Intelligent fire water monitor control system and method based on visual identification

Abstract

The invention discloses an intelligent fire water monitor control system and method based on visual identification, and relates to the technical field of fire water monitor control, and the method comprises the steps: analyzing the fire presentation effect of an infrared image on a target region, and obtaining a target infrared image; acquiring historical fire data of other areas where the fire occurs in the historical period, and analyzing the fire approximation degree between the target area and other areas to obtain other target areas; acquiring historical equipment control records of the fire water monitors in other areas of the target, acquiring historical fire extinguishing records of other areas of the target, and evaluating the extinguishing effect of the intelligent fire water monitors on the fire in the target area under different control parameters to obtain target control data; based on the target control data, control parameters of the intelligent fire water monitor are adjusted, the intelligent fire water monitor is used for extinguishing fire in the target area, the infrared thermal imager is used for shooting the target area, and the intelligent fire water monitor is controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of fire monitor control, and specifically to an intelligent fire monitor control system and method based on visual recognition. Background Art

[0002] In the field of fire protection, using visual recognition technology to monitor fire situations and then extinguish fires has become one of the common fire extinguishing methods. The application of visual recognition technology in fire fighting has the following advantages: 1. Actual monitoring: Visual recognition technology can monitor in real time and identify the occurrence of fires, improving the response speed and extinguishing fires in a timely manner; 2. Precise fire location: Through the analysis of on-site images, visual recognition technology can accurately identify the positions of smoke and flames in a fire; 3. Improvement of fire extinguishing efficiency: An intelligent fire monitor integrated with visual recognition technology can automatically track the fire source, intelligently adjust the fire extinguishing angle and water flow, thereby improving the fire extinguishing efficiency.

[0003] In daily life, when a visible light camera detects a fire source, although the captured images have rich color and texture information with high resolution, it is sensitive to strong light, and there are often obstacles or thick smoke at the fire scene. The images taken by the visible light camera may deviate greatly from the actual situation. Therefore, in the current visual recognition process of fires, it is common to use an infrared camera. Even infrared images obtained in harsh environments have good features. However, during the fire extinguishing process of a fire monitor, the infrared image only provides the specific position of the fire source in the fire, and it is impossible to intelligently control the spraying angle, direction, and pressure of the fire monitor based on the captured infrared image. This requires manual adjustment of the fire monitor according to the infrared image, which may not only prolong the fire extinguishing time and cause harm to firefighters, but also may further expand the fire and cause huge losses. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent fire monitor control system and method based on visual recognition to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An intelligent fire monitor control method based on visual recognition, the method comprising:

[0006] Step S100: Obtain the area where a fire occurs in the current period and record it as the target area, take a picture of the target area to obtain an infrared image, analyze the rendering effect of the infrared image on the fire in the target area, and obtain the target infrared image;

[0007] Step S200: Obtain the target infrared image, obtain the environmental data within the target area, obtain the historical fire data of other areas where fires occurred during the historical period, analyze the fire similarity degree between the target area and other areas, and obtain the target other areas;

[0008] Step S300: Obtain the historical equipment control records of the fire fighting water cannons in the target other areas, obtain the historical fire extinguishing records of the target other areas, evaluate the extinguishing effect of the intelligent fire fighting water cannon on the fire in the target area under different control parameters, and obtain the target control data;

[0009] Step S400: Based on the target control data, adjust the control parameters of the intelligent fire fighting water cannon, use the intelligent fire fighting water cannon to extinguish the fire in the target area, and use an infrared thermal imager to photograph the target area to control the intelligent fire fighting water cannon.

[0010] Further, step S100 includes:

[0011] Step S101: Use an infrared thermal imager to photograph the target area where a fire occurred during the current period, obtain several infrared images of the target area and gather them to obtain the target area image set;

[0012] Step S102: Perform Gaussian filtering on the infrared images of the target area, obtain the preset wavelength λ of the infrared thermal imager, obtain the radiation intensity of each pixel point in the infrared image, and calculate the characteristic temperature of each pixel point in the infrared image. Among them, the characteristic temperature T of the a-th pixel point in the infrared image is: a :

[0013]

[0014] Among them, k is the Boltzmann constant; c is the speed of light; h is the Planck constant; I a is the radiation intensity of the a-th pixel point;

[0015] Step S103: When the characteristic temperature T a is greater than the preset temperature threshold, determine that there is a fire source at the a-th pixel point, mark the a-th pixel point, and obtain the average value P of the radiation intensity of the marked pixel points in the infrared image: s :

[0016] Obtain the average value I′ of the radiation intensity of the unmarked pixel points in the infrared image, and calculate the image noise intensity P′ of the infrared image:

[0017]

[0018] Among them, n is the total number of unmarked pixel points in the infrared image; I′ iis the radiation intensity of the i-th unmarked pixel in the infrared image;

[0019] Calculate the image signal-to-noise ratio S = P s / P′, and perform normalization processing on the image signal-to-noise ratio S to obtain S′;

[0020] Step S104: Calculate the fire source area ratio D = B′ sum / B sum where B sum is the total number of pixels in the infrared image, and B′ sum is the total number of marked pixels in the infrared image;

[0021] Step S105: Obtain the median D′ of the fire source area ratios of each infrared image in the target area image set, and calculate the fire presentation score Q of the infrared image for the target area:

[0022]

[0023] where γ 1 and γ 2 are the preset first scoring coefficient and second scoring coefficient respectively;

[0024] Step S106: When the fire presentation score Q is greater than the preset threshold, determine that the infrared image has a presentation effect on the fire in the target area, record the infrared image as the target infrared image, and obtain the target infrared images in the target area image set.

[0025] Further, step S200 includes:

[0026] Step S201: Obtain each target infrared image in the target area image set of the target area;

[0027] Step S202: Obtain the fire fighting platform, and obtain the historical fire data of each other area where a fire occurred within the historical period from the fire fighting platform. The historical fire data includes the data corresponding to various environmental indicators of the other area, and each historical infrared image of the other area when extinguishing was not carried out;

[0028] Step S203: From the target infrared images of the target area, obtain the area where the pixels with fire sources are located and record it as the fire source area. Obtain the fire source areas of each historical infrared image in the other areas, and calculate the fire source approximation value K between the target infrared image and the other areas:

[0029]

[0030] where μ is the average value of the radiation intensity within the fire source area of the target infrared image; μ′ zis the average radiation intensity within the fire source area of the z-th historical infrared image in other regions; m is the total number of historical infrared images in other regions; σ′ is the variance of the radiation intensity within the fire source area of the target infrared image; σ′ z is the variance of the radiation intensity within the fire source area of the z-th historical infrared image; σ z is the covariance of the radiation intensity within the fire source area between the target infrared image and the z-th historical infrared image;

[0031] Step S204: Obtain each target infrared image of the target area and the fire source approximation in other regions, monitor the environment in the target area in the current period to obtain environmental data. The environmental data includes the average values of various environmental indicators in the target area. Based on the environmental data, construct an environmental feature vector F of the target area = {f 1 、f 2 、...、f α}, where f 1 、f 2 、...、f α are the average values of the 1st, 2nd,..., α-th environmental indicators in the target area respectively;

[0032] Step S205: Obtain the characteristic environmental feature vector F′ of other regions, and calculate the fire approximation score R between the target area and other regions:

[0033]

[0034] where η 1 、η 2 are the first coefficient and the second coefficient of the fire score preset respectively, η 1 +η 2 =1, η 1 >0, η 2 >0; K g is the g-th target infrared image of the target area and the fire source approximation in other regions; ε is the total number of target infrared images in the target area;

[0035] Step S206: When the fire approximation score R is greater than the preset fire approximation score threshold, determine that there is a fire approximation between the target area and other regions, record other regions as the target other regions of the target area, and obtain each target other region of the target area;

[0036] The above steps analyze the fire approximation degree between the target area and other areas from two perspectives: the target area itself and the environment. Because in the actual process, for the fire fighting water cannon to achieve the same effect under the same control parameters, environmental influencing factors are also important. For example, wind speed and wind direction will affect the spraying effect of the fire fighting water cannon. Through the analysis from two different perspectives, the other areas obtained are very similar to the target area in terms of both fire situation and environment, providing accurate data support for the subsequent control of the intelligent fire fighting water cannon.

[0037] Further, step S300 includes:

[0038] Step S301: Obtain the device data of the intelligent fire fighting water cannon for fire fighting in the target area. The device data includes the values corresponding to each performance index in the intelligent fire fighting water cannon;

[0039] Step S302: Obtain the historical device data of the fire fighting water cannons in each target other area of the target area. The historical device data includes the values corresponding to each performance index in the fire fighting water cannon;

[0040] Step S303: When the performance indexes of the fire fighting water cannon in a certain target other area of the target area are the same as those of the intelligent fire fighting water cannon, retain the certain target other area; otherwise, eliminate a certain target other area to obtain the retained target other areas in the target area;

[0041] Step S304: Obtain the historical device control records of the fire fighting water cannons in the target other areas, and obtain the data corresponding to each control parameter after the fire fighting water cannon extinguishes the fire from the historical device control records;

[0042] Step S305: Set the unit time duration, obtain the historical fire extinguishing records of the target other areas, and obtain the total number W of pixel points determined to have a fire source in the historical infrared images of the target other areas after the fire fighting water cannon extinguishes the fire for the unit time duration from the historical fire extinguishing records sum , obtain the total number W′ of pixel points determined to have a fire source in the historical infrared images of the target other areas when the fire fighting water cannon does not extinguish the fire sum , calculate the characteristic fire extinguishing value V of the target other area = W′ sum / W sum ;

[0043] Step S306: Obtain the maximum value of the characteristic fire extinguishing values of each target other area of the target area, obtain the data corresponding to each control parameter after the fire fighting water cannon extinguishes the fire in the target other area corresponding to the maximum value, and use it as the target control data of the intelligent fire fighting water cannon in the target area, and determine that the intelligent fire fighting water cannon has the best fire extinguishing effect on the target area under the target control data.

[0044] Further, step S400 includes:

[0045] Step S401: Obtain the target control data of the intelligent fire fighting water cannon, set the control parameters of the intelligent fire fighting water cannon based on the target control data within the current cycle, and use the intelligent fire fighting water cannon to extinguish the fire at the fire point in the target area;

[0046] Step S402: After the intelligent fire fighting water cannon extinguishes the fire in the target area, use the infrared thermal imager to take pictures of the target area every unit time to obtain characteristic infrared images. Based on the fire source position in the characteristic infrared images, adjust the spraying angle and spraying pressure in the intelligent fire fighting water cannon until there are no pixel points with fire sources in the characteristic infrared images, and perform intelligent control on the intelligent fire fighting water cannon.

[0047] In order to better implement the above method, an intelligent fire fighting water cannon control system based on visual recognition is also proposed. The system includes a target infrared image module, a fire approximation analysis module, a target control data module, and an intelligent control module;

[0048] The target infrared image module is used to analyze the fire presentation effect of the target area in the infrared image to obtain the target infrared image;

[0049] The fire approximation analysis module is used to analyze the fire approximation degree between the target area and other areas to obtain the target other areas;

[0050] The target control data module is used to evaluate the fire extinguishing effect of the intelligent fire fighting water cannon on the target area under different control parameters to obtain the target control data;

[0051] The intelligent control module is used to adjust the control parameters of the intelligent fire fighting water cannon according to the target control data, use the intelligent fire fighting water cannon to extinguish the fire in the target area, and use the infrared thermal imager to take pictures of the target area to perform intelligent control on the intelligent fire fighting water cannon.

[0052] Further, the target infrared image module includes a fire presentation scoring unit and a target infrared image unit;

[0053] The fire presentation scoring unit is used to calculate the fire presentation scores of each infrared image of the target area;

[0054] The target infrared image unit is used to analyze the fire presentation effect of each infrared image on the target area based on the fire presentation score to obtain the target infrared image.

[0055] Further, the fire approximation analysis module includes a fire approximation scoring unit and a fire approximation analysis unit;

[0056] The fire approximation scoring unit is used to obtain the historical fire data of each other area and calculate the fire approximation score between the target area and each other area;

[0057] The fire approximation analysis unit is used to analyze the fire approximation degree between the target area and each other area according to the fire approximation score, and obtain the target other area.

[0058] Furthermore, the target control data module includes a characteristic fire extinguishing value unit and a target control data unit;

[0059] The characteristic fire extinguishing value unit is used to calculate the characteristic fire extinguishing value of each target other area of the target area;

[0060] The target control data unit is used to obtain the maximum value of the characteristic fire extinguishing values of each target other area of the target area, and obtain the data corresponding to each control parameter after the fire fighting by the fire fighting water cannon in the target other area corresponding to the maximum value, and use it as the target control data of the intelligent fire fighting water cannon in the target area.

[0061] Furthermore, the intelligent control module includes an intelligent control unit;

[0062] The intelligent control unit is used to set each control parameter of the intelligent fire fighting water cannon according to the target control data, and use the intelligent fire fighting water cannon to extinguish the fire at the fire point in the target area. After the intelligent fire fighting water cannon extinguishes the fire in the target area, the infrared thermal imager is used to take pictures of the target area every unit time to obtain the characteristic infrared image until there are no pixel points with fire sources in the characteristic infrared image, and the intelligent control of the intelligent fire fighting water cannon is carried out.

[0063] Compared with the prior art, the beneficial effects of the present invention are: the present invention realizes the intelligent control of the intelligent fire fighting water cannon. By analyzing the captured infrared image, the target infrared image that can present the fire state in the target area is obtained, and according to the fire approximation degree between the target infrared image and the historical infrared images in other areas, the target other area is obtained. According to the fire extinguishing effect of the fire fighting water cannon in the target other area, the control parameters of the intelligent fire fighting water cannon are adjusted, so as to intelligently control the intelligent fire fighting water cannon to extinguish the fire in the target area, which not only ensures the safety of firefighters' lives, but also realizes the rapid and accurate extinguishment of the fire source. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 is the method flow chart of the intelligent fire fighting water cannon control method based on visual recognition of the present invention;

[0065] Figure 2 is the module schematic diagram of the intelligent fire fighting water cannon control system based on visual recognition of the present invention. Detailed implementation manners

[0066] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0067] Embodiment: As Figure 1 - Figure 2 shown, the present invention provides a technical solution, an intelligent fire fighting water cannon control method based on visual recognition, and the method includes:

[0068] Step S100: Obtain the area where a fire occurs in the current period and record it as the target area, take a picture of the target area to obtain an infrared image, and analyze the fire presentation effect of the target area in the infrared image to obtain a target infrared image;

[0069] Among them, step S100 includes:

[0070] Step S101: Use an infrared thermal imager to take pictures of the target area where a fire occurs in the current period, collect a number of infrared images of the target area, and obtain a target area image set;

[0071] Step S102: Perform Gaussian filtering on the infrared image of the target area, obtain the preset wavelength λ of the infrared thermal imager, obtain the radiation intensity of each pixel point in the infrared image, and calculate the characteristic temperature of each pixel point in the infrared image. Among them, the characteristic temperature T of the a-th pixel point in the infrared image a :

[0072]

[0073] Among them, k is the Boltzmann constant; c is the speed of light; h is the Planck constant; I a is the radiation intensity of the a-th pixel point;

[0074] Step S103: When the characteristic temperature T a is greater than the preset temperature threshold, it is determined that there is a fire source at the a-th pixel point, mark the a-th pixel point, and obtain the average value P of the radiation intensity of the marked pixel points in the infrared image s :

[0075] Obtain the average value I′ of the radiation intensity of the unmarked pixel points in the infrared image, and calculate the image noise intensity P′ of the infrared image:

[0076]

[0077] where n is the total number of unmarked pixel points in the infrared image; I′ i is the radiation intensity of the i-th unmarked pixel point in the infrared image;

[0078] Calculate the image signal-to-noise ratio S = P s / P′ of the infrared image, and perform normalization processing on the image signal-to-noise ratio S to obtain S′;

[0079] Step S104: Calculate the fire source area ratio D = B′ sum / B sum where B sum is the total number of all pixel points in the infrared image, and B′ sum is the total number of marked pixel points in the infrared image;

[0080] For example, the total number B sum of all pixel points in the infrared image is 1000; the total number B′ sum of marked pixel points in the infrared image is 400; calculate the fire source area ratio D = 400 / 1000 = 0.4 in the infrared image;

[0081] Step S105: Obtain the median D′ of the fire source area ratios of each infrared image in the target area image set, and calculate the fire presentation score Q of the infrared image for the target area:

[0082]

[0083] where γ 1 and γ 2 are the preset first scoring coefficient and second scoring coefficient respectively;

[0084] For example, S′ is 0.90; γ 1 and γ 2 are 0.6 and 0.4 respectively; the median D′ of the fire source area ratios of each infrared image in the target area image set is 0.5; D is 0.4;

[0085] Calculate the fire presentation score Q of the infrared image for the target area:

[0086]

[0087] Step S106: When the fire presentation score Q is greater than the preset threshold, determine that the infrared image has a presentation effect on the fire in the target area, record the infrared image as the target infrared image, and obtain the target infrared image in the target area image set;

[0088] Step S200: Obtain the target infrared image, obtain the environmental data within the target area, obtain the historical fire data of other areas where fires occurred during the historical period, analyze the fire approximation degree between the target area and other areas, and obtain the target other areas;

[0089] Among them, step S200 includes:

[0090] Step S201: Obtain each target infrared image in the target area image set of the target area;

[0091] Step S202: Obtain the fire fighting platform, and obtain the historical fire data of each other area where fires occurred during the historical period from the fire fighting platform. The historical fire data includes the data corresponding to various environmental indicators of other areas, and each historical infrared image of other areas when extinguishing was not carried out;

[0092] For example, various environmental indicators include wind speed, wind direction, humidity, etc.;

[0093] Step S203: From the target infrared image of the target area, obtain the area where the pixel points with a fire source are located, and record it as the fire source area. Obtain the fire source areas of each historical infrared image in other areas, and calculate the fire source approximation value K between the target infrared image and other areas:

[0094]

[0095] Among them, μ is the average value of the radiation intensity within the fire source area in the target infrared image; μ′ z is the average value of the radiation intensity within the fire source area of the z-th historical infrared image of other areas; m is the total number of each historical infrared image in other areas; σ′ is the variance of the radiation intensity within the fire source area in the target infrared image; σ′ z is the variance of the radiation intensity within the fire source area of the z-th historical infrared image; σ z is the covariance of the radiation intensity within the fire source area between the target infrared image and the z-th historical infrared image;

[0096] Step S204: Obtain the fire source approximation values of each target infrared image of the target area and other areas, monitor the environment of the target area in the current period, obtain the environmental data. The environmental data includes the average values of various environmental indicators of the target area. Based on the environmental data, construct the environmental feature vector F of the target area = {f 1 、f 2 、...、f α}, where f 1 、f 2 、...、f αare the average values of the 1st, 2nd, …, α-th environmental indicators of the target area respectively;

[0097] Step S205: Obtain the characteristic environmental feature vector F′ of other areas, and calculate the fire approximation score R between the target area and other areas:

[0098]

[0099] where η 1 and η 2 are the first coefficient of the fire score and the second coefficient of the fire score preset respectively, η 1 + η 2 = 1, η 1 > 0, η 2 > 0; K g is the approximation value of the fire source of the g-th target infrared image of the target area to other areas; ε is the total number of target infrared images of the target area;

[0100] Step S206: When the fire approximation score R is greater than the preset fire approximation score threshold, determine that there is a fire approximation between the target area and other areas, and record other areas as the target other areas of the target area, and obtain each target other area of the target area;

[0101] Step S300: Obtain the historical equipment control records of the fire fighting water cannons in the target other areas, obtain the historical fire fighting records of the target other areas, evaluate the extinguishing effect of the intelligent fire fighting water cannons on the fires in the target area under different control parameters, and obtain the target control data;

[0102] Among them, Step S300 includes:

[0103] Step S301: Obtain the equipment data of the intelligent fire fighting water cannons for fire fighting in the target area. The equipment data includes the corresponding values of various performance indicators in the intelligent fire fighting water cannons;

[0104] For example, the various performance indicators include the maximum water pressure, the maximum elevation angle, the maximum unit flow rate, etc.;

[0105] Step S302: Obtain the historical equipment data of the fire fighting water cannons in each target other area of the target area. The historical equipment data includes the corresponding values of various performance indicators in the fire fighting water cannons;

[0106] Step S303: When the performance indicators of the fire fighting water cannons in a certain target other area of the target area are the same as those of the intelligent fire fighting water cannons, retain a certain target other area, otherwise, eliminate a certain target other area, and obtain the target other areas retained in the target area;

[0107] Step S304: Obtain the historical equipment control records of the fire water monitors in the target other areas, and from the historical equipment control records, obtain the data corresponding to the various control parameters after the fire water monitors extinguish the fire.

[0108] Step S305: Set the unit time duration, obtain the historical fire extinguishing records of the target other areas, and from the historical fire extinguishing records, obtain the total number W of the pixel points determined to have a fire source in the historical infrared images of the target other areas after the fire water monitors extinguish the fire for a unit time duration. sum , obtain the total number W′ of the pixel points determined to have a fire source in the historical infrared images of the target other areas when the fire water monitors do not extinguish the fire. sum , calculate the characteristic fire extinguishing value V of the target other areas = W′ sum / W sum ;

[0109] Step S306: Obtain the maximum value of the characteristic fire extinguishing values of the various target other areas in the target area, and obtain the data corresponding to the various control parameters after the fire water monitors extinguish the fire in the target other area corresponding to the maximum value, and use it as the target control data of the intelligent fire water monitors in the target area, and determine that the intelligent fire water monitors have the best fire extinguishing effect on the target area under the target control data.

[0110] For example, the various control parameters include the spraying angle, spraying pressure, etc.

[0111] Step S400: Based on the target control data, adjust the control parameters of the intelligent fire water monitors, use the intelligent fire water monitors to extinguish the fire in the target area, and use an infrared thermal imager to photograph the target area to control the intelligent fire water monitors.

[0112] Among them, Step S400 includes:

[0113] Step S401: Obtain the target control data of the intelligent fire water monitors, and based on the target control data in the current cycle, set the various control parameters of the intelligent fire water monitors, and use the intelligent fire water monitors to extinguish the fire at the fire point in the target area.

[0114] Step S402: After the intelligent fire water monitors extinguish the fire in the target area, use an infrared thermal imager to photograph the target area every unit time duration to obtain characteristic infrared images, and based on the fire source position in the characteristic infrared images, adjust the spraying angle and spraying pressure in the intelligent fire water monitors until there are no pixel points determined to have a fire source in the characteristic infrared images, and perform intelligent control on the intelligent fire water monitors.

[0115] In order to better implement the above method, a smart fire fighting water cannon control system based on visual recognition is also proposed. The system includes a target infrared image module, a fire approximation analysis module, a target control data module, and an intelligent control module;

[0116] The target infrared image module is used to analyze the fire presentation effect of the target area in the infrared image to obtain the target infrared image;

[0117] The fire approximation analysis module is used to analyze the fire approximation degree between the target area and other areas to obtain the target other areas;

[0118] The target control data module is used to evaluate the fire extinguishing effect of the smart fire fighting water cannon on the target area under different control parameters to obtain the target control data;

[0119] The intelligent control module is used to adjust the control parameters of the smart fire fighting water cannon according to the target control data, use the smart fire fighting water cannon to extinguish the fire in the target area, and use an infrared thermal imager to take pictures of the target area to perform intelligent control on the smart fire fighting water cannon;

[0120] Among them, the target infrared image module includes a fire presentation scoring unit and a target infrared image unit;

[0121] The fire presentation scoring unit is used to calculate the fire presentation scores of each infrared image in the target area;

[0122] The target infrared image unit is used to analyze the fire presentation effect of each infrared image on the target area in the target area according to the fire presentation score to obtain the target infrared image;

[0123] Among them, the fire approximation analysis module includes a fire approximation scoring unit and a fire approximation analysis unit;

[0124] The fire approximation scoring unit is used to obtain the historical fire data of each other area and calculate the fire approximation scores between the target area and each other area;

[0125] The fire approximation analysis unit is used to analyze the fire approximation degree between the target area and each other area according to the fire approximation score to obtain the target other areas;

[0126] Among them, the target control data module includes a characteristic fire extinguishing value unit and a target control data unit;

[0127] The characteristic fire extinguishing value unit is used to calculate the characteristic fire extinguishing values of each target other area in the target area;

[0128] A target control data unit is used to obtain the maximum value of the characteristic fire extinguishing values of each target other area in the target area, and in the target other area corresponding to the maximum value, obtain the data corresponding to each control parameter after the fire fighting water monitor extinguishes the fire, and use it as the target control data of the intelligent fire fighting water monitor in the target area;

[0129] Among them, the intelligent control module includes an intelligent control unit;

[0130] The intelligent control unit is used to set each control parameter of the intelligent fire fighting water monitor according to the target control data, and use the intelligent fire fighting water monitor to extinguish the fire at the fire point in the target area. After the intelligent fire fighting water monitor extinguishes the fire in the target area, use the infrared thermal imager to take pictures of the target area every unit time length to obtain characteristic infrared images until there are no pixel points with fire sources in the characteristic infrared images, and perform intelligent control on the intelligent fire fighting water monitor.

[0131] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. An intelligent fire-fighting water cannon control method based on visual recognition, characterized in that: The method comprises: Step S100: acquiring an area where a fire occurs in the current cycle and recording it as a target area, photographing the target area to obtain an infrared image, analyzing the fire presentation effect of the infrared image on the target area, and obtaining a target infrared image; Step S200: acquiring the target infrared image, acquiring environmental data in the target area, acquiring historical fire data of other areas where fires occurred in historical periods, analyzing the fire similarity between the target area and the other areas, and obtaining other target areas; Step S300: Obtain historical equipment control records of fire monitors in other target areas, obtain historical fire extinguishing records of other target areas, evaluate the extinguishing effect of the intelligent fire monitors on the fire in the target area under different control parameters, and obtain target control data; Step S400: Based on the target control data, the control parameters of the intelligent fire water monitor are adjusted, the intelligent fire water monitor is used to extinguish the fire in the target area, and the target area is photographed using an infrared thermal imager to control the intelligent fire water monitor.

2. The intelligent fire-fighting water monitor control method based on visual recognition according to claim 1 is characterized in that: The step S100 includes: Step S101: using an infrared thermal imager to photograph a target area where a fire occurs in a current cycle, obtaining and collecting a plurality of infrared images of the target area to obtain a target area image set; Step S102: Gaussian filtering is performed on the infrared image of the target area to obtain the wavelength λ preset by the infrared thermal imager, obtain the radiation intensity of each pixel in the infrared image, and calculate the characteristic temperature of each pixel in the infrared image, wherein the characteristic temperature T of the ath pixel in the infrared image is a : Where k is the Boltzmann constant; c is the speed of light; h is the Planck constant; I a is the radiation intensity of the a-th pixel; Step S103: When the characteristic temperature T a If the temperature is greater than a preset temperature threshold, it is determined that there is a fire source at the a-th pixel point, the a-th pixel point is marked, and the average radiation intensity P of the marked pixel points in the infrared image is obtained. s : The average value I' of the radiation intensity of the unmarked pixels in the infrared image is obtained, and the image noise intensity P' of the infrared image is calculated: Where n is the total number of unmarked pixels in the infrared image; I′ i is the radiation intensity of the unmarked i-th pixel in the infrared image; Calculate the image signal-to-noise ratio S=P of the infrared image s / P′, and normalize the image signal-to-noise ratio S to obtain S′; Step S104: Calculate the fire source area ratio D=B′ in the infrared image sum / B sum , where B sum is the total number of pixels in the infrared image, B′ sum is the total number of marked pixels in the infrared image; Step S105: Obtain the median D′ of the fire source area ratio of each infrared image in the target area image set, and calculate the fire presentation score Q of the infrared image to the target area: Among them, γ1 and γ2 are the preset first scoring coefficient and second scoring coefficient respectively; Step S106: When the fire presentation score Q is greater than a preset threshold, it is determined that the infrared image has a presentation effect on the fire in the target area, the infrared image is recorded as a target infrared image, and the target infrared image in the target area image set is obtained.

3. The intelligent fire-fighting water monitor control method based on visual recognition according to claim 2 is characterized in that: The step S200 includes: Step S201: Acquire each target infrared image in the target area image set of the target area; Step S202: Obtain a fire fighting platform, and obtain historical fire data of various other areas where fires occurred in a historical period from the fire fighting platform, wherein the historical fire data includes data corresponding to various environmental indicators of other areas, and various historical infrared images of other areas when no fire extinguishing is performed; Step S203: From the target infrared image of the target area, obtain the area where the pixels of the fire source are located, and record it as the fire source area, obtain the fire source areas of each historical infrared image in other areas, and calculate the fire source approximation K between the target infrared image and the other areas: Wherein, μ is the average value of the radiation intensity in the fire source area in the target infrared image; μ′ z is the average value of the radiation intensity in the fire source area of ​​the zth historical infrared image of the other area; m is the total number of historical infrared images in the other area; σ′ is the variance of the radiation intensity in the fire source area of ​​the target infrared image; σ′ z is the variance of the radiation intensity in the fire source area of ​​the z-th historical infrared image; σ z is the covariance of the radiation intensity in the fire source area between the target infrared image and the z-th historical infrared image; Step S204: Obtain each target infrared image of the target area and the fire source approximation of other areas, monitor the environment of the target area in the current cycle, and obtain environmental data, the environmental data includes the average value of each environmental index of the target area, and construct the environmental feature vector F of the target area based on the environmental data. α }, where f1, f2, ..., f α are the average values ​​of the 1st, 2nd, ..., αth environmental indicators of the target area respectively; Step S205: Obtain the characteristic environmental feature vector F′ of the other regions, and calculate the fire approximation score R between the target region and the other regions: Wherein, η1 and η2 are the preset first fire scoring coefficient and the second fire scoring coefficient, respectively, η1+η2=1, η1>0, η2>0; K g is the g-th target infrared image of the target area and the fire source approximation of the other areas; ε is the total number of target infrared images of the target area; Step S206: When the fire similarity score R is greater than a preset fire similarity score threshold, it is determined that the target area and the other areas are fire similar, and the other areas are recorded as target other areas of the target area, and each target other area of ​​the target area is acquired.

4. The intelligent fire-fighting water monitor control method based on visual recognition according to claim 3 is characterized in that: The step S300 includes: Step S301: Acquire equipment data of an intelligent fire-fighting water monitor for fire extinguishing in the target area, wherein the equipment data includes values ​​corresponding to various performance indicators in the intelligent fire-fighting water monitor; Step S302: Acquire historical equipment data of fire water monitors in each other target area of ​​the target area, wherein the historical equipment data includes values ​​corresponding to various performance indicators in the fire water monitors; Step S303: When the fire monitors in one of the other target areas of the target area have the same performance indicators as the intelligent fire monitors, the one of the other target areas is retained; otherwise, the one of the other target areas is removed to obtain the retained other target areas in the target area; Step S304: Obtain historical equipment control records of fire monitors in other target areas, and obtain data corresponding to various control parameters of the fire monitors after fire extinguishing from the historical equipment control records; Step S305: Set the unit time, obtain the historical fire extinguishing records of other target areas, and obtain the unit time of fire extinguishing by the fire monitor from the historical fire extinguishing records. The total number of pixels W in the historical infrared images of other target areas that are determined to have fire sources sum , obtain the total number of pixels W′ that are determined to have fire sources in the historical infrared images of other areas of the target when the fire monitor is not extinguishing the fire sum , calculate the characteristic fire extinguishing value V=W′ of the other target areas sum / W sum ; Step S306: Obtain the maximum value of the characteristic fire extinguishing value of each other target area of ​​the target area, obtain the data corresponding to the various control parameters of the fire water cannon after the fire is extinguished in the other target areas corresponding to the maximum value, and use them as the target control data of the intelligent fire water cannon in the target area, and determine that the intelligent fire water cannon has the best fire extinguishing effect on the target area under the target control data.

5. The intelligent fire-fighting water monitor control method based on visual recognition according to claim 4 is characterized in that: The step S400 includes: Step S401: acquiring target control data of the intelligent fire water monitor, setting various control parameters of the intelligent fire water monitor based on the target control data in the current cycle, and using the intelligent fire water monitor to extinguish fire points in the target area; Step S402: After the smart fire water monitor extinguishes the fire in the target area, the target area is photographed using an infrared thermal imager at unit time intervals to obtain a characteristic infrared image, and based on the location of the fire source in the characteristic infrared image, the spray angle and spray pressure of the smart fire water monitor are adjusted until no pixel points indicating the presence of the fire source appear in the characteristic infrared image, and the smart fire water monitor is intelligently controlled.

6. An intelligent fire-fighting water monitor control system based on visual recognition, used to execute the intelligent fire-fighting water monitor control method based on visual recognition as described in any one of claims 1 to 5, characterized in that: The system includes a target infrared image module, a fire approximate analysis module, a target control data module, and an intelligent control module; The target infrared image module is used to analyze the fire presentation effect of the infrared image on the target area to obtain a target infrared image; The fire proximity analysis module is used to analyze the fire proximity between the target area and the other areas to obtain the target other areas; The target control data module is used to evaluate the fire extinguishing effect of the intelligent fire water monitor on the target area under different control parameters to obtain target control data; The intelligent control module is used to adjust the control parameters of the intelligent fire water cannon according to the target control data, use the intelligent fire water cannon to extinguish the fire in the target area, use an infrared thermal imager to photograph the target area, and intelligently control the intelligent fire water cannon.

7. The intelligent fire-fighting water monitor control system based on visual recognition according to claim 6 is characterized in that: The target infrared image module includes a fire presentation scoring unit and a target infrared image unit; The fire presentation scoring unit is used to calculate the fire presentation score of each infrared image of the target area; The target infrared image unit is used to analyze the fire presentation effects of each infrared image in the target area according to the fire presentation score to obtain a target infrared image.

8. The intelligent fire-fighting water monitor control system based on visual recognition according to claim 6 is characterized in that: The fire approximate analysis module includes a fire approximate scoring unit and a fire approximate analysis unit; The fire approximation scoring unit is used to obtain historical fire data of each other area and calculate the fire approximation score between the target area and each other area; The fire proximity analysis unit is used to analyze the fire proximity degree between the target area and each other area according to the fire proximity score to obtain the target other area.

9. The intelligent fire-fighting water monitor control system based on visual recognition according to claim 6 is characterized in that: The target control data module includes a characteristic fire extinguishing value unit and a target control data unit; The characteristic fire extinguishing value unit is used to calculate the characteristic fire extinguishing value of each target other area of ​​the target area; The target control data unit is used to obtain the maximum value of the characteristic fire extinguishing value of each target other area of ​​the target area, obtain the data corresponding to the various control parameters of the fire water cannon after the fire is extinguished in the target other area corresponding to the maximum value, and use it as the target control data of the intelligent fire water cannon in the target area.

10. The intelligent fire-fighting water monitor control system based on visual recognition according to claim 6 is characterized in that: The intelligent control module includes an intelligent control unit; The intelligent control unit is used to set various control parameters of the intelligent fire-fighting water monitor according to the target control data, and use the intelligent fire-fighting water monitor to extinguish fire points in the target area. After the intelligent fire-fighting water monitor extinguishes fires in the target area, an infrared thermal imager is used to photograph the target area at unit time intervals to obtain a characteristic infrared image, until no pixel points with fire sources appear in the characteristic infrared image, and the intelligent fire-fighting water monitor is intelligently controlled.

Images