Fire Detection Systems

Abstract

To provide a fire probe system with which it is possible to suppress the misrecognition of a fire attributable to sunlight.SOLUTION: The fire probe system comprises: an infrared camera 12 for capturing an inside image of a range in charge so as to acquire detection information that includes temperature information corresponding to each of a plurality of pixels; and a controller 11 for executing an identification process of identifying a fire source position inside of the range in charge on the basis of the detection information. The controller determines whether or not the time-series data regarding a feature quantity calculated from the detection information having been acquired by the infrared camera indicates a degree of dispersion within a prescribed range that is comparable to a cyclical fluctuation characteristic unique to a flame, and thereby suppresses misrecognition of a fire attributable to sunlight.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] FIELD OF THE DISCLOSURE The present disclosure relates to a fire detection system for locating a fire source within a fire monitoring area. [Background technology]

[0002] There is a system in which a fire detection device is placed in each compartment in a large space and the location of a fire source is calculated (see, for example, Patent Document 1). The fire detection system in Patent Document 1 has a configuration in which multiple detection devices are placed within a fire monitoring range and the fire source location data identified by the detection device located closest to the same fire source location is used.

[0003] By being provided with such a configuration, the fire exploration system according to Patent Document 1 can prevent the accuracy of calculating the fire source position from deteriorating and can also prevent the time required to calculate the fire source position from increasing. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2020-119014 Summary of the Invention [Problem to be solved by the invention]

[0005] Domed baseball stadiums are an example of large spaces where fire monitoring is required. Some domed baseball stadiums use natural turf. In such cases, a roof that can be opened and closed to allow sunlight in may be used to help the natural turf grow.

[0006] When fire monitoring is performed with the ceiling open, there is a risk that reflected sunlight may be mistaken for a fire. If a fire is mistakenly recognized, it may not only trigger an alarm to notify the fire, but also trigger water spraying to extinguish the fire. Therefore, in order to improve the reliability of fire detection systems, it is important to suppress false recognition of fires caused by sunlight and to avoid unnecessary alarms or water spraying.

[0007] In addition, reducing false positives is a common issue not only in dome stadiums but also in various fire monitoring areas. Therefore, in fire monitoring areas affected by reflected sunlight, there is a strong demand for a fire detection system that reduces false positives caused by sunlight.

[0008] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a fire detection system that can suppress false recognition of fires caused by sunlight. [Means for solving the problem]

[0009] The fire detection system according to the present disclosure includes an infrared camera that captures an image of a pre-assigned assigned range within a fire monitoring range to obtain detection information including temperature information corresponding to each of a plurality of pixels, and a controller that executes an identification process to identify the location of a fire source within the assigned range based on the detection information obtained by the infrared camera, and the controller performs a first process to compare the temperature information of each pixel within the assigned range with a pre-set threshold temperature to identify a high temperature state, and when a pixel area having a temperature equal to or higher than the threshold temperature is identified, From each image acquired in time series from the infrared camera, The same field of view including the specified pixel area The area is extracted, and the multiple data obtained are treated as time series data. a second process of Each extracted area included in the time series data The highest temperature is extracted at In the same visual field Chronological Na If the statistical value obtained for the maximum temperature indicates a degree of variation within a predetermined range, a third process is executed to identify the pixel region as the fire source position, thereby executing the identification process. The fire detection system according to the present disclosure also includes an infrared camera that acquires detection information including temperature information corresponding to each of a plurality of pixels by capturing an image of a pre-assigned assigned range within a fire monitoring range, and a controller that executes an identification process to identify the location of a fire source within the assigned range based on the detection information acquired by the infrared camera, and the controller performs a first process to compare the temperature information of each pixel within the assigned range with a pre-set threshold temperature to identify a high temperature state, and when a pixel area having a temperature equal to or higher than the threshold temperature is identified, From each image acquired in time series from the infrared camera, The same field of view including the specified pixel area The area is extracted, and the multiple data obtained are treated as time series data. a second process of Each extracted area included in the time series data A circumscribing rectangle of a pixel region having a temperature equal to or higher than the threshold temperature is obtained in In the same visual field Chronological Na circumscribed rectangle each If the statistical value calculated for the number of pixels constituting the pixel region indicates a degree of variation within a predetermined range, a third process is executed to identify the pixel region as the fire source position. Furthermore, the fire detection system according to the present disclosure includes an infrared camera that acquires detection information including temperature information corresponding to each of a plurality of pixels by capturing an image of a pre-assigned assigned range within the fire monitoring range, and a controller that executes an identification process to identify the location of a fire source within the assigned range based on the detection information acquired by the infrared camera, and the controller performs a first process to compare the temperature information of each pixel within the assigned range with a pre-set threshold temperature to identify a high temperature state, and when a pixel area having a temperature equal to or higher than the threshold temperature is identified, From each image acquired in time series from the infrared camera, The same field of view including the specified pixel area The area is extracted, and the multiple data obtained are treated as time series data. a second process of Each extracted area included in the time series data The fire source position showing the highest temperature Identify death, In the same visual field Chronological Location of the fire source If the statistical values ​​obtained for the first and second pixels indicate a degree of variation within a predetermined range, a third process is executed to identify the pixel region as the fire source position. [Effects of the Invention]

[0010] According to the present disclosure, a fire detection system can be obtained that can suppress false recognition of fires caused by sunlight. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is an overall configuration diagram of a fire exploration system according to a first embodiment of the present disclosure. [Figure 2] 3 is an explanatory diagram relating to an instantaneous monitoring range in the vertical direction and an instantaneous monitoring range in the horizontal direction of the infrared camera according to the first embodiment of the present disclosure. FIG. [Figure 3] 3 is an explanatory diagram illustrating a specific process 1 executed by a controller according to the first embodiment of the present disclosure. FIG. [Figure 4] 4 is an explanatory diagram showing a specific process 2 executed by a controller according to the first embodiment of the present disclosure. FIG. [Figure 5] 10 is an explanatory diagram showing a specific process 3 executed by a controller according to the first embodiment of the present disclosure. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0012] The present disclosure relates to a fire detection system, and more particularly to a fire detection system for detecting fires caused by sunlight, which is a fire that has been detected by an infrared camera.

[0013] Embodiment 1 Fig. 1 is an overall configuration diagram of a fire exploration system according to a first embodiment of the present disclosure. The fire exploration system shown in Fig. 1 is configured to include one exploration device 10. The exploration device 10 is arranged to explore the location of a fire source within its assigned coverage area.

[0014] The exploration device 10 is configured to include a controller 11, an infrared camera 12, and a drive mechanism 13 for moving the imaging area of ​​the infrared camera 12 and changing the imaging position.

[0015] The number of detection devices 10 is not limited to one, and a configuration using a plurality of detection devices arranged to detect the fire source location within each assigned area in the fire monitoring area may be adopted. When adopting such a configuration, it is conceivable to add an overall controller that receives detection information for each assigned area from the controller 11 in each detection device 10 and monitors the entire fire monitoring area.

[0016] However, the fire detection system according to the present disclosure has a technical feature of suppressing erroneous recognition of fires caused by sunlight by executing a process of identifying the actual fire cause and non-fire cause caused by sunlight by each controller 11 based on the time-series data of the detection results of the infrared camera 12. Therefore, the following description will explain in detail the process of identifying the fire caused by one detection device 10.

[0017] The infrared camera 12 has a function of capturing an image of its assigned area and acquiring detection information including temperature information for each of a plurality of pixels. The controller 11 executes a process for identifying the fire source position within its assigned area based on the detection information acquired by the infrared camera 12. The drive mechanism 13 will be described later.

[0018] Next, a specific example of the viewing angle of infrared camera 12 and the monitoring area associated with the movement of infrared camera 12 using drive mechanism 13 will be described with reference to Fig. 2. Fig. 2 is an explanatory diagram of an instantaneous monitoring range AV in the vertical direction and an instantaneous monitoring range AH ​​in the horizontal direction of infrared camera 12 according to the first embodiment of the present disclosure.

[0019] Here, the instantaneous vertical monitoring range AV corresponds to the vertical viewing range, i.e., the vertical viewing angle, when the infrared camera 12 is fixed. The specific example in the upper part of Figure 2 illustrates a case where the vertical viewing angle is 37.0 degrees.

[0020] The instantaneous horizontal monitoring range AH ​​corresponds to the horizontal viewing range, i.e., the horizontal viewing angle, when the infrared camera 12 is fixed. The specific example in the lower part of Figure 2 illustrates a case where the horizontal viewing angle is 50.0 degrees.

[0021] By performing fire detection using the infrared camera 12, which has an instantaneous field of view of 50.0 degrees horizontally and 37.0 degrees vertically, and positioning it at a total of 12 positions using the pan head of the drive mechanism 13, for example, as shown below, it is possible to monitor fires over a wide monitoring area with just one unit.

[0022] (1) Fix the head elevation angle at -19.5 degrees from horizontal and rotate it horizontally to search for the fire source in four positions. (2) Next, fix the elevation angle of the platform at -47.5° from the horizontal and rotate it horizontally to search for the fire source in four positions. (3) Next, fix the elevation angle of the platform at -75.5° from the horizontal and rotate it horizontally to search for the fire source in four positions.

[0023] Next, a detailed description will be given of the process of identifying the actual cause of a fire and the non-fire cause caused by sunlight based on the detection information acquired by the infrared camera 12.

[0024] Flames, which are the original cause of fires, are characterized by periodic fluctuations over time.On the other hand, non-fire causes caused by sunlight are assumed to be incident on the infrared camera 12 as reflected sunlight, and do not have the characteristic of periodic fluctuations over time that is unique to flames.

[0025] Therefore, the controller 11 performs a specific process on the detection information acquired by the infrared camera 12 to detect whether or not there are characteristics related to periodic fluctuations, and can distinguish between the actual fire causes and non-fire causes caused by sunlight based on the detection information acquired by the infrared camera 12.

[0026] Therefore, three specific identification processes for detecting characteristics related to periodic fluctuations will be described below. Note that, regardless of which identification process is performed, the controller 11 has a common technical feature of detecting characteristics related to periodic fluctuations using time-series data of detection information acquired by the infrared camera 12.

[0027] <Specification process 1: A method to detect periodic fluctuations by focusing on the maximum temperature> The controller 11 extracts the maximum temperature from each piece of time-series data as a feature amount, and can determine whether or not there is a periodic fluctuation characteristic of a flame from the degree of time-series variation in the maximum temperature.

[0028] 3A, 3B, and 3C are explanatory diagrams showing specific processing 1 executed by controller 11 according to embodiment 1 of the present disclosure. 3A, 3B, and 3C respectively show images G1, G2, and G3 acquired as time-series data by infrared camera 12 at a predetermined sampling period.

[0029] Specific processing in the identification process 1 will be described with reference to Fig. 3. The controller 11 performs the identification process 1 by sequentially executing the following first to third processes.

[0030] (1) First treatment The controller 11 compares the temperature information of each pixel in the image G1 acquired from the infrared camera 12 with a threshold temperature set in advance to identify a high temperature state.

[0031] (2) Second Processing When the controller 11 identifies a pixel area in image G1 that has a temperature above the threshold temperature, it extracts an area of ​​the same field of view that includes the identified pixel area from each of images G2 and G3, which were acquired in chronological order following image G1.

[0032] In the examples of Figures 3(A) to 3(C), a pixel area in image G1 having a temperature above the threshold temperature is identified as area S11 shown in a white frame in Figure 3(A), and then areas S12 and S13 shown in the same white frame are extracted in chronological order from images G2 and G3, respectively, as images containing the same pixel area.

[0033] (3) Third treatment The controller 11 extracts the maximum temperatures in the white framed regions S11 to S13 in the images G1 to G3, which are time-series data. In the example of Fig. 3(A) to Fig. 3(C), 90.6 degrees is extracted as the maximum temperature in the region S11 of the image G1 in Fig. 3(A), 110.9 degrees is extracted as the maximum temperature in the region S12 of the image G2 in Fig. 3(B), and 110.9 degrees is extracted as the maximum temperature in the region S13 of the image G3 in Fig. 3(C). 85.5 10 shows the case where the degree is extracted.

[0034] Furthermore, if the maximum temperatures extracted in a time series show a degree of variation within a predetermined range, the controller 11 specifies the pixel area specified in the second process as the fire source position. For example, the controller 11 calculates statistics such as the variance value and the difference value between the maximum temperature and the minimum temperature for the maximum temperatures extracted from each of the three images G1 to G3, and compares them with a preset threshold value, thereby quantitatively determining the degree of variation in the maximum temperatures.

[0035] Flames, which are the cause of fires that we want to detect, have the characteristic of fluctuating periodically. Therefore, when a flame is detected in the above-mentioned identification process 1, the degree of variation over time of the maximum temperature, which is a feature, will indicate a degree of variation within a predetermined range. On the other hand, non-fire causes caused by sunlight do not have the characteristic of fluctuating periodically like a flame.

[0036] Therefore, the controller 11 can distinguish between the actual cause of a fire and a non-fire cause caused by sunlight, based on the degree of time-series variation in the maximum temperature, which is a feature amount.

[0037] <Identification process 2: A method for detecting periodic fluctuations by focusing on pixel regions with temperatures above a threshold temperature> The controller 11 extracts pixel areas above a predetermined threshold temperature as features to identify high temperature conditions for each piece of time series data, and can determine whether or not there are periodic fluctuations characteristic of a flame from the degree of time series variation in pixel areas above the threshold temperature.

[0038] 4 is an explanatory diagram showing specific process 2 executed by controller 11 according to embodiment 1 of the present disclosure. Fig. 4 shows images G1 to G4 acquired as time-series data by infrared camera 12 at times T1 to T4 for each predetermined sampling period, respectively.

[0039] Specific processing in the identification process 2 will be described with reference to Fig. 4. The controller 11 performs the identification process 2 by sequentially executing the following first to third processes.

[0040] (1) First treatment The controller 11 compares the temperature information for each pixel in the image G1 acquired from the infrared camera 12 with a threshold temperature that is set in advance to identify a high temperature state.

[0041] (2) Second Processing When the controller 11 identifies a pixel area in image G1 that has a temperature above the threshold temperature, it extracts an area of ​​the same field of view that includes the identified pixel area from each of images G2 to G4, which are acquired in chronological order following image G1.

[0042] The example in Figure 4 shows that after a pixel area having a temperature above a threshold temperature is identified in image G1, time series data having the same field of view is extracted from each of images G2, G3, and G4 as images containing the identified pixel area.

[0043] (3) Third treatment The controller 11 determines a circumscribing rectangle of a pixel area having a temperature equal to or higher than the threshold temperature in each of the images G1 to G4, which are time-series data. In each of the images G1 to G4, the pixels shown in white correspond to pixels having a temperature lower than the threshold temperature.

[0044] On the other hand, pixels shown in multiple shades of gray other than white correspond to pixels having a temperature above the threshold temperature, with darker gray indicating higher temperatures. Furthermore, pixel regions S21 to S24 shown in bold frames in each of images G1 to G4 correspond to circumscribing rectangles.

[0045] The example in Figure 4 shows a case where a pixel area S21, which is a circumscribing rectangle of 7 pixels x 3 pixels, is extracted from image G1, a pixel area S22, which is a circumscribing rectangle of 4 pixels x 3 pixels, is extracted from image G2, a pixel area S23, which is a circumscribing rectangle of 5 pixels x 3 pixels, is extracted from image G3, and a pixel area S24, which is a circumscribing rectangle of 7 pixels x 3 pixels, is extracted from image G4.

[0046] Furthermore, if the bounding rectangles extracted in time series show a degree of variation within a predetermined range, the controller 11 identifies the pixel region identified in the second process as the fire source position. For example, the controller 11 calculates statistics such as the variance value and the difference between the maximum and minimum number of pixels for the number of pixels constituting the bounding rectangles extracted from each of the four images G1 to G4, and compares these with a preset threshold to quantitatively determine the degree of variation of the bounding rectangles. The degree of variation of the bounding rectangles can also be quantitatively determined by calculating the position of the center of gravity of the bounding rectangle as a statistical quantity.

[0047] Flames, which are the fire causes we want to detect, have the characteristic of fluctuating periodically. Therefore, when a flame is detected in the above-mentioned identification process 2, the degree of variation over time of the circumscribed rectangle, which is a feature, will indicate a degree of variation within a predetermined range. On the other hand, non-fire causes caused by sunlight do not have the characteristic of fluctuating periodically like a flame.

[0048] Therefore, the controller 11 can distinguish between the original cause of a fire and a non-fire cause caused by sunlight, based on the degree of time-series variation of the circumscribing rectangle, which is a feature amount.

[0049] <Specification process 3: A method for detecting periodic fluctuations by focusing on the position of the pixel showing the highest temperature> The controller 11 extracts the pixel position showing the highest temperature from the pixel area above a predetermined threshold temperature set in advance to identify a high temperature state for each piece of time series data, and can determine whether or not there are periodic fluctuation characteristics specific to a flame from the degree of time series variation in the pixel position showing the highest temperature.

[0050] 5 is an explanatory diagram showing specific process 3 executed by controller 11 according to embodiment 1 of the present disclosure. Fig. 5 shows images G1 to G4 acquired as time-series data by infrared camera 12 at times T1 to T4 for each predetermined sampling period, respectively.

[0051] Specific processing in the identification process 3 will be described with reference to Fig. 5. The controller 11 performs the identification process 3 by sequentially executing the following first to third processes.

[0052] (1) First process The controller 11 compares the temperature information for each pixel in the image G1 acquired from the infrared camera 12 with a threshold temperature that is set in advance to identify a high temperature state.

[0053] (2) Second Processing When the controller 11 identifies a pixel area in image G1 that has a temperature above the threshold temperature, it extracts an area of ​​the same field of view that includes the identified pixel area from each of images G2 to G4, which are acquired in chronological order following image G1.

[0054] The example in Figure 5 shows that after a pixel area having a temperature above a threshold temperature is identified in image G1, time series data having the same field of view is extracted from each of images G2, G3, and G4 as images containing the identified pixel area.

[0055] (3) Third treatment The controller 11 extracts the fire source position information showing the highest temperature from each of the images G1 to G4, which are time-series data. In each of the images G1 to G4, the pixels shown in white correspond to pixels having a temperature below the threshold temperature.

[0056] On the other hand, pixels shown in gray tones other than white correspond to pixels having a temperature below the threshold temperature, and the darker the gray, the higher the temperature. Furthermore, in each of images G1 to G4, the pixels shown in bold frames correspond to the pixel positions showing the highest temperatures.

[0057] In the example of FIG. 5, pixel position P1 is extracted from image G1, pixel position P2 is extracted from image G2, pixel position P3 is extracted from image G3, and pixel position P4 is extracted from image G4.

[0058] Furthermore, if the pixel positions P1 to P4 extracted in time series show a degree of variation within a predetermined range, the controller 11 specifies the pixel region specified in the second process as the fire source position. For example, the controller 11 calculates a statistical quantity such as the variance of the distance from the position of the reference pixel to each of pixel positions P1 to P4 for pixel positions P1 to P4 extracted from each of the four images G1 to G4, and compares it with a preset threshold value, thereby making it possible to quantitatively determine the degree of variation in the pixel positions showing the highest temperatures.

[0059] Flames, which are the cause of fires that we want to detect, have the characteristic of fluctuating periodically. Therefore, when a flame is detected in the above-mentioned identification process 3, the degree of chronological variation in pixel positions indicating the maximum temperature, which is a feature, will show a degree of variation within a predetermined range. On the other hand, non-fire causes caused by sunlight do not have the characteristic of fluctuating periodically like a flame.

[0060] Therefore, the controller 11 can distinguish between the actual cause of a fire and a non-fire cause caused by sunlight, based on the degree of time-series variation in the pixel position indicating the maximum temperature, which is a feature amount.

[0061] As described above, according to the first embodiment, by determining whether the degree of variation over time of the maximum temperature, the degree of variation over time of the circumscribing rectangle, or the degree of variation over time of the pixel position indicating the maximum temperature exhibits a degree of variation within a predetermined range comparable to the periodic fluctuation characteristics unique to a flame, it is possible to suppress the misidentification of a fire caused by sunlight.

[0062] It is also possible to simultaneously perform two or more of the above-described identification processes 1 to 3 to distinguish between the actual cause of the fire and a non-fire cause caused by sunlight. [Explanation of symbols]

[0063] 10. Exploration device, 11. Controller, 12. Infrared camera, 13. Drive mechanism.

Claims

1. an infrared camera that captures images of a pre-assigned area within the fire monitoring range and obtains detection information including temperature information corresponding to each of a plurality of pixels; a controller that executes a process for identifying a fire source position within the coverage area based on the detection information acquired by the infrared camera; Equipped with The controller a first process of comparing the temperature information of each pixel in the assigned range with a threshold temperature set in advance to identify a high temperature state; a second process of extracting a region of the same field of view including the identified pixel region from each of the images acquired in time series from the infrared camera when the pixel region having a temperature equal to or higher than the threshold temperature is identified, and treating the plurality of extracted data as time series data; a third process of extracting a maximum temperature in each extracted region included in the time-series data, and specifying the pixel region as the fire source position when a statistical value obtained for the time-series maximum temperature in the same visual field region shows a degree of variation within a predetermined range; A fire detection system that executes the specific processing by sequentially executing the above steps.

2. an infrared camera that captures images of a pre-assigned area within the fire monitoring range and obtains detection information including temperature information corresponding to each of a plurality of pixels; a controller that executes a process for identifying a fire source position within the coverage area based on the detection information acquired by the infrared camera; Equipped with The controller a first process of comparing the temperature information of each pixel in the assigned range with a threshold temperature set in advance to identify a high temperature state; a second process of extracting a region of the same field of view including the identified pixel region from each of the images acquired in time series from the infrared camera when the pixel region having a temperature equal to or higher than the threshold temperature is identified, and treating the plurality of extracted data as time series data; a third process of determining a circumscribing rectangle of a pixel area having a temperature equal to or higher than the threshold temperature in each extracted area included in the time-series data, and specifying the pixel area as the fire source position when a statistical value determined for the number of pixels constituting each of the time-series circumscribing rectangles in the same field of view area shows a degree of variation within a predetermined range; A fire detection system that executes the specific processing by sequentially executing the above steps.

3. an infrared camera that captures images of a pre-assigned area within the fire monitoring range and obtains detection information including temperature information corresponding to each of a plurality of pixels; a controller that executes a process for identifying a fire source position within the coverage area based on the detection information acquired by the infrared camera; Equipped with The controller a first process of comparing the temperature information of each pixel in the assigned range with a threshold temperature set in advance to identify a high temperature state; a second process of extracting a region of the same field of view including the identified pixel region from each of the images acquired in time series from the infrared camera when the pixel region having a temperature equal to or higher than the threshold temperature is identified, and treating the plurality of extracted data as time series data; a third process of identifying a fire source position showing the highest temperature in each extracted region included in the time-series data, and identifying the pixel region as the fire source position when a statistical quantity obtained for the time-series fire source positions in the same field of view region shows a degree of variation within a predetermined range; A fire detection system that executes the specific processing by sequentially executing the above steps.

Images