times, acquiring candidate leakage source positions from the multiple leakage directions at different times, and determining the leakage source position using a determination method depending on whether the angle formed by the multiple leakage directions at different times is less than a threshold value or greater than a threshold value.

[0010] The present invention also provides a leakage source position calculation method that executes the steps of acquiring leakage directions of at least two or more leaking fluids at different times, acquiring candidate leakage source positions from each of the multiple leakage directions at different times, and determining the leakage source position using a determination method depending on whether the angle formed by the multiple leakage directions at different times is less than a threshold value or greater than a threshold value. Effect of the Invention

[0011] According to the present invention, the leak source position can be determined by a determination method according to whether an angle between a plurality of leak directions at different times is less than a threshold value or is equal to or greater than a threshold value, thereby making it possible to determine the leak source position by a determination method that can suppress an increase in the error between the actual leak source position and the estimated leak source position. [Brief description of the drawings]

[0012] [Figure 1] 1 is a block diagram showing a configuration of a leakage source location calculation system according to an embodiment of the present invention; [Diagram 2] 4 is a flowchart showing an example of a leakage source position calculation method according to the present embodiment. [Figure 3A] FIG. 1 is an explanatory diagram showing an example of time-series infrared images. [Figure 3B] FIG. 1 is an explanatory diagram showing an example of time-series infrared images. [Figure 3C] FIG. 1 is an explanatory diagram showing an example of time-series infrared images. [Figure 3D] FIG. 1 is an explanatory diagram showing an example of time-series infrared images. [Figure 4A] FIG. 11 is an explanatory diagram showing an example of an image from which a gas signal is extracted. [Figure 4B] FIG. 11 is an explanatory diagram showing an example of an image from which a gas signal is extracted. [Figure 4C] FIG. 11 is an explanatory diagram showing an example of an image from which a gas signal is extracted. [Figure 4D] FIG. 11 is an explanatory diagram showing an example of an image from which a gas signal is extracted. [Figure 5A] FIG. 13 is an explanatory diagram showing an example of an image in which a gas signal has a maximum value. [Figure 5B] FIG. 2 is an explanatory diagram showing an example of a gas cloud region. [Figure 6] 6 is a time chart showing an example of timing for acquiring a gas leakage direction and a potential leakage source position. [Figure 7] FIG. 11 is an explanatory diagram showing the relationship between the angle between a plurality of leakage directions and an error. [Figure 8] 11 is an explanatory diagram showing an example of a method for determining a leakage source when a plurality of leakage directions are substantially the same; FIG. [Figure 9A] 11 is an explanatory diagram showing an example of a method for determining a leakage source when a plurality of leakage directions are not substantially the same; FIG. [Figure 9B] 11 is an explanatory diagram showing an example of a method for determining a leakage source when a plurality of leakage directions are not substantially the same; FIG. [Figure 10A] 11 is an explanatory diagram showing an example of a method for determining a leakage source when a plurality of leakage directions are not substantially the same; FIG. [Figure 10B] 11 is an explanatory diagram showing an example of a method for determining a leakage source when a plurality of leakage directions are not substantially the same; FIG. [Figure 11] 11A and 11B are explanatory diagrams showing another example of a method for determining a leak direction, a leak source candidate position, and a leak source. [Figure 12] 11A and 11B are explanatory diagrams showing another example of a method for determining a leak direction, a leak source candidate position, and a leak source. [Figure 13A] 11A and 11B are explanatory diagrams showing another example of a method for determining a leak direction, a leak source candidate position, and a leak source. [Figure 13B] 11A and 11B are explanatory diagrams showing another example of a method for determining a leak direction, a leak source candidate position, and a leak source. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, with reference to the drawings, a leakage source position calculation device of the present invention and a leakage source position calculation system including the leakage source position calculation device will be described. Also, an embodiment of a program executed by a processing unit of the leakage source position calculation device and a leakage source position calculation method will be described.

[0014] <Configuration Examples of Leak Source Location Calculation Device, Leak Source Location Calculation System, and Program> FIG. 1 is a block diagram showing a configuration of a leakage source position calculation system according to this embodiment. The leakage source position calculation system 1 includes an image acquisition unit 2, a leakage source position calculation device 3, and an output unit 5. The leakage source position calculation device 3 may include both or either of the image acquisition unit 2 and the output unit 5. The leakage source position calculation system 1 may also include an input unit. The input unit may be a keyboard, a touch panel, a mouse, or the like. The leakage source position calculation device 3 may also include the input unit.

[0015] The image acquisition unit 2 is an infrared camera equipped with an optical system, a filter, an image sensor, and a signal processing unit (not shown). The image acquisition unit 2 is configured to pass only infrared light of a specific wavelength corresponding to the monitoring target through a filter and receive it with the image sensor. The image acquisition unit 2 captures time-series infrared images, videos of infrared images, etc. of subjects including the monitoring target, and generates image data of a digital signal. The monitoring target is a leaking fluid, for example, a point where gas transport pipes are connected to each other.

[0016] The leakage source location calculation device 3 is an information processing device such as a personal computer, a smartphone, a tablet terminal, etc. The leakage source location calculation device 3 includes an arithmetic processing unit 30 and a storage unit 4.

[0017] The calculation processing unit 30 is composed of a CPU (Central Processing Unit) etc. The calculation processing unit 30 includes, as functional blocks, a fluid region extraction unit 31, a leakage direction calculation unit 32, a leakage source candidate calculation unit 33, and a leakage source determination unit .

[0018] The fluid region extraction unit 31 acquires a gas region from an image acquired by the image acquisition unit 2. The gas region is an example of a fluid region. Gas is also an example of a leaking fluid. The fluid region extraction unit 31 acquires the gas region two or more times at different times. In this way, the fluid region extraction unit 31 acquires two or more gas regions. The fluid region extraction unit 31 is realized by a program. The program that realizes the fluid region extraction unit 31 is executed by the calculation processing unit 30.

[0019] The leak direction calculation unit 32 is an example of a leak direction acquisition unit, and acquires the gas leak direction from the gas region acquired by the fluid region extraction unit 31. The leak direction calculation unit 32 acquires the gas leak direction from each of multiple gas regions acquired at different times. In this way, the leak direction calculation unit 32 acquires the leak directions of multiple gases. The leak direction calculation unit 32 is realized by a program. The program that realizes the leak direction calculation unit 32 is executed by the calculation processing unit 30.

[0020] The leak source candidate calculation unit 33 is an example of a leak source candidate position acquisition unit, and acquires the gas leak source candidate position from the gas leak direction acquired by the leak direction calculation unit 32. The leak source candidate calculation unit 33 acquires the leak source candidate position of each gas from the leak directions of multiple gases acquired at different times. In this way, the leak source candidate calculation unit 33 acquires the leak source candidate positions of multiple gases. The leak source candidate calculation unit 33 is realized by a program. The program that realizes the leak source candidate calculation unit 33 is executed by the arithmetic processing unit 30.

[0021] The leak source determination unit 34 switches the leak source determination method depending on the angle between the leak directions of the multiple gases. The leak source determination unit 34 switches the leak source determination method depending on whether the angle between the leak directions of the multiple gases is less than a predetermined threshold value or is equal to or greater than a threshold value. The leak source determination unit 34 is realized by a program. The program that realizes the leak source determination unit 34 is executed by the calculation processing unit 30.

[0022] The storage unit 4 is composed of a RAM (Random Access Memory), a ROM (Read Only Memory), etc. The storage unit 4 may also include a HDD (Hard Disk Drive), etc. The storage unit 4 includes a leak direction storage unit 40 and a leak source candidate storage unit 41. The leak direction storage unit 40 stores the leak directions of multiple gases. The leak source candidate storage unit 41 stores the leak source candidate positions of multiple gases. In addition, when the fluid region extraction unit 31, the leak direction calculation unit 32, the leak source candidate calculation unit 33, and the leak source determination unit 34 are realized by programs, the storage unit 4 stores these programs.

[0023] The output unit 5 is configured with a display etc. The output unit 5 is, for example, a liquid crystal display. Moreover, the output unit 5 may be an organic EL display (Organic Light Emitting Diode display), a plasma display, etc., instead of a liquid crystal display.

[0024] Programs for implementing the fluid region extraction unit 31, the leakage direction calculation unit 32, the leakage source candidate calculation unit 33, and the leakage source determination unit 34 are stored in advance in the storage unit 4, but are not limited thereto. For example, a storage medium storing these programs may be prepared, and the programs stored in this storage medium may be stored in the storage unit 4. The storage medium storing these programs may be, for example, an external storage medium such as a magnetic disk, an optical disk, or an SSD (Solid State Drive). In addition, these programs may be stored in a server connected to the leakage source position calculation device 3 via a wired or wireless network. These programs are sent from the server to the leakage source position calculation device 3 via the network, and stored in the storage unit 4.

[0025] The calculation processing unit 30 reads out and executes these programs from the storage unit 4. In this way, a fluid region extraction unit 31, a leakage direction calculation unit 32, a leakage source candidate calculation unit 33, and a leakage source determination unit 34 are realized.

[0026] However, these functional blocks may be realized by processing by a DSP (Digital Signal Processor) instead of processing by the arithmetic processing unit 30. Some of these functional blocks may be realized by processing by a DSP, or all of them may be realized by processing by a DSP. Furthermore, these functional blocks may be realized by processing by a DSP together with processing by the arithmetic processing unit 30. Similarly, some or all of each functional block may be realized by processing by a dedicated hardware circuit instead of or together with processing by software.

[0027] <Examples of operation of the leakage source location calculation device, leakage source location calculation system, and program> 2 is a flowchart showing an example of a leakage source position calculation method according to the present embodiment. Next, as the leakage source position calculation method, an operation example of the leakage source position calculation system 1, the leakage source position calculation device 3, and a program executed by the leakage source position calculation device 3 will be described.

[0028] The calculation processing unit 30 reads out from the storage unit 4 and executes programs that realize a fluid region extraction unit 31, a leakage direction calculation unit 32, a leakage source candidate calculation unit 33, and a leakage source determination unit .

[0029] The fluid region extraction unit 31 acquires a gas region from the image acquired by the image acquisition unit 2 in step SA1 of FIG.

[0030] 2, the leak direction calculation unit 32 acquires the gas leak direction from the gas region acquired by the fluid region extraction unit 31. The arithmetic processing unit 30 stores the gas leak direction acquired by the leak direction calculation unit 32 in the leak direction storage unit 40 of the storage unit 4 in step SA3 of FIG.

[0031] The leak source candidate calculation unit 33 acquires the gas leak source candidate positions from the gas leak direction acquired by the leak direction calculation unit 32 in step SA4 of Fig. 2. The arithmetic processing unit 30 stores the gas leak source candidate positions acquired by the leak source candidate calculation unit 33 in the leak source candidate storage unit 41 of the storage unit 4 in step SA5 of Fig. 2.

[0032] The calculation processing unit 30 judges whether the above-mentioned processing of steps SA1 to SA5 for acquiring the gas leakage direction and the leakage source candidate position has been performed a predetermined number of times in step SA6 of Fig. 2. If the calculation processing unit 30 judges that the processing for acquiring the gas leakage direction and the leakage source candidate position has not been performed a predetermined number of times, the calculation processing unit 30 repeats the processing of steps SA1 to SA5 a predetermined number of times.

[0033] The leak source determination unit 34 judges whether the angle between the leak directions of the multiple gases is less than a predetermined threshold value or is equal to or greater than the threshold value in step SA7 of Fig. 2. If the angle between the leak directions of the multiple gases is less than the predetermined threshold value, the leak source determination unit 34 determines a leak source candidate position from among the multiple leak source candidate positions in step SA8 of Fig. 2. If the angle between the leak directions of the multiple gases is equal to or greater than the threshold value, the leak source determination unit 34 determines a leak source candidate position from the multiple leak directions in step SA9 of Fig. 2.

[0034] The operation will be described in detail below. The image acquisition unit 2 acquires a plurality of infrared images MD(1) to MD(4) in time series. Fig. 3A, Fig. 3B, Fig. 3C, and Fig. 3D are explanatory diagrams showing an example of infrared images in time series. The infrared images in Fig. 3A, Fig. 3B, Fig. 3C, and Fig. 3D include the gas region to be monitored. The infrared image MD(1) is the first infrared image at the start of imaging. The infrared image MD(2) is the infrared image acquired 0.2 seconds later, the infrared image MD(3) is the infrared image acquired 0.4 seconds later, and the infrared image MD(4) is the infrared image acquired 0.6 seconds later. The number of infrared images acquired and the time intervals are examples.

[0035] These infrared images are subjected to image processing, for example, as described in Japanese Patent No. 6245418. In the image processing described in the publication, changes in gas as it sways in the wind are captured. As a result, images MD(10) to MD(40) from which gas signals have been extracted are obtained. FIGS. 4A, 4B, 4C, and 4D are explanatory diagrams showing examples of images from which gas signals have been extracted. Image MD(10) is an image from which a gas signal has been extracted from infrared image MD(1). Image MD(20) is an image from which a gas signal has been extracted from infrared image MD(2). Image MD(30) is an image from which a gas signal has been extracted from infrared image MD(3). Image MD(40) is an image from which a gas signal has been extracted from infrared image MD(4).

[0036] Next, the fluid region extraction unit 31 acquires a gas region from the image from which the gas signal is extracted. The fluid region extraction unit 31 acquires the maximum value of the gas signal from the image for n seconds from which the gas signal is extracted, for example, from the image for 1 second. The fluid region extraction unit 31 acquires the gas region CL by performing noise removal and threshold processing on the image MD(max) in which the gas signal is at its maximum value using a known method. FIG. 5A is an explanatory diagram showing an example of an image in which the gas signal is at its maximum value, and FIG. 5B is an explanatory diagram showing an example of a gas region. The method for acquiring the gas signal and the method for acquiring the gas region are merely examples, and other methods may be used. The gas region may also be a region designated by the user.

[0037] The leak direction calculation unit 32 acquires the gas leak direction from the gas region CL acquired by the fluid region extraction unit 31, for example, by the following method. The leak direction calculation unit 32 obtains the optical flow for the pixels in the gas region CL. The optical flow is calculated using a known technique such as the gradient method or the Lucas-Kanade method. The leak direction calculation unit 32 calculates the average value of the optical flow of the pixels in the gas region CL. The leak direction calculation unit 32 acquires the average value of the optical flow of the pixels in the gas region CL as the leak direction of the entire gas region CL. The calculation processing unit 30 records the gas leak direction acquired by the leak direction calculation unit 32 in the leak direction memory unit 40 of the memory unit 4.

[0038] The method for calculating the gas leakage direction is not limited to optical flow. For example, the gas leakage direction may be calculated by template matching such as Sum of Absolute Difference (SAD) or Normalized Cross-Correlation (NCC).

[0039] The leak source candidate calculation unit 33 acquires the gas leak source candidate position from the gas leak direction acquired by the leak direction calculation unit 32, for example, by the following method. The leak source candidate calculation unit 33 calculates the leak source candidate position by tracing the pixel movement vector, as in the method of WO2020 / 110410. The arithmetic processing unit 30 records the gas leak source candidate position acquired by the leak source candidate calculation unit 33 in the leak source candidate storage unit 41 of the storage unit 4.

[0040] The gas leakage source candidate position may be calculated using the methods of JP 2002-296142, JP 5861416, and JP 2016-114500. The gas leakage source candidate position may be calculated using the methods of WO2016 / 143754, WO2016 / 208317, WO2017 / 150565, and WO2021 / 246130.

[0041] The calculation processing unit 30 repeats the above-mentioned operations from extraction of the gas region to calculation of the leak source candidate position at least two times at different times. In this example, the operations from extraction of the gas region to calculation of the leak source candidate position are repeated 12 times at different times to obtain the leak directions and leak source candidate positions of multiple gases.

[0042] FIG. 6 is a time chart showing an example of the timing for acquiring the gas leakage direction and the leakage source candidate position.

[0043] The calculation processing unit 30 repeats the process of steps SA1 to SA5 in the flowchart of Fig. 2 every 10 seconds for, for example, two minutes. As a result, the calculation processing unit 30 acquires the gas leakage direction and the leakage source candidate position 12 times and records them in the memory unit 4. Then, the calculation processing unit 30 proceeds to the process of step SA7 in Fig. 2. The calculation processing unit 30 judges whether a plurality of gas leakage directions among the 12 times of gas leakage directions are approximately the same. Note that the number, intervals, and time for acquiring the gas leakage direction and the leakage source candidate position are merely examples and are not limited thereto.

[0044] The arithmetic processing unit 30 determines that the multiple leak directions are substantially the same if the angle between the multiple leak directions is less than a predetermined threshold value, and determines that the multiple leak directions are not substantially the same if the angle between the multiple leak directions is equal to or greater than a threshold value.

[0045] If it is determined that the multiple leak directions are not substantially the same, the leak source determination unit 34 determines the leak source by, for example, a method of finding an intersection of the multiple leak directions. On the other hand, if it is determined that the multiple leak directions are substantially the same, the method of finding an intersection of the multiple leak directions is not applied to the determination of the leak source. The reason for this is explained below.

[0046] 7 is an explanatory diagram showing the relationship between the angle between multiple leak directions and the error. When multiple leak directions are approximately the same and the angle between the multiple leak directions is an acute angle, there is a possibility that there will be a large error between the actual leak source and the leak source estimated at the intersection of the leak directions.

[0047] As shown in Fig. 7, the angle between a line L1 indicating the leak direction at a certain time and a line L2 indicating the leak direction at another time is defined as θ. Also, the amount of deviation between an intersection point P1 of the lines L1 and L2 and an intersection point P2 of the line L1 and a line L2' that is parallel to the line L2 and shifted by Δx is defined as Δd. Δd can be calculated using the following formula (1).

[0048]

number

[0049] Suppose that line L2' is translated by, for example, 5 pixels relative to line L2. If the angle θ is, for example, 30°, the intersection point P2 of line L1 and line L2' will be shifted by 10 pixels from the intersection point P1 of line L1 and line L2. If the angle θ is even more acute, an even larger error will occur between the two intersection points.

[0050] On the other hand, when the angle θ is 90°, if the line L2′ is translated by five pixels relative to the line L2, the amount of deviation of the intersection point P2 from the intersection point P1 can be suppressed to five pixels.

[0051] Therefore, the threshold for determining whether the multiple leak directions are substantially the same is set to, for example, 30°. Then, the method for determining the gas leak source candidates is switched depending on whether the angle between the multiple leak directions is less than the threshold or greater than the threshold. Note that the threshold for determining whether the multiple leak directions are substantially the same is not limited to 30°.

[0052] <Example of method for determining leak source when multiple leak directions are approximately the same> FIG. 8 is an explanatory diagram showing an example of a method for determining a leakage source when a plurality of leakage directions are substantially the same.

[0053] In step SA1 of FIG. 2, the fluid region extraction unit 31 acquires a gas region CL1 at time t1, acquires a gas region CL2 at time t2, and acquires a gas region CL3 at time t3.

[0054] 2, the leak direction calculation unit 32 acquires a gas leak direction L10 from the gas region CL1 acquired by the fluid region extraction unit 31. The leak direction calculation unit 32 also acquires a gas leak direction L20 from the gas region CL2 acquired by the fluid region extraction unit 31. The leak direction calculation unit 32 also acquires a gas leak direction L30 from the gas region CL3 acquired by the fluid region extraction unit 31. The calculation processing unit 30 records the gas leak directions L10, L20, and L30 acquired by the leak direction calculation unit 32 in the leak direction storage unit 40 of the storage unit 4 in step SA3 of FIG.

[0055] 2, the leak source candidate calculation unit 33 acquires a gas leak source candidate position P10 from the gas leak direction L10 acquired by the leak direction calculation unit 32. The leak source candidate calculation unit 33 also acquires a gas leak source candidate position P20 from the gas leak direction L20 acquired by the leak direction calculation unit 32. The leak source candidate calculation unit 33 also acquires a gas leak source candidate position P30 from the gas leak direction L30 acquired by the leak direction calculation unit 32. The arithmetic processing unit 30 records the gas leak source candidate positions P10, P20, P30 acquired by the leak source candidate calculation unit 33 in the leak source candidate storage unit 41 of the storage unit 4 in step SA5 of FIG.

[0056] In step SA7 of FIG. 2, if the angle formed by the multiple leak directions is less than a predetermined threshold, the calculation processing unit 30 determines that the multiple leak directions are substantially the same.

[0057] In this case, the leak source determination unit 34 determines a leak source from among a plurality of leak source candidate positions determined from each leak direction in step SA8 of Fig. 2. The leak source determination unit 34 determines an average leak direction Lc using the leak directions L10, L20, and L30. The leak source determination unit 34 selects the leak source candidate position P10, P20, and P30 that is located at the most upstream of the average leak direction Lc. For example, when the leak source candidate position P10 is located at the most upstream of the average leak direction Lc, the leak source determination unit 34 determines the leak source candidate position P10 as the leak source position Pa.

[0058] The leak source determination unit 34 determines whether or not either the x-coordinate or the y-coordinate of the leak source candidate position is outside the interquartile range. The leak source determination unit 34 excludes the leak source candidate positions that are outside the interquartile range from the leak source determination targets. Then, the leak source determination unit 34 selects the leak source candidate position that is located most upstream in the average leak direction from among the remaining leak source candidate positions, and determines the leak source. This makes it possible to eliminate the effects of noise, etc.

[0059] <Example of method for determining leak source when multiple leak directions are not approximately the same> 9A, 9B, 10A, and 10B are explanatory diagrams showing an example of a method for determining a leakage source when a plurality of leakage directions are not substantially the same.

[0060] 2, the fluid region extraction unit 31 acquires the gas region CL10 at time t1, the gas region CL20 at time t2, and the gas region CL30 at time t3. The calculation processing unit 30 acquires the centers of gravity of the gas regions CL10, CL20, and CL30.

[0061] 2, the leak direction calculation unit 32 acquires the gas leak direction L11 from the gas region CL10 acquired by the fluid region extraction unit 31. The leak direction calculation unit 32 also acquires the gas leak direction L21 from the gas region CL20 acquired by the fluid region extraction unit 31. The leak direction calculation unit 32 also acquires the gas leak direction L31 from the gas region CL30 acquired by the fluid region extraction unit 31. The leak directions L11, L21, and L31 are indicated by straight lines. The calculation processing unit 30 records the gas leak directions L11, L21, and L31 acquired by the leak direction calculation unit 32 in the leak direction storage unit 40 of the storage unit 4 in step SA3 of FIG. 2.

[0062] 2, the leak source candidate calculation unit 33 acquires a gas leak source candidate position P11 from the gas leak direction L11 acquired by the leak direction calculation unit 32. The leak source candidate calculation unit 33 also acquires a gas leak source candidate position P21 from the gas leak direction L21 acquired by the leak direction calculation unit 32. The leak source candidate calculation unit 33 also acquires a gas leak source candidate position P31 from the gas leak direction L31 acquired by the leak direction calculation unit 32. The arithmetic processing unit 30 records the gas leak source candidate positions P11, P21, P31 acquired by the leak source candidate calculation unit 33 in the leak source candidate storage unit 41 of the storage unit 4 in step SA5 of FIG.

[0063] In step SA7 of FIG. 2, if the angle formed by the multiple leak directions is equal to or greater than a predetermined threshold, the calculation processing unit 30 determines that the multiple leak directions are not substantially the same.

[0064] In this case, the leakage source determination unit 34 determines the leakage source in step SA9 of FIG. 2, for example, by a method of obtaining an intersection of a plurality of leakage directions. As a specific method, the leakage direction L11 is set as a straight line passing through the center of gravity of the gas region CL10 from which the leakage direction L11 is obtained. Or, the leakage direction L11 is set as a straight line passing through the leakage source candidate position P11 obtained from the leakage direction L11. Also, the leakage direction L21 is set as a straight line passing through the center of gravity of the gas region CL20 from which the leakage direction L21 is obtained. Or, the leakage direction L21 is set as a straight line passing through the leakage source candidate position P21 obtained from the leakage direction L21. Furthermore, the leakage direction L31 is set as a straight line passing through the center of gravity of the gas region CL30 from which the leakage direction L31 is obtained. Or, the leakage direction L31 is set as a straight line passing through the leakage source candidate position P31 obtained from the leakage direction L31.

[0065] 9A, the leak source determination unit 34 determines an intersection CP12 between a line of the leak direction L11 and a line of the leak direction L21. The leak source determination unit 34 also determines an intersection CP13 between the line of the leak direction L11 and a line of the leak direction L31. The leak source determination unit 34 also determines an intersection CP23 between the line of the leak direction L21 and a line of the leak direction L31. The leak source determination unit 34 determines the center of gravity of the intersections CP12, CP13, and CP23, and determines the center of gravity as the leak source position Pb.

[0066] In the method shown in Fig. 9B, the straight lines of a plurality of leak directions acquired at different times are obtained so as to pass through the center of gravity of the gas region or the position of the possible leak source. Up to this point, the method is the same as that shown in Fig. 9A.

[0067] In the method shown in Fig. 9B, the leak source determination unit 34 calculates the point that is the shortest distance from the straight line of the leak direction L11, the straight line of the leak direction L21, and the straight line of the leak direction L31. The leak source determination unit 34 determines that point as the leak source position Pc. The point that is the shortest distance from the straight line of each leak direction can be found by minimizing the following equation (2).

[0068]

number

[0069] In equation (2), c is a neighboring point of multiple lines, ui is a point on the line, and vi is a directional vector of the line. The point with the shortest distance from the line in each leak direction is found by minimizing the sum of the distance between the neighboring point and the line.

[0070] 10A and 10B, the straight lines of the multiple leak directions acquired at different times are obtained so as to pass through the center of gravity of the gas region or the position of the potential leak source. Up to this point, this is the same as the method described in FIG. 9A.

[0071] In the method shown in FIG. 10A and FIG. 10B, a probability map MP11 is created by weighting the upstream side in a straight line of the leak direction L11. In the probability map MP11, for example, a weight of 0.5% is added for each pixel away from the leak source candidate position P11 to the upstream side. Conversely, the weight is reduced by 0.5% for each pixel away from the leak source candidate position P11 to the downstream side. Furthermore, a probability map MP21 is created by weighting the upstream side in a straight line of the leak direction L21. In the probability map MP21, for example, a weight of 0.5% is added for each pixel away from the leak source candidate position P21 to the upstream side. Conversely, the weight is reduced by 0.5% for each pixel away from the downstream side. Furthermore, a probability map MP31 is created by weighting the upstream side in a straight line of the leak direction L31. In the probability map MP31, for example, a weight of 0.5% is added for each pixel away from the leak source candidate position P31 to the upstream side. Conversely, the weight is reduced by 0.5% for each pixel away from the downstream side.

[0072] The leak source determination unit 34 finds the position with the largest value in the probability maps MP11, MP21, and MP31 created in this way, and determines the highest probability position as the leak source position Pd.

[0073] <Other examples of methods for determining leak direction, potential leak source location, and leak source> 11 is an explanatory diagram showing another example of a method for determining the leak direction, the leak source candidate position, and the leak source. The fluid region extraction unit 31 acquires the gas region CL10 in step SA1 of FIG.

[0074] The leak direction calculation unit 32 searches for points where the gas is highly concentrated to determine the leak direction in step SA2 in Fig. 2. The leak source candidate calculation unit 33 searches for points where the gas is highly concentrated to determine the leak source candidate positions. In the method described in Fig. 11, vertices N1 to N4 of a rectangle CL14 shown by a dashed line including the gas region CL10 are set as starting points of interest, and circles C1 to C4 of an arbitrary radius of 1 are set.

[0075] Next, the circumference of each of the circles C1 to C4 is searched, and the point of interest is moved to the point that has a higher density than the pixel of interest and has the highest density. This is repeated until there are no more points to move to.

[0076] Among the search trajectories from vertices N1 to N4, the starting point of the trajectory with the shortest straight-line distance between the starting point and the end point is considered to be the starting point closest to the gas leakage source. However, since this point is likely to be a point where a high-concentration point has not been found, it is excluded before executing the subsequent processes. In FIG. 11, for example, the search trajectory NL4 from vertex N4 is the trajectory with the shortest straight-line distance between the starting point and the end point. Therefore, vertex N4 is excluded.

[0077] For the remaining search trajectories NL1 to NL3, if each of the search trajectories NL1 to NL3 is approximated to a straight line, a straight line set from the gas concentration can be drawn. For example, a straight line NL10 set from the gas concentration can be drawn from the search trajectory NL1. A straight line NL20 set from the gas concentration can be drawn from the search trajectory NL2. Furthermore, a straight line NL30 set from the gas concentration can be drawn from the search trajectory NL3. Note that the straight line set from the gas concentration can be not only an approximation of the search trajectory, but also a straight line connecting the start point and end point of the search trajectory.

[0078] The leak direction calculation unit 32 determines the line NL10 as the gas leak direction with the starting point side of the search trajectory NL1 as the downstream direction. The leak direction calculation unit 32 also determines the line NL20 as the gas leak direction with the starting point side of the search trajectory NL2 as the downstream direction. The leak direction calculation unit 32 also determines the line NL30 as the gas leak direction with the starting point side of the search trajectory NL3 as the downstream direction. In step SA3 of FIG. 2, the calculation processing unit 30 records the gas leak direction acquired by the leak direction calculation unit 32 in the leak direction memory unit 40 of the memory unit 4.

[0079] 2, the leak source candidate calculation unit 33 determines the end point NL11 of the search trajectory NL1 as the leak source candidate position. The leak source candidate calculation unit 33 also determines the end point NL21 of the search trajectory NL2 as the leak source candidate position. The leak source candidate calculation unit 33 also determines the end point NL31 of the search trajectory NL3 as the leak source candidate position. The arithmetic processing unit 30 records the gas leak source candidate position acquired by the leak source candidate calculation unit 33 in the leak source candidate memory unit 41 of the memory unit 4 in step SA5 of FIG. 2.

[0080] In step SA6 of FIG. 2, the calculation processing unit 30 repeats the process of searching for a point where the gas is highly concentrated and determining the leak direction and the leak source candidate position at least twice or more at different times. In step SA7 of FIG. 2, if the angle between the multiple leak directions is less than a predetermined threshold, the calculation processing unit 30 determines that the multiple leak directions are substantially the same. In this case, in step SA8 of FIG. 2, the leak source determination unit 34 determines the leak source from the multiple leak source candidate positions determined from each leak direction. In addition, if the angle between the multiple leak directions is equal to or greater than a predetermined threshold, the calculation processing unit 30 determines that the multiple leak directions are not substantially the same. In this case, in step SA9 of FIG. 2, the leak source determination unit 34 determines the leak source position Pe by, for example, a method of determining an intersection point of the multiple leak directions.

[0081] 12, 13A, and 13B are explanatory diagrams showing other examples of the method for determining the leak direction, the leak source candidate position, and the leak source. In step SA1 of FIG. 2, the fluid region extraction unit 31 acquires a gas region CL10 at time t1. In addition, the fluid region extraction unit 31 acquires a gas region CL20 at time t2 different from time t1.

[0082] 12, the calculation processing unit 30 uses the gas cloud center line to obtain a straight line ML10 that indicates the gas cloud shape of the gas region CL10. Similarly, the calculation processing unit 30 uses the gas cloud center line to obtain a straight line ML20 that indicates the gas cloud shape of the gas region CL20. The gas cloud center line is an approximated straight line that is drawn by binarizing the gas region and minimizing the sum of squares of the distances from each point in the binarized gas region.

[0083] The leak direction calculation unit 32 determines the gas leak direction from the straight line indicating the gas cloud shape in step SA2 in Fig. 2. The calculation processing unit 30 records the gas leak direction in the leak direction storage unit 40 of the storage unit 4 in step SA3 in Fig. 2.

[0084] The leak source candidate calculation unit 33 determines the leak source candidate positions from the straight lines indicating the gas cloud shape in step SA4 in Fig. 2. The arithmetic processing unit 30 records the gas leak source candidate positions in the leak source candidate storage unit 41 of the storage unit 4 in step SA5 in Fig. 2.

[0085] The calculation processing unit 30 repeats the process of obtaining a straight line indicating the gas cloud shape using the gas cloud center line at least two or more times at different times in step SA6 of FIG.

[0086] The calculation processor 30 determines whether the angle between the lines showing the gas cloud shapes at different times is less than a predetermined threshold value in step SA7 of Fig. 2. The angle between the lines showing the gas cloud shapes at different times is the angle between the leak directions.

[0087] The calculation processing unit 30 determines that the multiple leak directions are substantially the same if the angle between the multiple straight lines showing the gas cloud shapes at different times is less than a predetermined threshold. In this case, the leak source determination unit 34 determines the leak source from the multiple leak source candidate positions obtained from each leak direction in step SA8 of Fig. 2. In addition, the calculation processing unit 30 determines that the multiple leak directions are not substantially the same if the angle between the multiple straight lines showing the gas cloud shapes at different times is equal to or greater than a predetermined threshold. In this case, the leakage source determination unit 34 determines, for example, an intersection point of straight lines indicating a gas cloud shape set from the contour lines or center lines of multiple fluid regions in step SA9 in Fig. 2. Then, the leakage source position Pf is determined by a method of determining an intersection point of multiple leakage directions.

[0088] 13A and 13B, the calculation processing unit 30 uses the gas cloud contour line to obtain gas cloud contour straight lines CL11a and CL11b of the gas region CL10. Similarly, the calculation processing unit 30 uses the gas cloud contour line to obtain gas cloud contour straight lines CL21a and CL21b of the gas region CL20.

[0089] The calculation processing unit 30 binarizes the gas region CL10 to obtain the gas cloud center line ML10, and extracts the contour of the gas region CL10 from the binarized image. The contour of the gas region CL10 can be extracted using a Sobel filter or a Laplacian filter. The calculation processing unit 30 divides the contour of the gas region CL10 into two by the gas cloud center line ML10 to obtain a first contour line CL10a and a second contour line CL10b. The first contour line CL10a and the second contour line CL10b are one contour line and the other contour line obtained by dividing the contour of the gas region CL10 by the gas cloud center line ML10. In FIG. 13B, the first contour line CL10a is shown by a solid line, and the second contour line CL10b is shown by a dashed line. The calculation processor 30 then obtains a first gas cloud contour straight line CL11a by approximating the area surrounded by the first contour line CL10a and the gas cloud center line ML10 with a straight line drawn so as to minimize the sum of squares of the distances from each point in the area. The calculation processor 30 also obtains a second gas cloud contour straight line CL11b by approximating the area surrounded by the second contour line CL10b and the gas cloud center line ML10 with a straight line drawn so as to minimize the sum of squares of the distances from each point in the area.

[0090] Similarly, the calculation processing unit 30 binarizes the gas region CL20 to obtain a gas cloud center line ML20, and extracts the contour of the gas region CL20 from the binarized image. The calculation processing unit 30 divides the contour of the gas region CL20 into two by the gas cloud center line ML20 to obtain a first contour line CL20a and a second contour line CL20b. The calculation processing unit 30 then approximates the first contour line CL20a with a straight line to obtain a first gas cloud contour straight line CL21a. The calculation processing unit 30 also approximates the second contour line CL20b with a straight line to obtain a second gas cloud contour straight line CL21b.

[0091] The leak direction calculation unit 32 obtains the gas leak direction from the straight line indicating the gas cloud contour line in step SA2 of Fig. 2. The calculation processing unit 30 records the gas leak direction in the leak direction storage unit 40 of the storage unit 4 in step SA3 of Fig. 2.

[0092] The leak source candidate calculation unit 33 determines the leak source candidate positions from the straight line indicating the gas cloud contour line in step SA4 of Fig. 2. The arithmetic processing unit 30 records the gas leak source candidate positions in the leak source candidate storage unit 41 of the storage unit 4 in step SA5 of Fig. 2.

[0093] The calculation processing unit 30 repeats the process of acquiring the gas cloud contour straight line in step SA6 of FIG. 2 at least two or more times at different times.

[0094] In step SA7 of Fig. 2, the calculation processor 30 determines whether the angle between the lines showing the gas cloud contour at different times is less than a predetermined threshold value. The angle between the lines showing the gas cloud contour at different times is the angle between the leak directions.

[0095] If the angle between the straight lines showing the gas cloud contours at different times is less than a predetermined threshold, the calculation processing unit 30 determines that the multiple leak directions are substantially the same. In this case, the leak source determination unit 34 determines the leak source from among the multiple leak source candidate positions obtained from each leak direction in step SA8 of FIG. 2. Also, if the angle between the straight lines showing the gas cloud contours at different times is equal to or greater than a predetermined threshold, the calculation processing unit 30 determines that the multiple leak directions are not substantially the same. In this case, the leak source determination unit 34 determines the leak source position Pg in step SA9 of FIG. 2, for example, by a method of obtaining the center of gravity of the intersection of the straight lines showing the gas cloud contours at different times.

[0096] The monitoring target is not limited to gas, but may be a fluid such as liquid. The process of determining the leakage source position is not limited to an infrared image, but may be performed by acquiring multiple pieces of fluid information in time series from visible images or laser light. When acquiring fluid information from laser light, for example, the method of JP 2002-296142 A is used. In the method of JP 2002-296142 A, multiple laser lights are screened in time series, and a velocity vector obtained by time-space differentiation or particle correlation of a moving image is used. In addition, when acquiring fluid information from other images, for example, the method of WO2016 / 208317 is used. In the method of WO2016 / 208317, a movement trajectory acquired by utilizing optical flow or the like based on multiple images acquired in time series is used. [Explanation of symbols]

[0097] REFERENCE SIGNS LIST 1 leak source position calculation system, 2 image acquisition unit, 3 leak source position calculation device, 30 calculation processing unit, 31 fluid region extraction unit, 32 leak direction calculation unit, 33 leak source candidate calculation unit, 34 leak source determination unit, 4 storage unit, 40 leak direction storage unit, 41 leak source candidate storage unit, 5 output unit

Claims

1. A leakage direction acquisition unit that acquires leakage directions of at least two or more leaking fluids at different times; a leakage source candidate position acquisition unit that acquires leakage source candidate positions from the plurality of leakage directions at different times; a leakage source determining unit that determines a leakage source position by a determination method according to whether an angle between the plurality of leakage directions at different times is less than a threshold value or is equal to or greater than a threshold value; A leak source location calculation device comprising:

2. When the angle between the multiple leak directions at different times is less than a threshold value, the leak source determination unit determines, as a leak source position, a leak source candidate position located at the most upstream of the leak directions from among the multiple leak source candidate positions acquired from the multiple leak directions at different times. The leakage source location calculation device according to claim 1 .

3. If the angle between the multiple leak directions at different times is equal to or greater than a threshold value, the leak source determination unit obtains an intersection point between the lines based on the multiple leak directions at different times, and determines the center of gravity of each intersection point as the leak source position. The leakage source location calculation device according to claim 1 .

4. If the angle between the multiple leak directions at different times is equal to or greater than a threshold value, the leak source determination unit calculates the concentration of the multiple leak directions at different times, calculates the intersection points of the lines set based on the concentration, and determines the center of gravity of each intersection point as the leak source position. The leakage source location calculation device according to claim 1 .

5. When the angle between the multiple leak directions at different times is equal to or greater than a threshold value, the leak source determination unit obtains an intersection point between each straight line set from the contour lines or center lines of the multiple fluid regions from which the multiple leak directions at different times are obtained, and determines the center of gravity of each intersection point as the leak source position. The leakage source location calculation device according to claim 1 .

6. When the angle between the multiple leak directions at different times is equal to or greater than a threshold value, the leak source determination unit determines a point with the shortest distance from each straight line set from the multiple leak directions at different times as a leak source position. The leakage source location calculation device according to claim 1 .

7. If the angle between the multiple leak directions at different times is equal to or greater than a threshold value, the leak source determination unit creates a probability map weighted toward the upstream side of the multiple leak directions at different times, and determines the highest probability position as the leak source position. The leakage source location calculation device according to claim 1 .

8. The line based on the leak direction passes through the center of gravity of the fluid domain from which this leak direction is obtained.

8. The leakage source location calculation device according to claim 3, 6 or 7.

9. A straight line based on the leak direction passes through the leak source candidate position obtained from this leak direction.

8. The leakage source location calculation device according to claim 3, 6 or 7.

10. an image acquisition unit for acquiring an image of a subject including a monitoring target; a leakage source position calculation device that determines a leakage source position from time-series images acquired by the image acquisition unit, The leakage source location calculation device comprises: A leakage direction acquisition unit that acquires leakage directions of at least two or more leaking fluids at different times; a leakage source candidate position acquisition unit that acquires leakage source candidate positions from the plurality of leakage directions at different times; a leakage source determining unit that determines the leakage source position by a determination method according to whether an angle formed by the plurality of leakage directions at different times is less than a threshold value or is equal to or greater than a threshold value. Leak source location calculation system.

11. A program executed by a processor of a leakage source position calculation device that determines a leakage source position from time-series images, Obtaining leakage directions of at least two or more leaking fluids at different times; acquiring leakage source candidate positions from the plurality of leakage directions at different times; determining a leakage source position by a determination method according to whether an angle between the plurality of leakage directions at different times is less than a threshold value or is greater than or equal to a threshold value; The calculation processing unit executes program.

12. Obtaining leakage directions of at least two or more leaking fluids at different times; acquiring leakage source candidate positions from the plurality of leakage directions at different times; determining a leakage source position using a determination method according to whether an angle between the plurality of leakage directions at different times is less than a threshold value or is greater than or equal to a threshold value; Leak source location calculation method.

Images