JP2020085829A - Vibration detection method - Google Patents

Abstract

To provide a vibration detection method that can detect vibration generating displacement which is smaller than size of one pixel of a camera, for the camera installed in a predetermined place.SOLUTION: A vibration detection method, which is a method to detect vibration acting on a first camera 2 installed in a predetermined place, comprises: an imaging step of imaging a time-series image V2 using a second camera 6, in a state where a second camera 6 having a smaller field angle per one pixel than the first camera 2 is installed in the predetermined place; and a generating step of generating vibration data D indicating the vibration acting on the second camera 6 on the basis of the time-series image V2 imaged using the second camera 6.SELECTED DRAWING: Figure 1

Description

The present invention relates to a method for detecting vibrations acting on a camera.

Facilities that handle a large amount of gas during operation include, for example, gas plants, petrochemical plants, thermal power plants, and steelmaking-related facilities. In these facilities, a gas detector may be used because gas leakage may occur due to deterioration of the facility over time, operation error of the facility, or the like. As a result, the gas leak can be found in a state where the gas leak is small, so that a major accident can be prevented.

As a gas detection device, a probe type gas detection device is widely used, but recently, an infrared image type gas detection device has come into use. The former detects a gas based on the change in the electrical characteristics of the probe when a gas molecule comes in contact with the probe. The latter detects gas based on the property that the gas absorbs or emits light.

The infrared image type gas detection device will be described in detail. In the presence of gas, a part of electromagnetic waves (mainly electromagnetic waves in the infrared region) generated by the background body having an absolute temperature of 0 K or more radiating a black body is absorbed by the gas, and the gas itself emits the electromagnetic wave by the black body radiation. Because it radiates, the amount of electromagnetic waves changes. Therefore, when gas exists in the imaging range of the infrared image, the pixel value (brightness value) of the pixel corresponding to the region where the gas exists in the infrared image (hereinafter, gas region) changes in the amount of electromagnetic waves. Is a value that reflects. Therefore, the gas can be detected and the gas can be visualized. According to the infrared image type gas detection device, gas detection can be performed without bringing the infrared camera close to an imaging target (for example, a gas transport pipe in a plant).

An infrared image type gas detection device is disclosed in Patent Document 1, for example. Patent Document 1 discloses a data quality assurance system including an imaging unit (infrared camera) and a gas monitoring processing unit that performs image processing on detection data output by the imaging unit to detect gas. The data quality assurance system invalidates the detection data output by the imaging unit if the blur of the imaging unit (infrared camera) exceeds the correctable range.

JP, 2016-170029, A

The infrared camera is installed at a predetermined place where a gas leakage monitoring target can be photographed. The blurring of the infrared camera is caused by the vibration acting on the infrared camera installed at a predetermined place. The cause of the vibration is the operation of the above facility, wind, etc. The reason why the accuracy of gas detection deteriorates when vibration acts on the infrared camera will be briefly described. When vibration acts on the infrared camera, each pixel forming the image sensor mounted on the infrared camera vibrates. As a result, the background area imaged by each pixel changes in a minute time period, and the change in the signal indicating the change in the amount of electromagnetic waves in the background increases. In the background, the amount of electromagnetic waves in the high-contrast area is significantly different from that in the low-contrast area.

The change in the amount of electromagnetic waves caused by the presence of gas is slight. Therefore, when the change in the signal indicating the change in the background electromagnetic wave becomes large, the signal indicating the change in the electromagnetic wave amount caused by the presence of gas is difficult to separate from the signal indicating the change in the background electromagnetic wave amount. Become. Therefore, it is not desirable to install the infrared camera in such a place even if the vibration causes a deviation smaller than one pixel size of the infrared camera (a deviation of sub-pixel level).

Patent Document 1 describes an acceleration sensor, a gyro sensor, or a means using feature point extraction of an image as means for detecting vibration of an infrared camera. However, it is difficult for these means to accurately detect the vibration that causes the subpixel level shift. There is a demand for a technique capable of detecting vibration that causes a subpixel level shift.

It is an object of the present invention to provide a vibration detection method for a camera installed in a predetermined place, which can detect a vibration that causes a shift smaller than the one pixel size of the camera.

A vibration detecting method according to the present invention that achieves the above object is a method for detecting vibration acting on a first camera installed at a predetermined location, wherein an angle of view per pixel is greater than that of the first camera. A photographing step of photographing a time-series image using the second camera in a state where a small second camera is installed in the predetermined place; and the time-series photographed using the second camera. A generation step of generating vibration data indicating vibration acting on the second camera based on the image.

According to the present invention, it is possible to detect, for a camera installed at a predetermined location, a vibration that causes a shift smaller than the one pixel size of the camera.

It is a block diagram showing each of a vibration detection system and a gas detection system concerning an embodiment. It is a functional block diagram of a vibration detection device. It is a block diagram which shows the hardware constitutions of the vibration detection apparatus shown in FIG. 2A. It is an explanatory view explaining time series pixel data. It is a flow chart explaining the image processing which extracts a gas field. It is explanatory drawing explaining the step which produces|generates the horizontal vibration data D_L from the horizontal displacement data d_L. It is a flow chart explaining the vibration detection method concerning an embodiment. It is a block diagram which shows the modification of a vibration data generation part.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each of the drawings, the components denoted by the same reference numerals indicate the same components, and the description of the components already described will be omitted.

FIG. 1 is a block diagram showing each of a vibration detection system 5 and a gas detection system 1 according to the embodiment. The gas detection system 1 will be described first. The gas detection system 1 includes an infrared camera 2 and a gas detection device 3.

The infrared camera 2 (an example of a first camera) is connected to the gas detection device 3 and shoots a moving image V1 (an example of a time-series image) of an infrared image of the imaging target T under the control of the gas detection device 3, The moving image V1 is transmitted to the gas detection device 3. The time-series images have a structure in which images captured at different times are arranged in time series. In the embodiment, as the imaging target T, a gas leakage monitoring target (for example, a gas transport pipe) will be described as an example. The infrared camera 2 is installed in a predetermined place where the subject T can be photographed (for example, a place where the subject T can be overlooked). Such places include, for example, rooftops of buildings, steel towers, chimneys, poles, fences and concrete columns. The height and angle of the infrared camera 2 are set at a predetermined location so that the imaging target T is included in the imaging range R1 of the infrared camera 2.

The infrared camera 2 includes a two-dimensional image sensor 20 having sensitivity to the light absorption wavelength of gas to be detected. For example, in the case of a hydrocarbon gas, the two-dimensional image sensor 20 having sensitivity in the wavelength band of about 3 μm is used. Examples of such a sensor include a cooled type indium antimony (InSb) image sensor and a cooled type mercury cadmium tellurium (HgCdTe) image sensor.

When the wavelength band of electromagnetic waves absorbed or radiated by the gas to be detected is other than the infrared wavelength band (for example, the ultraviolet wavelength band), a camera having sensitivity in this wavelength band is used.

The gas detection device 3 is a computer device that performs various image processes necessary for gas detection using the moving image V1 captured by the infrared camera 2. The gas detection device 3 performs image processing for extracting a region (gas region) in which gas appears in the frame forming the moving image V1. If gas appears in the imaging range R due to gas leaking from the imaging target T, the gas region is extracted.

The vibration detection system 5 includes a visible light camera 6 (an example of a second camera) and a vibration detection device 7. The visible light camera 6 is installed in the above-mentioned predetermined place. The visible light camera 6 may be installed at the same place as the infrared camera 2 or may not be exactly the same as long as it can detect vibrations acting on the infrared camera 2. An example in which the infrared camera 2 is installed on a pole will be described. The visible light camera 6 may be attached to the mounting jig for the infrared camera fixed to the pole (strictly the same), or the same pole with the infrared camera 2 attached to the mounting jig fixed to the pole. The visible light camera 6 may be attached to another attachment jig fixed to.

The visible light camera 6 includes a two-dimensional image sensor 60 having a sensitivity in the wavelength range of visible light. The two-dimensional image sensor 60 is, for example, a two-dimensional CMOS (Complementary Metal Oxide Semiconductor) image sensor or a two-dimensional CCD (Charge Coupled Device) image sensor.

The second camera is not limited to the visible light camera 6 as long as it can capture a moving image that can generate vibration data. The second camera may be a visible light camera, a near infrared camera, or an infrared camera.

The shooting range R2 of the visible light camera 6 is included in the shooting range R1 of the infrared camera 2. The imaging target of the visible light camera 6 is a stationary object (for example, a building or a pipe) in the imaging range of the infrared camera 2.

The frame rate of the visible light camera 6 is higher than the frame rate of the infrared camera 2. This is because if the frame rate of the visible light camera 6 is smaller than the frame rate of the infrared camera 2, there will be a leak in the detection of vibration that acts on the image captured in the moving image V1 captured by the infrared camera 2. Preferably, the frame rate of the visible light camera 6 is an integral multiple of the frame rate of the infrared camera 2.

The angle of view per pixel of the visible light camera 6 is smaller than the angle of view per pixel of the infrared camera 2. Let θ1 be the angle of view per pixel of the infrared camera 2, and θ2 be the angle of view per pixel of the visible light camera 6. The following expression 1 is established.
θ2<θ1/N Equation 1

Here, N represents the number of magnifications. The magnification number N is larger than 1, and preferably 10 or more. Since the displacement of the image due to the vibration is the displacement of the sub-pixel size in the pixel of the infrared camera 2 by the expression 1 being satisfied, the displacement of the image in the pixel of the visible light camera 6 is equal to or larger than the pixel size. It is possible to obtain vibration data indicating a deviation smaller than the pixel size of the infrared camera 2. When the magnification N increases, vibration data indicating a smaller deviation can be obtained, and the magnification N can be increased by attaching a telephoto lens to the visible light camera 6. In this case, the magnification N increases as the magnification of the telephoto lens increases.

α1 is the angle of view of the infrared camera 2, α2 is the angle of view of the visible light camera 6, n1_L is the number of horizontal pixels of the infrared camera 2, n1_V is the number of vertical pixels of the infrared camera 2, and n2_L is the visible light camera 6. The number of pixels in the horizontal direction and n2_V are the number of pixels in the vertical direction of the visible light camera 6.

In the horizontal direction, θ1=α1/n1_L and θ2=α2/n2_L. In the vertical direction, θ1=α1/n1_V and θ2=α2/n2_V. Expression 1 can be rewritten into Expression 2-1 and Expression 2-2.
(Α2/n2_L)<(α1/n1_L)/N Equation 2-1
(Α2/n2_V)<(α1/n1_V)/N Equation 2-2

Expression 2-1 and expression 2-2 can be rewritten as expression 3-1 and expression 3-2, respectively.
α2<{α1×(n2_L/n1_L)}/N... Equation 3-1
α2<{α1×(n2_V/n1_V)}/N... Formula 3-2

The pixel number ratio is n2_L/n1_L (horizontal pixel number of visible light camera 6/horizontal pixel number of infrared camera 2) or n2_V/n1_V (vertical pixel number of visible light camera 6/infrared camera) 2 in the vertical direction). Expression 3-1 and Expression 3-2 can be rewritten as Expression 4.

α2<(α1×pixel number ratio)÷N...Equation 4

From the above, Expression 1 can be rewritten as Expression 4.

In the embodiment, the first camera and the second camera are separate, but the same camera may be used as the first camera and the second camera. That is, the infrared camera 2, which is the first camera, may be used as the second camera. When the infrared camera 2 is used as the second camera, Equation 1 can be established by attaching a telephoto lens to the infrared camera 2.

Referring to FIG. 1, the vibration detection device 7 is a computer device having a function of generating vibration data based on the moving image V2 captured by the visible light camera 6. FIG. 2A is a functional block diagram of the vibration detection device 7. The vibration detection device 7 is provided in a personal computer, a smartphone, a tablet terminal, or the like, and includes a main body unit 70, an operation unit 71, and a display unit 72 as functional blocks. The main body unit 70 includes a control processing unit 700, an interface unit 701, a vibration data generation unit 702, an installation determination unit 703, a storage unit 704, and a display control unit 705.

The control processing unit 700 controls each unit of the vibration detection device 7 (the interface unit 701, the vibration data generation unit 702, the installation determination unit 703, the storage unit 704, the display control unit 705) according to the function of each unit. It is a device.

The interface unit 701 is an interface for communicating between the vibration detection device 7 and a device (visible light camera 6, operation unit 71, display unit 72) external to this device under the control of the control processing unit 700. .. The interface unit 701 and the external device may be directly connected or may be connected via a LAN.

The vibration data generation unit 702 generates vibration data D indicating the vibration acting on the visible light camera 6, based on the moving image V2 (an example of a time series image) captured using the visible light camera 6. The vibration data generation unit 702 includes a displacement detection unit 7020 and a displacement data processing unit 7021.

The positional shift detection unit 7020 generates positional shift data d that indicates the positional shift amount of the image captured in the moving image V2 in time series. explain in detail. The positional shift detection unit 7020 performs template matching with respect to each frame constituting the moving image V2 by using a stationary object (for example, a building in the plant, piping) in the imaging range R2 (FIG. 1) as a template image. Thus, the amount of deviation of the image of the stationary object is measured for each frame. This is the positional shift data d indicating the positional shift amount. The positional deviation amount is measured in each of the horizontal direction (x direction) and the vertical direction (y direction) of the frame. Therefore, the positional deviation detecting unit 7020 generates the lateral positional deviation data d_L and the vertical positional deviation data d_V as the positional deviation data d.

The positional shift data processing unit 7021 performs a predetermined process on the positional shift data to generate vibration data D. The positional shift data processing unit 7021 generates horizontal vibration data D_L and vertical vibration data D_V as the vibration data D. Details of the displacement data processing unit 7021 will be described later.

The installation determination unit 703 determines whether or not the infrared camera 2 can be installed based on the vibration data D.

The storage unit 704 stores various images, data, and programs necessary to execute the functions of the control processing unit 700, the vibration data generation unit 702, the installation determination unit 703, and the display control unit 705.

The display control unit 705 causes the display unit 72 to display various images. For example, the display control unit 705 causes the display unit 72 to display an image indicating the vibration data D, or causes the display unit 72 to display an image indicating the determination result of whether or not the infrared camera 2 can be installed.

The operation unit 71 is a device that is connected to the interface unit 701 and allows a user to input various commands and various data to the vibration detection device 7. The command is, for example, a command for performing vibration detection. The data is, for example, data necessary for calculation of the vibration threshold value (angle of view θ1, angle of view θ2, etc.). The display unit 72 displays the various images described above.

FIG. 2B is a block diagram showing a hardware configuration of the vibration detection device 7 shown in FIG. 2A. The vibration detection device 7 includes a CPU (Central Processing Unit) 7a, a RAM (Random Access Memory) 7b, a ROM (Read Only Memory) 7c, a HDD (Hard Disk Drive) 7d, a liquid crystal display 7e, a communication interface 7f, a keyboard, and the like. And a bus 7h for connecting them. The liquid crystal display 7e is hardware that realizes the display unit 72. Instead of the liquid crystal display 7e, an organic EL display (Organic Light Emitting Diode display), a plasma display or the like may be used. The communication interface 7f is hardware that realizes the interface unit 701, and is, for example, a communication card for wired communication or wireless communication. The keyboard 7 g is hardware that implements the operation unit 71. A touch panel or a mouse may be used instead of the keyboard.

In the HDD 7d, programs for realizing the respective functional blocks of the control processing unit 700, the vibration data generation unit 702, the installation determination units 703 and 9 and the display control unit 705, and various data (for example, the moving image V2). Is stored. The program may be stored in the ROM 7c instead of the HDD 7d. The vibration detection device 7 may include a flash memory instead of the HDD 7d, and these programs and data may be stored in the flash memory. The HDD 7d is hardware that realizes the storage unit 704. A flash memory may be used instead of the HDD 7d.

The CPU 7a is an example of a hardware processor, and reads the above-mentioned program from the HDD 7d, expands it in the RAM 7b, and executes the expanded program to control the processing unit 700, the vibration data generation unit 702, and the installation determination unit 703. And the display control unit 705 is realized. However, with respect to these functions, some or all of the functions may be realized by a process by a DSP (Digital Signal Processor) instead of or together with the process by the CPU 7a. Further, similarly, a part or all of each function may be realized by a process by a dedicated hardware circuit instead of or together with the process by software.

The positional deviation data processing unit 7021 will be described in detail with reference to FIG. 2A. The positional shift data processing unit 7021 performs image processing for gas region extraction performed by the gas detection device 3 on the horizontal positional shift data d_L and the vertical positional shift data d_V (an example of image processing for image recognition. The horizontal vibration data D_L and the vertical vibration data D_V are generated by performing the same processing as the above).

A known technique can be applied to the image processing for gas region extraction. For example, this known technique is disclosed in International Publication WO201773430. This internationally disclosed technique is a technique invented by the present inventor, and this technique will be described by taking the moving image V1 taken by the infrared camera 2 shown in FIG. 1 as an example.

The moving image V1 has a structure in which a plurality of frames are arranged in time series. Data in which pixel data of pixels at the same position in a plurality of frames are arranged in time series is referred to as time series pixel data. The time-series pixel data will be specifically described. FIG. 3 is an explanatory diagram illustrating time-series pixel data. Let K be the number of frames of the moving image V1. It is assumed that one frame is composed of M pixels. Pixel data indicates a brightness value of a pixel or a temperature value obtained by converting the brightness value into a temperature value.

Pixels at the same position in a plurality (K) of frames mean pixels in the same order. For example, to describe the first pixel, the pixel data of the first pixel included in the first frame, the pixel data of the first pixel included in the second frame,..., K−1th frame The data obtained by arranging the pixel data of the first pixel included in the first pixel and the pixel data of the first pixel included in the Kth frame in time series is the time series pixel data of the first pixel. The number of time-series pixel data is the same as the number of pixels that configure one frame, and the moving image V1 is configured by a plurality (M) of these time-series pixel data.

Referring to FIG. 1, when a gas leak occurs in the state where the infrared camera 2 is capturing the moving image V1, the moving image V1 includes first frequency component data indicating a temperature change due to the leaked gas. .. The gas detection device 3 performs image processing on the moving image V1 to extract a gas region including the first frequency component data. FIG. 4 is a flowchart illustrating this image processing.

3 and 4, the gas detection device 3 extracts M second frequency component data corresponding to each of the M time-series pixel data (step S1). The second frequency component data has a lower frequency than the first frequency component data, and is data indicating a temperature change in the background. The second frequency component data is time-series pixel data obtained by calculating a simple moving average in units of a first predetermined number (eg, 21) of frames less than K frames with respect to the time-series pixel data. Data extracted from.

The gas detection device 3 extracts M third frequency component data corresponding to each of the M time-series pixel data (step S4). The third frequency component data is high frequency noise. For the third frequency component data, a simple moving average is calculated for the time-series pixel data in units of a third predetermined number (for example, 3) of frames smaller than the first predetermined number (for example, 21). This is the data extracted from the time-series pixel data.

The gas detection device 3 calculates M first difference data corresponding to each of the M time-series pixel data (step S2). The first difference data is data obtained by calculating the difference between the time-series pixel data and the second frequency component data extracted from this time-series pixel data.

The gas detection device 3 calculates M second difference data corresponding to each of the M time-series pixel data (step S5). The second difference data is data obtained by calculating the difference between the time-series pixel data and the third frequency component data extracted from this time-series pixel data.

The gas detection device 3 calculates M first variation data corresponding to each of the M time-series pixel data (step S3). The first variation data is data calculated by performing a predetermined calculation on the first difference data in units of a second predetermined number (for example, 21) of frames, and the first difference data. It is data indicating variations in data waveforms. The variation is, for example, a moving standard deviation.

The gas detection device 3 calculates M second variation data corresponding to each of the M time-series pixel data (step S6). The second variation data is data calculated by performing a predetermined calculation in units of a fourth predetermined number (for example, 21) of frames on the second difference data, and the second difference data. It is data indicating variations in data waveforms. The variation is, for example, a moving standard deviation.

The gas detection device 3 calculates M third difference data corresponding to each of the M time-series pixel data (step S7). The third difference data is data obtained by calculating the difference between the first variation data and the second variation data obtained from the same time series pixel data. The moving image composed of the M pieces of third difference data is the moving image in which the gas region is extracted.

The above is a description of an example of image processing for extracting a gas region. With reference to FIG. 2A, since the positional deviation data d indicates a time-series change in the positional deviation amount, it can be treated in the same manner as the time-series pixel data (FIG. 3) and image processing for gas region extraction (steps S1 to S1 in FIG. 4). The signal processing corresponding to S7) is possible (in other words, the same signal processing as the image processing of gas region extraction can be applied). The positional deviation data processing unit 7021 performs the same signal processing as the image processing (FIG. 4) of the gas region extraction on the lateral positional deviation data d_L and the vertical positional deviation data d_V, respectively, to obtain the lateral vibration data. D_L and vertical vibration data D_V are generated.

The lateral displacement data d_L will be described as an example. FIG. 5 is an explanatory diagram illustrating a step in which the lateral vibration data D_L is generated from the lateral position displacement data d_L. FIG. 5 includes graphs G1 to G8. The horizontal axis of each graph shows the order of frames. The vertical axis of each graph shows the size of the waveform.

The graph G1 shows the waveform of the lateral displacement data d_L. The graph G2 shows the waveform of the second frequency component data. The graph G3 shows the waveform of the third frequency component data. The graph G4 shows the waveform of the first difference data. The graph G5 shows the waveform of the second difference data. The graph G6 shows the waveform of the first variation data. The graph G7 shows the waveform of the second variation data. The graph G8 shows the waveform of the lateral vibration data D_L.

The positional shift data processing unit 7021 extracts the second frequency component data (graph G2) from the lateral positional shift data d_L (graph G1). This is the process of step S1 in FIG. The displacement data processing unit 7021 extracts the third frequency component data (graph G3) from the lateral displacement data d_L (graph G1). This is the process of step S4 in FIG.

The displacement data processing unit 7021 calculates first difference data (graph G4) obtained by calculating the difference between the lateral displacement data d_L (graph G1) and the second frequency component data (graph G2). .. This is the process of step S2 in FIG. The displacement data processing unit 7021 calculates second difference data (graph G5) obtained by calculating the difference between the lateral displacement data d_L (graph G1) and the third frequency component data (graph G3). .. This is the process of step S5 of FIG.

The positional deviation data processing unit 7021 performs a predetermined calculation in units of a second predetermined number (for example, 21) of frames on the first difference data (graph G4) to obtain the first variation data. (Graph G6) is calculated. This is the process of step S3 in FIG. The positional deviation data processing unit 7021 performs a predetermined calculation in units of a fourth predetermined number (for example, 21) of frames on the second difference data (graph G5) to obtain the second variation data. (Graph G7) is calculated. This is the process of step S6 of FIG.

The positional deviation data processing unit 7021 calculates the difference between the first variation data (graph G6) and the second variation data (graph G7) obtained from the same time series pixel data, and obtains third difference data. (Graph G8) is calculated. This is the process of step S7 in FIG. The third difference data (graph G8) is the lateral vibration data D_L.

The positional shift data processing unit 7021 similarly generates vertical vibration data D_V based on the vertical positional shift data d_V.

As described above, the vibration data D is generated by performing the same signal processing (signal processing on time series data) as the image processing of gas region extraction on the positional deviation data d that is time series data. .. According to the embodiment, the vibration data D is data in which vibrations that hinder gas region extraction are extracted. The determination accuracy can be improved by determining whether or not the infrared camera 2 can be installed based on this data.

When the frame rate of the visible light camera 6 is higher than the frame rate of the infrared camera 2, there is a possibility that the vibration data D also includes vibrations that do not hinder gas region extraction. Therefore, the frame rate of the visible light camera 6 is preferably the same as the frame rate of the infrared camera 2. Therefore, when the frame rate of the visible light camera 6 is higher than the frame rate of the infrared camera 2, the vibration data generation unit 702 performs the process of making the frame rate of the moving image V2 the same as the frame rate of the infrared camera 2, and then the moving image V2. The displacement data d may be generated based on The process of making the frame rates the same may be, for example, a process of thinning out the frames forming the moving image V2 or a process of obtaining a moving average of the frames forming the moving image V2.

The vibration data generation unit 702 generates horizontal vibration data D_L and vertical vibration data D_V as the vibration data D. The vibration data generation unit 702 may generate the vibration data D (representative vibration data) based on these. For example, the vibration data generation unit 702 generates the vibration data D (representative vibration data) by summing the squares of the horizontal vibration data D_L and the vertical vibration data D_V at the same time, and obtaining the square root of the square sum. Alternatively, the vibration data D (representative vibration data) may be generated by comparing the horizontal vibration data D_L and the vertical vibration data D_V at the same time and selecting the larger data. ..

The vibration threshold Th (an example of a threshold) will be described. The storage unit 704 (FIG. 2A) stores the vibration threshold Th in advance. The vibration threshold Th is represented by the following Expression 5.
Th=L/g...Equation 5

The description will start from L. L is represented by the following Equation 6.
L=S/κ...Equation 6
S indicates a gas detection threshold value. κ indicates a safety factor. The safety factor κ is a ratio for distinguishing the gas detection signal S from the signal generated by vibration. As the safety factor κ increases, the false detection of gas decreases.

The gas detection threshold value S is expressed by Equation 7 below.
S=γ{P(T+ΔK)-P(T)}...Equation 7
γ indicates the light absorption rate in the concentration-thickness product of the detection gas in the gas detection specifications. P(T) indicates the infrared brightness corresponding to the black body radiance under the detected gas temperature condition in the gas detection specifications. P(T+ΔK) indicates the infrared brightness corresponding to the black body radiance under the background temperature condition in the gas detection specifications.

γ, P(T), and P(T+ΔK) are values that depend on the gas concentration/thickness product, the gas temperature, and the background temperature, respectively. Therefore, in order to determine the gas detection threshold value S, a predetermined value is required for each. As the values determined by the camera detection specifications, the respective predetermined values are, for example, 10% LELm (=0.005m), 300K, and 303K.

Next, g will be described. The vibration amount h of the infrared camera 2 is expressed by Equation 8 below.
h=(θ2/θ1)×D...Equation 8
θ1 represents the angle of view per pixel of the infrared camera 2. θ2 indicates the angle of view per pixel of the visible light camera 6. D indicates the vibration data generated by the vibration data generation unit 702.

When the infrared camera 2 vibrates, each pixel of the infrared camera 2 (each pixel of the two-dimensional image sensor 20 (FIG. 1)) vibrates. As a result, in each frame forming the moving image V1, the amount of infrared rays (luminance) detected by the pixels forming the edge of the image captured in the frame changes. The maximum value of the infrared ray amount (luminance) detected by one pixel forming the edge due to the vibration is Imax, and the minimum value thereof is Imin. The amount of change in the luminance of this pixel is h×(Imax-Imin). Assuming the maximum value of the brightness change amount, the maximum brightness value is Imax and the minimum brightness value is Imin in the range of brightness values that can be received by each pixel of the infrared camera 2. Accordingly, the variation amount C of the brightness of each pixel of the infrared camera 2 can be expressed by the following Expression 9.
C=h×(Imax−Imin)...Equation 9

When Imax-Imin is β, Expression 9 is expressed by Expression 10 below.
C=h×β...Equation 10

The following expression 11 is derived from the expressions 8 and 10.
C={(θ2/θ1)×D}×β... Equation 11

g is defined by the following Expression 12 included on the right side of Expression 11.
g=β×(θ2/θ1)...Equation 12

The following expression 13 is derived from the expressions 11 and 12.
C=g×D...Equation 13

C is the amount of change in luminance caused by vibration. If this value is smaller than the value obtained by dividing the gas detection threshold value S by the safety factor κ, that is, L, gas detection is possible, and it is determined that the infrared camera 2 can be installed.
g×D<L...Equation 14

Divide both sides of Equation 14 by g.
D<L/g...Equation 15
Since the vibration data D satisfying Expression 15 becomes the vibration threshold Th, the vibration threshold Th is represented by Expression 5.

The vibration detection method according to the embodiment will be described. FIG. 6 is a flowchart explaining this. With reference to FIG. 1, FIG. 2A and FIG. 6, the user installs the visible light camera 6 at the place where the infrared camera 2 is installed (step S11). The angle of view θ2 per pixel of the visible light camera 6 is smaller than the angle of view θ1 per pixel of the infrared camera 2.

The user operates the operation unit 71 to input a vibration detection command to the vibration detection device 7. As a result, the control processing unit 700 operates the visible light camera 6 to capture the moving image V2 (step S12). The visible light camera 6 transmits the moving image V2 to the vibration detection device 7. The interface unit 701 receives the transmitted moving image V2, and the control processing unit 700 transfers the received moving image V2 to the vibration data generating unit 702.

The vibration data generation unit 702 generates vibration data D based on the moving image V2 (step S13).

The installation determination unit 703 reads the vibration threshold Th from the storage unit 704 and compares it with the vibration data D (step S14). When the vibration data D is larger than the vibration threshold Th (Yes in step S14), the installation determination unit 703 determines that the infrared camera 2 cannot be installed (step S15). The display control unit 705 causes the display unit 72 to display an image indicating that installation is impossible. When the vibration data D is less than or equal to the vibration threshold Th (No in step S14), the installation determination unit 703 determines that the infrared camera 2 can be installed (step S16). The display control unit 705 causes the display unit 72 to display an image indicating that installation is possible.

It should be noted that the user may determine whether or not the infrared camera 2 can be installed by looking at the vibration data D, or the user may make a determination by comparing the vibration threshold Th with the vibration data D.

The main effects of the embodiment will be described. The vibration data D is generated based on the moving image V2 captured by the visible light camera 6 installed at the place where the infrared camera 2 is installed. The visible light camera 6 has a smaller angle of view per pixel than the infrared camera 2. Therefore, the vibration data D can indicate a shift smaller than the one pixel size of the infrared camera 2 (subpixel level shift). Therefore, according to the vibration detecting method according to the embodiment, it is possible to detect the vibration of the infrared camera 2 installed at a predetermined place, which causes a deviation smaller than one pixel size of the camera.

A modified example of the vibration data generation unit 702 will be described. FIG. 7 is a block diagram showing this modification. In the modification, a gain/offset processing unit is provided instead of the position shift data processing unit 7021. The positional shift data processing unit 7021 performs the same processing as the image processing shown in FIG. 4 on the positional shift data d to generate vibration data D. On the other hand, the gain/offset processing unit uses the following Expression 16 to generate the vibration data D from the displacement data d (generates the lateral vibration data D_L from the lateral displacement data d_L, and the vertical direction). Vertical vibration data D_V is generated from the positional deviation data d_V).
D=ad+b... Equation 16
a shows a gain. b indicates an offset.

The gain a and the offset b will be specifically described. The positional deviation data d is calculated based on the angle of view θ2 per pixel of the visible light camera 6, and therefore indicates the amount of deviation in units of pixels of the visible light camera 6. For example, if the shift amount is 2.5, it means a shift of 2.5 pixels of the visible light camera 6. Suppose that there is a case where it is desired to obtain the vibration data D indicating the amount of deviation in units of pixels of the infrared camera 2 (the amount of deviation in units of pixels of the infrared camera 2 understands the amount of deviation due to vibration of the infrared camera 2). Easy). In this case, the gain/offset processing unit substitutes a=θ1/θ2 and b=0 into the equation 16 to generate the vibration data D from the displacement data d.

It is assumed that there is a case where the displacement data d is desired to be the vibration data D. The gain/offset processing unit substitutes a=1 and b=0 into the equation 16 to generate the vibration data D from the displacement data d.

Suppose there is a case where the vibration amplitude for a specific period is desired to be obtained. The gain/offset processing unit first substitutes b=0 into Expression 16 to generate vibration data D from the positional deviation data d (D=ad). The gain/offset processing unit then calculates the average value m of the specific period for this vibration data D and substitutes b=−m into Equation 16 to obtain the vibration data D (D=ad−m). To generate. The vibration data D indicates the vibration amplitude of the specific period. In this case, the offset b will be changed by the position shift data d.

The positional deviation data d usually indicates both positive and negative values. It is assumed that the vibration data D may be desired to have a positive value. The gain/offset processing unit substitutes into Expression 14 the value that is equal to or larger than the value obtained by subtracting the minimum value of ad from 0, and substitutes the value into Expression 14 to generate the vibration data D from the displacement data d. Also in this case, the offset b is changed by the positional shift data d.

D Vibration data D_L Horizontal vibration data D_V Vertical vibration data d Position shift data d_L Horizontal position shift data d_V Vertical position shift data R1 Infrared camera shooting range R2 Visible light camera shooting range T Shooting target V1 Infrared camera shooting Video V2 Video taken by visible light camera θ1 Angle of view per pixel of infrared camera θ2 Angle of view per pixel of visible light camera

Claims (5)

A method for detecting vibration acting on a first camera installed at a predetermined place, comprising:
A photographing step of photographing a time-series image using the second camera in a state where a second camera whose angle of view per pixel is smaller than the first camera is installed at the predetermined place;
A vibration detecting method, comprising: generating vibration data indicating vibration acting on the second camera based on the time-series image captured by the second camera. The vibration detection method according to claim 1, wherein the following expression holds with the second camera in which the angle of view per pixel is smaller than that of the first camera.
Angle of view of second camera <(angle of view of first camera x pixel number ratio)/number of magnifications Here, the pixel number ratio is the number of pixels in the horizontal direction of the second camera/the lateral direction of the first camera Or the number of pixels in the vertical direction of the second camera/the number of pixels in the vertical direction of the first camera, and the scaling factor indicates a value greater than 1. In the generating step, position shift data indicating the amount of position shift of the images captured in the time-series images captured by the second camera in time series is generated, and captured by the first camera. 3. The vibration data is generated by performing signal processing corresponding to image processing for image recognition performed on a time series image, which is performed on the position shift data, to generate the vibration data. Vibration detection method. The vibration detection method according to claim 3, wherein the image processing is extraction processing of a gas region. The method further comprises a determination step of comparing the vibration data with a threshold value determined based on a signal indicating the gas region and determining whether or not the first camera can be installed after the generation step. The vibration detection method according to item 4.