Flame composite detection method, device, computer equipment and storage medium

Abstract

The present application relates to a flame composite detection method, device, computer equipment and storage medium, the method comprising: identifying flames and / or smoke in the image information of the target area by training a complete image recognition model to obtain a recognition result; updating a first flame mark variable according to the recognition result; collecting detection results of a sensor group arranged in the target area; the sensor group comprises a main sensor and an auxiliary sensor respectively responding to different radiation bands; comparing and analyzing the detection results of the main sensor and the auxiliary sensor to update a second flame mark variable according to the analysis result; determining the flame detection result according to the first flame mark variable and the second flame mark variable. Through the present application, it is possible to perform composite detection of the flame in the target area according to the recognition result of the image recognition model and the detection result of the sensor group, reduce false detection or missed detection, and solve the problem of inaccurate flame detection.

Description

Technical Field

[0001] The present application relates to the technical field of fire safety monitoring, and in particular to a flame composite detection method, device, computer equipment and storage medium. Background Art

[0002] In fire safety monitoring, flame detection technology has been widely used in multiple scenarios due to its high efficiency and accuracy. Whether it is a home environment, public places, industrial production, smart fire protection systems, or specific workplaces, this technology provides strong support for early warning and alarm of fires, effectively reducing the risk and loss of fires.

[0003] Most existing flame detection methods rely on single image recognition, sound recognition or infrared pyroelectric sensors, which are prone to false detection or missed detection. For example, pictures containing flames, or heat sources such as sunlight and car lights are mistakenly detected as flames, resulting in inaccurate flame detection.

[0004] With respect to the problem of inaccurate flame detection in related technologies, no effective solution has been proposed so far. Summary of the invention

[0005] Based on this, it is necessary to provide a flame composite detection method, device, computer equipment and storage medium that can improve the accuracy of flame detection in response to the above technical problems.

[0006] In a first aspect, a flame composite detection method is provided in this embodiment, comprising:

[0007] By training a complete image recognition model, flames and / or smoke in the image information of the target area are recognized to obtain a recognition result;

[0008] Update the first flame mark variable according to the recognition result;

[0009] Collecting detection results of a sensor group arranged in the target area; the sensor group includes a main sensor and an auxiliary sensor that respond to different radiation bands respectively;

[0010] Comparing and analyzing the detection results of the main sensor and the auxiliary sensor to update the second flame mark variable according to the analysis result;

[0011] A flame detection result is determined according to the first flame flag variable and the second flame flag variable.

[0012] In some of the embodiments, the flame and / or smoke in the image information of the target area is identified by training a complete image recognition model to obtain the recognition result, including:

[0013] Acquiring image information of the target area through a camera;

[0014] The image information is input into the image recognition model, and the recognition result is output; the recognition result indicates whether the image information contains flames or smoke.

[0015] In some embodiments, the comparing and analyzing the detection results of the primary sensor and the secondary sensor includes:

[0016] Performing Fourier processing on the detection results of the main sensor and the auxiliary sensor to obtain the amplitude of the detection result at each frequency point;

[0017] Frequency analysis is performed according to the amplitude at each frequency point, a target frequency point that meets a preset amplitude condition is determined, and the analysis result is obtained.

[0018] In some embodiments, updating the second flame flag variable according to the analysis result includes:

[0019] Determine whether the number of target frequency points in the analysis result is greater than or equal to a preset threshold;

[0020] If the judgment is yes, the analysis result indicates that the sensor group detects flame, and the second flame flag variable is updated accordingly.

[0021] In some of the embodiments, after collecting the detection results of the sensor group disposed in the target area, the method further includes:

[0022] Determining whether the detection result of the main sensor meets a preset voltage condition;

[0023] If the judgment is yes, flame is detected by the sensor group, and the second flame mark variable is updated accordingly.

[0024] In some embodiments, the first flame mark variable is set corresponding to the recognition result of the image recognition model; the second flame mark variable is set corresponding to the detection result of the sensor group.

[0025] In some of the embodiments, it also includes:

[0026] When the flame detection result is that a flame is detected, an alarm is triggered accordingly; and the alarm information and the image information of the target area are reported to the platform end.

[0027] In a second aspect, a flame composite detection device is provided in this embodiment, including:

[0028] An image recognition module is used to recognize flames and / or smoke in the image information of the target area by training a complete image recognition model to obtain a recognition result; and update the first flame mark variable according to the recognition result;

[0029] A sensor detection module is used to collect detection results of a sensor group arranged in the target area; the sensor group includes a main sensor and an auxiliary sensor that respond to different radiation bands respectively; the detection results of the main sensor and the auxiliary sensor are compared and analyzed to update the second flame mark variable according to the analysis results;

[0030] The composite detection module is used to determine a flame detection result according to the first flame mark variable and the second flame mark variable.

[0031] In a third aspect, a computer device is provided in this embodiment, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the flame composite detection method described in the first aspect when executing the computer program.

[0032] In a fourth aspect, in this embodiment, a storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the flame composite detection method described in the first aspect is implemented.

[0033] Compared with the related art, the flame composite detection method, device, computer equipment and storage medium provided in this embodiment, by training a complete image recognition model, recognizes the flame and / or smoke in the image information of the target area to obtain a recognition result; updates the first flame mark variable according to the recognition result; collects the detection result of the sensor group set in the target area; the sensor group includes a main sensor and an auxiliary sensor that respond to different radiation bands respectively; compares and analyzes the detection results of the main sensor and the auxiliary sensor to update the second flame mark variable according to the analysis result; determines the flame detection result according to the first flame mark variable and the second flame mark variable. Through this embodiment, it is possible to perform composite detection of the flame in the target area according to the recognition result of the image recognition model and the detection result of the sensor group, reduce false detection or missed detection, and solve the problem of inaccurate flame detection.

[0034] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0036] Figure 1 is a hardware structure block diagram of a terminal of a flame composite detection method in an embodiment;

[0037] Figure 2 is a flow chart of a flame composite detection method in one embodiment;

[0038] Figure 3 is a schematic diagram of a flame composite detection system in one embodiment;

[0039] Figure 4 is a flow chart of a flame composite detection method in another embodiment;

[0040] Figure 5 It is a structural block diagram of a flame composite detection device in an embodiment.

[0041] In the figure: 102, processor; 104, memory; 106, transmission device; 108, input and output device; 10, image recognition module; 20, sensor detection module; 30, composite detection module. DETAILED DESCRIPTION

[0042] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0043] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meaning understood by people with general skills in the technical field to which this application belongs. The words "one", "a", "the", "these" and the like in this application do not indicate a quantitative limitation, and they may be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone. Generally, the character " / " indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", "third", etc. in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0044] The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. For example, running on a terminal, Figure 1 FIG. 1 is a hardware structure diagram of a terminal of the flame composite detection method of this embodiment. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 and a memory 104 for storing data, wherein the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0045] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the flame composite detection method in the present embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0046] The transmission device 106 is used to receive or send data via a network. The above network includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0047] In fire safety monitoring, flame detection technology has been widely used in multiple scenarios due to its high efficiency and accuracy. Whether it is a home environment, public places, industrial production, smart fire protection systems, or specific workplaces, this technology provides strong support for early warning and alarm of fires, effectively reducing the risk and loss of fires.

[0048] Most existing flame detection methods rely on single image recognition, sound recognition or infrared pyroelectric sensors, which are prone to false detection or missed detection. For example, pictures containing flames, or heat sources such as sunlight and car lights are mistakenly detected as flames, resulting in inaccurate flame detection.

[0049] In this embodiment, a flame composite detection method is provided. Figure 2 is a flow chart of the flame composite detection method in this embodiment, such as Figure 2 As shown, the method comprises the following steps:

[0050] Step S210, identifying flames and / or smoke in the image information of the target area by training a complete image recognition model to obtain a recognition result.

[0051] Specifically, monitoring and acquisition equipment is set up in the target area to collect image information of the target area, and the flame and / or smoke in the image information are identified by the image recognition model to obtain the recognition result. The target area can be a production site such as a machine room, warehouse, factory, or a crowded place such as a hospital or school.

[0052] The image recognition model is obtained by training a neural network. Using images and videos containing flames and / or smoke as a data set, an image recognition model for identifying flames, an image recognition model for identifying smoke, or an image recognition model for identifying flames and smoke can be trained separately. Since flames and smoke are both significant features related to fire conditions, in order to reduce missed detection when flame features are not obvious but smoke features are obvious, in some embodiments, the Yolo V8 model is used to annotate the flames and smoke in the data set, respectively, to obtain an image recognition model that can identify and classify flames and smoke.

[0053] Step S220: updating the first flame mark variable according to the recognition result.

[0054] Specifically, a first flame flag variable corresponding to the recognition result is pre-set, and the first flame flag variable is initialized to 0. Optionally, when the first flame flag variable is 0, it indicates that the recognition result is that flame or smoke is not recognized by the image recognition model, and when the first flame flag variable is 1, it indicates that the recognition result is that flame and / or smoke is recognized by the image recognition model.

[0055] Step S230, collecting detection results of a sensor group arranged in the target area; the sensor group includes a main sensor and auxiliary sensors that respond to different radiation bands respectively.

[0056] Specifically, a sensor group is set in the target area, and the sensor group includes a main sensor and auxiliary sensors that respond to different radiation bands respectively. Each sensor can be an infrared pyroelectric sensor, which is used to convert infrared radiation within the detection range of the sensor group into an electrical signal to obtain a detection result.

[0057] Among them, the main sensor is a sensor sensitive to the radiation band of the flame, and its response wavelength is usually 4.3um (micrometer), and there can be multiple auxiliary sensors, all of which are sensors insensitive to the radiation band of the flame, and their response wavelengths can be 3.8um, 5.0um, etc. Exemplarily, one auxiliary sensor with a response wavelength of 3.8um or 5.0um can be set, and two auxiliary sensors with response wavelengths of 3.8um and 5.0um can be set respectively, or even more auxiliary sensors. In this embodiment, there is no specific limitation on the number of auxiliary sensors.

[0058] Step S240: Compare and analyze the detection results of the main sensor and the auxiliary sensor to update the second flame mark variable according to the analysis result.

[0059] Specifically, since the main sensor and the auxiliary sensor are sensors that are sensitive and insensitive to the radiation band of the flame, respectively, the auxiliary sensor is used as an auxiliary comparison sensor to compare and analyze the detection results of the main sensor and the auxiliary sensor to obtain the analysis result. The detection results can be compared and analyzed in the time domain and the frequency domain, for example, the voltage value of the detection result is first analyzed in the time domain, and then the detection result is further analyzed in the frequency domain through Fourier operation.

[0060] The second flame flag variable is updated correspondingly according to the analysis result of the detection result of the sensor group, the second flame flag variable corresponding to the detection result of the sensor group is pre-set, and the second flame flag variable is initialized to 0. Optionally, when the second flame flag variable is 0, it indicates that the recognition result is that the flame is not detected by the sensor group, and when the second flame flag variable is 1, it indicates that the recognition result is that the flame is detected by the sensor group.

[0061] Step S250, determining a flame detection result according to the first flame mark variable and the second flame mark variable.

[0062] Specifically, the flame detection result is determined according to the first flame flag variable and the second flame flag variable, the recognition result of the image recognition model and the detection result of the sensor group are integrated. When the first flame flag variable and the second flame flag variable both indicate that the flame is recognized and detected, the flame detection result is that the flame is detected in the target area, otherwise, the flame detection result is that the flame is not detected in the target area.

[0063] Through the above steps, the recognition result of flame and / or smoke is obtained by the image recognition model, and the detection result of flame is obtained by the sensor group. The recognition result of the image recognition model and the detection result of the sensor group are combined to perform composite detection on the flame in the target area, thereby reducing false detection or missed detection and solving the problem of inaccurate flame detection.

[0064] In some embodiments, the step S210 above identifies flames and / or smoke in the image information of the target area by training a complete image recognition model to obtain a recognition result, including the following steps:

[0065] Step S211, acquiring image information of the target area through a camera.

[0066] Step S212, input the image information into the image recognition model and output the recognition result; the recognition result indicates whether the image information contains flames or smoke.

[0067] Specifically, image information is acquired by a camera disposed in the target area, and the image information includes the target area. The image recognition model outputs a recognition result of flame and / or smoke. When the recognition result indicates that the image information includes flame or smoke, the first flame flag variable is correspondingly updated to 1, otherwise the first flame flag variable is updated to 0.

[0068] Among them, the image recognition model is trained using the Yolo V8 model. The training process of the image recognition model that can recognize and classify flames and smoke is as follows:

[0069] 1. Data collection: Collect various pictures containing flames and smoke as data sets;

[0070] 2. Label the flames and smoke in the picture separately;

[0071] 3. Download the yolov8 source code; create a new classification.yaml file in the source code root directory, and specify the training image set path, verification image path, and the total number of classifications as 2, namely flame and smoke;

[0072] 4. Set the training weights for training and output the training result model;

[0073] 5. Test and verify the model best.pt trained in step 4 to see if the model can accurately identify the flames and smoke in the picture;

[0074] 6. Adjust the training parameters and repeat steps 4 and 5 to get the best training model;

[0075] 7. Convert the trained model file best.pt to onnx (Open Neural Network Exchange) format to facilitate subsequent conversion into a format that can be used by the device.

[0076] In this embodiment, a visual solution is adopted, and flame features and smoke features can be used for training. The trained image recognition model can be divided into flame type and smoke type. Flame and smoke are identified by the image recognition model to avoid missed identification in cases where the flame features are not obvious but the smoke features are obvious.

[0077] In some embodiments, the comparison and analysis of the detection results of the primary sensor and the secondary sensor in step S240 includes the following steps:

[0078] Step S241, performing Fourier processing on the detection results of the main sensor and the auxiliary sensor to obtain the amplitude of the detection results at each frequency point.

[0079] Specifically, the voltage values ​​of the main sensor and the auxiliary sensor are collected as detection results, and the detection results are analyzed in the frequency domain by Fourier processing. After Fourier processing, the voltage value of each sensor after Fourier operation is obtained, and the detection result of each sensor contains the voltage value corresponding to each frequency point. Exemplarily, the resolution of Fourier processing is set to 0.5. Since the frequency range of flame fluctuation is 3Hz-7Hz (Hertz), the amplitude of the first 10Hz is retained after processing. The detection result of each sensor contains the voltage values ​​corresponding to 20 frequency points, and the frequency points start from 0 and increase in steps of 0.5. Among them, the smaller the resolution, the higher the accuracy of frequency analysis. The resolution can also be set to 0.25, 0.1, etc., without specific limitation.

[0080] Step S242, performing frequency analysis according to the amplitude at each frequency point, determining the target frequency point that meets the preset amplitude condition, and obtaining the analysis result.

[0081] Specifically, the preset amplitude condition is: for each frequency point, the amplitude of the main sensor at the frequency point is greater than the preset multiple of the amplitude of the auxiliary sensor at the frequency point. Among them, since the auxiliary sensor is not as sensitive to the radiation band of the flame as the main sensor, the auxiliary sensor is used for auxiliary comparison, and the preset multiple can be set to 3 times.

[0082] The amplitude of the main sensor and the auxiliary sensor at each frequency point is analyzed, and the frequency point that meets the preset amplitude condition is used as the target frequency point to obtain the analysis result.

[0083] In this embodiment, frequency analysis is performed on the detection results of the main sensor and the auxiliary sensor in the frequency domain through Fourier processing to eliminate interference from other heat sources, obtain analysis results, and use the analysis results to update the second flame mark variable in subsequent steps.

[0084] In some embodiments, updating the second flame mark variable according to the analysis result in step S240 includes the following steps:

[0085] Determine whether the number of target frequency points in the analysis result is greater than or equal to a preset threshold;

[0086] If the judgment is yes, the analysis result indicates that the sensor group detects a flame, and the second flame flag variable is updated accordingly.

[0087] Specifically, the preset threshold can be set in a certain proportion according to the number of frequency points in the detection results of each sensor. For example, if each detection result in the above embodiment contains 20 frequency points, the preset threshold can be set to 5 accordingly, that is, the target number of frequency points is greater than or equal to 5.

[0088] If the judgment is yes, the sensor group detects the flame, and the second flame flag variable is updated to 1 accordingly.

[0089] If the judgment is no, the analysis result indicates that the sensor group has not detected a flame, and the second flame flag variable remains at 0.

[0090] In this embodiment, it is possible to determine whether flame is detected by the sensor group based on the analysis result, and update the second flame mark variable accordingly, so that the second flame mark variable is combined with the first flame mark variable for composite detection in subsequent steps.

[0091] In some of the embodiments, after collecting the detection results of the sensor group disposed in the target area, the following steps are also included:

[0092] It is determined whether the detection result of the main sensor meets the preset voltage condition; if it is determined to be yes, the flame is detected by the sensor group, and the second flame mark variable is updated accordingly.

[0093] Specifically, the voltage value of the detection result is analyzed in the time domain to determine whether the voltage value of the main sensor meets the preset voltage condition. The preset voltage condition is: the voltage value exceeding a certain proportion is the voltage limit full value, wherein the proportion can be set to 50%-80%.

[0094] If the judgment is yes, the flame is detected by the sensor group, and the second flame flag variable is updated accordingly. This judgment is for the case where the external flame or non-flame heat source signal changes strongly, and the voltage value of the sensor is a square wave. In this case, Fourier processing is invalid. Therefore, it is judged that the flame is detected by the sensor group, and the second flame flag variable is updated to 1 accordingly.

[0095] If the judgment is negative, the second flame mark variable is kept unchanged at 0, and the detection result is further analyzed in the frequency domain through Fourier operation.

[0096] For example, each sensor collects 2048 voltage values ​​in the same second, and the voltage range varies within 0-3.3V, that is, each sensor collects 2048 voltage values. When more than 80% of all the voltage values ​​of each sensor are the voltage limit full value 3.3V, it is determined that the preset voltage condition is met.

[0097] In this embodiment, by analyzing the voltage value of the detection result in the time domain, it is possible to make a judgment on the situation where the external flame or non-flame heat source signal changes strongly, thereby avoiding missed detection in the frequency domain analysis.

[0098] In some of the embodiments, the first flame sign variable is set corresponding to the recognition result of the image recognition model; the second flame sign variable is set corresponding to the detection result of the sensor group.

[0099] Specifically, the first flame flag variable and the second flame flag variable are set and initialized to 0. The first flame flag variable is set corresponding to the recognition result of the image recognition model. When the first flame flag variable is 0, it indicates that the recognition result is that the flame or smoke is not recognized by the image recognition model. When the first flame flag variable is 1, it indicates that the recognition result is that the flame and / or smoke is recognized by the image recognition model. The second flame flag variable is set corresponding to the detection result of the sensor group. When the second flame flag variable is 0, it indicates that the recognition result is that the flame is not detected by the sensor group. When the second flame flag variable is 1, it indicates that the recognition result is that the flame is detected by the sensor group.

[0100] In this embodiment, the first flame mark variable and the second flame mark variable are set to indicate the situations in which the flame is detected by the image recognition model and the sensor group respectively, so as to comprehensively perform flame composite detection.

[0101] In some embodiments, the method further comprises:

[0102] When the flame detection result is that a flame is detected, an alarm is triggered accordingly; and the alarm information and image information of the target area are reported to the platform.

[0103] Specifically, when a flame is detected by both the camera and the sensor group, the flame detection result is that a flame is detected, which triggers a fire alarm and reports the fire alarm information to the platform through MQTT (Message Queuing Telemetry Transport). At the same time, the platform also receives and stores image information of the target area. The platform includes client terminals such as web clients and mobile clients for viewing fire alarm information and videos. In addition, the client can also send fire alarm reset information to the front-end sensor group and camera.

[0104] In this embodiment, by triggering a fire alarm and reporting it to the platform, the on-site video and on-site fire information can be viewed in real time, and the fire video can be automatically recorded to facilitate subsequent on-site review.

[0105] The present embodiment is described and illustrated below through preferred embodiments.

[0106] In this embodiment, a flame composite detection system is provided. Figure 3 Schematic diagram of the flame composite detection system in this embodiment. Figure 3 As shown, the system includes: a flame identification detector and a platform end.

[0107] Among them, the front-end flame recognition detector is set in the required application scenario, such as computer room, warehouse, etc., and is connected to the platform through wired or wireless (4G / 5G), and transmits fire alarm information and video stream information through MQTT protocol and RTSP (Real Time Streaming Protocol).

[0108] The flame recognition detector includes a processor, a camera, three sensors that respond to different radiation bands (including a main sensor with a response band of 4.3um and two auxiliary sensors with response bands of 5.0um and 3.8um), a buzzer, an alarm light, a 4G / 5G module and a wired network module (RJ45 interface).

[0109] The platform includes a server, a web client, and a mobile client. The server is responsible for receiving the fire alarm information sent by the flame recognition detector and pushing it to the web client and mobile client. The server also receives and stores the video stream pushed by the flame recognition detector device. The web client and mobile client are used to view the fire alarm information and video. In addition, the client can also send the fire alarm reset information to the flame recognition detector device.

[0110] By combining the video image information obtained by the camera with the detection results of the three sensors, it is comprehensively judged whether a flame is detected in the target area. If the first flame flag variable fireVideo is 1, and the second flame flag variable fireIR is 1, it is determined that a flame is detected, and the buzzer and alarm light are triggered by the PWM (Pulse Width Modulation) signal, and the fire alarm information is reported to the platform end. The server pushes the fire alarm information to the client through HTTPS (Hypertext Transfer Protocol Secure). If the flame recognition detector receives the fire reset information, the buzzer and alarm light are stopped, and the real-time video stream is pushed to the server through the RTSP protocol.

[0111] Figure 4 is a flow chart of the flame composite detection method of this embodiment, which is applied to the above flame composite detection system, such as Figure 4 As shown, the flame composite detection method includes the following steps:

[0112] Step S410, setting and initializing a first flame flag variable and a second flame flag variable.

[0113] Among them, the first flame flag variable fireVideo for recognizing flame or smoke through the camera is 0. When fireVideo is 0, it means that the flame or smoke is not recognized by the camera. When fireVideo is 1, it means that the flame or smoke is recognized by the camera; the second flame flag variable fireIR for detecting flame through three sensors is 0. When the second flame flag variable fireIR is 0, it means that the flame is not detected by the three sensors. When the second flame flag variable fireIR is 1, it means that the flame is detected by three infrared pyroelectric sensors.

[0114] Step S420, acquiring image information of the target area through a camera; inputting the image information into an image recognition model, and outputting a recognition result; and updating the first flame mark variable according to the recognition result.

[0115] Among them, when the first flame mark variable is 1, the indicator diagram recognition model recognizes flame or smoke, and at the same time checks the second flame mark variable. When both flame mark variables are 1, it is confirmed that flame is detected.

[0116] Step S430, collecting the voltage values ​​of the three sensors, and determining whether more than 80% of the voltage values ​​of the main sensor are the full voltage limit value.

[0117] The sensor includes a main sensor with a response band of 4.3um and two auxiliary sensors with response bands of 5.0um and 3.8um. ​​Each sensor collects 2048 voltage values ​​in the same second, and the voltage range varies from 0 to 3.3V, that is, each sensor collects 2048 voltage values.

[0118] If the judgment is yes, step S431 is executed, the second flame mark variable is updated to 1, and the first flame mark variable is checked at the same time. When both flame mark variables are 1, it is confirmed that the flame is detected.

[0119] If the judgment is negative, the second flame mark variable is kept unchanged and step S440 is executed.

[0120] Step S440: Perform Fourier processing on the voltage values ​​of the three sensors to obtain the voltage value of each sensor at each frequency point.

[0121] Step S450, determining whether the amplitude of the main sensor at the same frequency point is more than three times the amplitude of the auxiliary sensor, and whether the frequency point number is greater than or equal to a preset threshold.

[0122] If the judgment is yes, step S431 is executed to update the second flame mark variable to 1, and the second flame mark variable is checked at the same time. When both flame mark variables are 1, it is confirmed that the flame is detected.

[0123] If the judgment is no, step S451 is executed to keep the second flame flag variable at 0.

[0124] It should be noted that the steps shown in the above process or the flowchart of the attached figure can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here. For example, the update process of the first flame mark variable and the second flame mark variable is performed simultaneously, regardless of the order, and once the flame is recognized in the image recognition model, the second flame mark variable of the sensor group is synchronously obtained to comprehensively confirm whether the flame is detected.

[0125] In this embodiment, the recognition results of flame and smoke are obtained through the image recognition model, the detection results of the flame are obtained through the sensor group, and the detection results are analyzed in the time domain and frequency domain. The recognition results of the image recognition model and the detection results of the sensor group are combined to perform composite detection of the flame in the target area, thereby reducing false detection or missed detection and solving the problem of inaccurate flame detection.

[0126] In this embodiment, a flame recombination device is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. The terms "module", "unit", "subunit" and the like used below can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0127] Figure 5 is a structural block diagram of the flame composite detection device of this embodiment, such as Figure 5 As shown, the device comprises:

[0128] The image recognition module 10 is used to recognize flames and / or smoke in the image information of the target area by training a complete image recognition model to obtain a recognition result; and update the first flame mark variable according to the recognition result.

[0129] The sensor detection module 20 is used to collect the detection results of the sensor group set in the target area; the sensor group includes a main sensor and an auxiliary sensor that respond to different radiation bands respectively; the detection results of the main sensor and the auxiliary sensor are compared and analyzed to update the second flame mark variable according to the analysis results.

[0130] The composite detection module 30 is used to determine the flame detection result according to the first flame mark variable and the second flame mark variable.

[0131] Through the device provided in this embodiment, the image recognition model obtains the recognition result of the flame and / or smoke, and the detection result of the flame is obtained through the sensor group. By combining the recognition result of the image recognition model and the detection result of the sensor group, the flame in the target area is compositely detected, thereby reducing false detection or missed detection and solving the problem of inaccurate flame detection.

[0132] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0133] In this embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0134] Optionally, the computer device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0135] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and will not be repeated in this embodiment.

[0136] In addition, in combination with the flame composite detection method provided in the above embodiments, a storage medium may be provided in this embodiment to implement the flame composite detection method. The storage medium stores a computer program; when the computer program is executed by a processor, any flame composite detection method in the above embodiments is implemented.

[0137] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0138] It should be understood that the specific embodiments described herein are only used to explain the application, rather than to limit it. Based on the embodiments provided in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the protection scope of this application.

[0139] Obviously, the drawings are only some examples or embodiments of the present application. For ordinary technicians in the field, the present application can also be applied to other similar situations based on these drawings without creative work. In addition, it is understandable that although the work done in this development process may be complicated and lengthy, for ordinary technicians in the field, certain changes in design, manufacturing or production based on the technical content disclosed in this application are only conventional technical means and should not be regarded as insufficient content disclosed in this application.

[0140] The term "embodiment" in this application refers to a specific feature, structure or characteristic described in conjunction with the embodiment that can be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily mean the same embodiment, nor does it mean that it is mutually exclusive with other embodiments and is independent or optional. It is clearly or implicitly understood by those of ordinary skill in the art that the embodiments described in this application can be combined with other embodiments without conflict.

[0141] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of patent protection. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the attached claims.

Claims

1. A flame composite detection method, characterized in that: include: By training a complete image recognition model, the flame and smoke in the image information of the target area are identified to obtain the recognition result; Update the first flame mark variable according to the recognition result; Collecting detection results of a sensor group arranged in the target area; The sensor group includes a main sensor and an auxiliary sensor that respond to different radiation bands respectively; Determining whether the detection result of the main sensor meets a preset voltage condition; If the judgment is yes, the flame is detected by the sensor group, and the second flame mark variable is updated accordingly; the preset voltage condition is: the voltage value exceeding a certain proportion is the voltage limit full value; If the judgment is no, the second flame mark variable is kept unchanged, and the detection results of the main sensor and the auxiliary sensor are further compared and analyzed to update the second flame mark variable according to the analysis result; The steps include: Performing Fourier processing on the detection results of the main sensor and the auxiliary sensor to obtain the amplitude of the detection result at each frequency point; Perform frequency analysis according to the amplitude at each frequency point, determine the target frequency point that meets the preset amplitude condition, and obtain the analysis result; the preset amplitude condition is: for each frequency point, the amplitude of the main sensor at the frequency point is greater than the preset multiple of the amplitude of the auxiliary sensor at the frequency point; A flame detection result is determined according to the first flame flag variable and the second flame flag variable.

2. The flame composite detection method according to claim 1, characterized in that: The flame and smoke in the image information of the target area are identified by training a complete image recognition model to obtain the recognition results, including: Acquiring image information of the target area through a camera; The image information is input into the image recognition model, and the recognition result is output; the recognition result indicates whether the image information contains flames or smoke.

3. The flame composite detection method according to claim 1, characterized in that: The updating of the second flame mark variable according to the analysis result comprises: Determine whether the number of target frequency points in the analysis result is greater than or equal to a preset threshold; If the judgment is yes, the analysis result indicates that the sensor group detects flame, and the second flame flag variable is updated accordingly.

4. The flame composite detection method according to claim 1, characterized in that: The first flame mark variable is set corresponding to the recognition result of the image recognition model; the second flame mark variable is set corresponding to the detection result of the sensor group.

5. The flame composite detection method according to claim 1, characterized in that: Also includes: When the flame detection result is that flame is detected, an alarm is triggered accordingly; And report the warning information and the image information of the target area to the platform.

6. A flame composite detection device, characterized in that: include: An image recognition module is used to identify flames and smoke in the image information of the target area by training a complete image recognition model to obtain a recognition result; Update the first flame mark variable according to the recognition result; A sensor detection module, used to collect detection results of a sensor group arranged in the target area; the sensor group includes a main sensor and an auxiliary sensor respectively responding to different radiation bands; and determine whether the detection result of the main sensor meets a preset voltage condition; If the judgment is yes, the flame is detected by the sensor group, and the second flame mark variable is updated accordingly; the preset voltage condition is: the voltage value exceeding a certain proportion is the voltage limit full value; If the judgment is no, the second flame mark variable is kept unchanged, and the detection results of the main sensor and the auxiliary sensor are further compared and analyzed to update the second flame mark variable according to the analysis result; The steps include: Performing Fourier processing on the detection results of the main sensor and the auxiliary sensor to obtain the amplitude of the detection result at each frequency point; Perform frequency analysis according to the amplitude at each frequency point, determine the target frequency point that meets the preset amplitude condition, and obtain the analysis result; the preset amplitude condition is: for each frequency point, the amplitude of the main sensor at the frequency point is greater than the preset multiple of the amplitude of the auxiliary sensor at the frequency point; The composite detection module is used to determine a flame detection result according to the first flame mark variable and the second flame mark variable.

7. A computer device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to execute the flame composite detection method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the flame composite detection method according to any one of claims 1 to 5 are implemented.

Images