Perception method and device of fire-fighting field data, electronic equipment and storage medium

Abstract

The invention relates to a method and device for sensing fire-fighting scene data, electronic equipment and a storage medium, and belongs to the technical field of fire-fighting emergency rescue, and the method comprises the steps: obtaining sensor data, collected by a plurality of different types of sensors, of a fire-fighting scene and image data collected by an image collection device; fusing the sensor data collected by the sensors of the same type to obtain fused data corresponding to the sensors of the same type; and inputting the image data and the fusion data into the trained BP neural network to obtain the danger level of the fire-fighting scene output by the BP neural network. According to the fire-fighting field data sensing method provided by the invention, the image data and the fusion data are input into the trained BP neural network, comprehensive sensing of multiple different types of sensor data and image data is realized, implicit association among multi-modal data is mined by using the neural network, and finally the danger level is output. And a reliable basis is provided for fire rescue.

Description

Technical Field

[0001] This invention relates to the field of fire emergency rescue technology, and in particular to a method, device, electronic equipment and storage medium for sensing fire scene data. Background Technology

[0002] Fire scene environments are complex and ever-changing. High temperatures, dense smoke, toxic gases, and the risk of structural collapse pose serious threats to the safety of rescue personnel, while also placing extremely high demands on on-site data perception. During fire rescue operations, timely and accurate acquisition of various types of data from the fire scene is crucial for making rescue decisions, ensuring the safety of rescue personnel, and improving rescue efficiency. Currently, fire scene data perception mainly relies on traditional sensor equipment or manual observation.

[0003] Traditional sensing methods have many shortcomings. Single-type sensors often only acquire a specific type of data; for example, a temperature sensor can only sense temperature information and cannot comprehensively reflect the complex conditions at the scene, resulting in insufficient data comprehensiveness. Firefighting scenes are harsh environments with high temperatures, dense smoke, and strong electromagnetic interference. Traditional sensors are prone to malfunction or data deviation in such environments, leading to low reliability. Furthermore, data transmission can be affected by interference, causing data delays or loss, impacting the timeliness of rescue decisions. In addition, data obtained through manual observation is highly subjective and limited by the rescuers' field of vision and experience, making it difficult to achieve accurate and comprehensive perception of the scene. Summary of the Invention

[0004] In view of this, it is necessary to provide a method, device, electronic equipment and storage medium for sensing fire scene data, so as to achieve the purpose of comprehensively sensing fire scene data and providing a reliable basis for fire rescue.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for sensing fire scene data, comprising: Acquire sensor data from multiple different types of sensors and image data from image acquisition devices at the fire scene; Sensor data collected by the same type of sensors are fused to obtain fused data corresponding to the same type of sensors; The image data and the fused data are input into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

[0006] In one possible implementation, the fusion of sensor data collected by sensors of the same type to obtain fused data corresponding to the same type of sensors includes: Based on the accuracy of the same type of sensor, determine the corresponding weights for the same type of sensor; Based on the weights, sensor data collected by sensors of the same type are fused to obtain the fused data.

[0007] In one possible implementation, the expression for the fused data is as follows:

[0008] in, Indicates data fusion. This indicates the weight corresponding to the high-precision sensor. This indicates the number of high-precision sensors. Indicates the first Data collected by a high-precision sensor This indicates the weight corresponding to the low-precision sensor. This indicates the number of low-precision sensors. Indicates the first Data collected by a low-precision sensor.

[0009] In one possible implementation, the BP neural network is trained in the following manner: The hazard level of the sample area is marked to obtain the sample label; The initial BP neural network is trained based on the sample fusion data, sample image data, and sample labels of the sample region, with the minimum loss function as the learning objective, to obtain the BP neural network.

[0010] In one possible implementation, the loss function is expressed as follows:

[0011] in, This indicates the number of categories for the sample label. Indicates sample label, This represents the predicted value.

[0012] In one possible implementation, the different types of sensors include: Temperature sensors, smoke sensors, gas sensors, and sound sensors.

[0013] Secondly, the present invention also provides a sensing device for fire scene data, comprising: The acquisition unit is used to acquire sensor data from multiple different types of sensors at the fire scene and image data acquired by image acquisition devices; The fusion unit is used to fuse sensor data collected by the same type of sensors to obtain fused data corresponding to the same type of sensors; The prediction unit is used to input the image data and the fused data into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

[0014] Thirdly, the present invention also provides an electronic device, including a memory and a processor, wherein, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the fire scene data sensing method described in any of the above implementations.

[0015] Fourthly, the present invention also provides a computer-readable storage medium for storing a computer-readable program or instruction, which, when executed by a processor, can implement the steps in the fire scene data sensing method described in any of the above implementations.

[0016] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the fire scene data sensing method described in any of the above implementations.

[0017] The beneficial effects of this invention are as follows: The fire scene data sensing method, device, electronic equipment, and storage medium provided by this invention acquire sensor data from multiple different types of sensors and image data from image acquisition devices at the fire scene. By fusing data from similar types of sensors, the influence of single-point outliers is effectively suppressed. The image data and fused data are then input into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network. This achieves comprehensive perception of multiple different types of sensor data and image data, reduces the risk of single sensor failure through data fusion, and utilizes neural networks to uncover the implicit correlations between multimodal data. Finally, it outputs an objective and quantitative hazard level assessment result, providing a reliable basis for fire rescue. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic flowchart of an embodiment of the fire scene data sensing method provided by the present invention; Figure 2 A schematic diagram of an embodiment of the fire scene data sensing device provided by the present invention; Figure 3 A schematic diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] In the description of the embodiments of the present invention, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0022] The terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a technical feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.

[0023] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0024] This invention provides a method, device, electronic equipment, and storage medium for sensing fire scene data, which will be described below.

[0025] Figure 1 A schematic flowchart of an embodiment of the fire scene data sensing method provided by the present invention is shown below. Figure 1 As shown, the methods for sensing fire scene data include: S101. Acquire sensor data from multiple different types of sensors at the fire scene and image data from image acquisition devices; S102. Fusion of sensor data collected by the same type of sensors to obtain fused data corresponding to the same type of sensors; S103. Input the image data and the fused data into the trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

[0026] In S101, different types of sensors may include temperature sensors, smoke sensors, gas sensors, and sound sensors, etc.

[0027] Multiple different types of sensors can be used to collect similar sensor data at the fire scene, including temperature data, smoke concentration data, gas concentration data, and sound data.

[0028] Image acquisition devices are used to collect image data of the fire scene. These devices can be monitoring equipment with infrared imaging capabilities, such as a combination of thermal imaging cameras and visible light cameras, to capture the spatial distribution characteristics of the fire scene.

[0029] In S102, the fused data is the comprehensive data obtained by weighting the measurement values ​​of the same type of sensors. For example, it can be implemented by using a hierarchical weighting algorithm of high-precision sensors and low-precision sensors to eliminate single-point measurement errors.

[0030] In S103, the BP neural network is a multilayer perceptron model trained by the backpropagation algorithm. It can adopt a network structure consisting of an input layer, a hidden layer, and an output layer to establish a mapping relationship between multimodal data and hazard levels.

[0031] Image data and fused data are input into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

[0032] For example, after the temperature sensor collects temperature data from multiple points, a weighted average is performed based on the accuracy of each sensor to generate fused temperature data. The smoke sensor and sound sensor undergo the same processing to eliminate outliers caused by high-temperature interference from individual sensors. A thermal imaging camera acquires images of the fire's thermal distribution, which are then input into a backpropagation neural network along with the structured fused data.

[0033] The BP neural network extracts spatial correlation features between temperature trends, smoke concentration gradients and high-temperature areas in thermal images through hidden layer nodes, and finally maps them into quantitative hazard levels in the output layer. The hazard level can be divided into five levels: no hazard, low hazard, medium hazard, high hazard and emergency.

[0034] In summary, the fire scene data perception method provided by this invention acquires sensor data from multiple different types of sensors and image data from image acquisition devices at the fire scene. It effectively suppresses the influence of single-point outliers by fusing data from similar types of sensors, and inputs the image data and fused data into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network. This achieves comprehensive perception of multiple different types of sensor data and image data, reduces the risk of single sensor failure through data fusion, and utilizes neural networks to uncover the implicit correlations between multimodal data. Finally, it outputs an objective and quantitative hazard level assessment result, providing a reliable basis for fire rescue.

[0035] In some embodiments of the present invention, the fusion of sensor data collected by sensors of the same type to obtain fused data corresponding to the same type of sensors includes: Based on the accuracy of the same type of sensor, determine the corresponding weights for the same type of sensor; Based on the weights, sensor data collected by sensors of the same type are fused to obtain the fused data.

[0036] In some embodiments of the present invention, the expression for the fused data is as follows:

[0037] in, Indicates data fusion. This indicates the weight corresponding to the high-precision sensor. This indicates the number of high-precision sensors. Indicates the first Data collected by a high-precision sensor This indicates the weight corresponding to the low-precision sensor. This indicates the number of low-precision sensors. Indicates the first Data collected by a low-precision sensor.

[0038] Based on the factory-calibrated accuracy level or real-time calibration results of the sensors, corresponding weights are assigned to sensors of the same type with different accuracies. For example, the weight for a high-precision sensor is set to 0.6, and the weight for a low-precision sensor can be set to 0.4. Then, a weighted average algorithm is used to multiply the high-precision sensor data by its corresponding weight and the low-precision sensor data by their corresponding weights, sum them, and then calculate the mean to obtain the fused data.

[0039] For example, when two high-precision temperature sensors measure 100℃ and 90℃ respectively, and a low-precision sensor measures 85℃, the fused temperature value can be calculated as (100×0.6+90×0.6+85×0.4) / (0.6×2+0.4)=92.5℃.

[0040] Understandably, the methods for fusing smoke concentration data, gas concentration data, and sound data are the same as those for fusing temperature data.

[0041] This invention effectively suppresses the impact of low-precision sensor measurement deviations on the final result by increasing the contribution of high-precision data. It retains the complementary characteristics of multi-sensor data and suppresses the error propagation of low-precision data through weight allocation.

[0042] Traditional methods typically employ equal-weighted averaging or simple majority voting for data fusion, neglecting individual sensor differences. For example, when fusion temperature data, directly taking the arithmetic mean dilutes the value of high-precision sensor data compared to low-precision data. This invention, however, utilizes a precision-oriented weighting mechanism, giving higher weight to measurements from high-precision sensors, significantly improving data reliability. This invention addresses the issue of uneven data reliability caused by differences in device precision in multi-sensor systems. In complex fire-fighting environments such as high temperatures and electromagnetic interference, where low-precision sensors are susceptible to environmental influences and measurement deviations, this invention automatically reduces the weight of abnormal data, preventing overall data failure due to a single sensor malfunction. Simultaneously, by preserving the redundancy of multi-sensor data, it ensures system availability even when some sensors fail, significantly enhancing the robustness and environmental adaptability of the fire monitoring system.

[0043] This invention effectively solves the problem of decreased reliability caused by improper weight allocation when fusing multi-precision sensor data. In complex environments such as high temperature and electromagnetic interference at fire scenes, it can improve the accuracy and stability of fused data, providing reliable input for subsequent hazard level assessment.

[0044] In some embodiments of the present invention, the BP neural network is trained in the following manner: The hazard level of the sample area is marked to obtain the sample label; The initial BP neural network is trained based on the sample fusion data, sample image data, and sample labels of the sample region, with the minimum loss function as the learning objective, to obtain the BP neural network.

[0045] In some embodiments of the present invention, the expression for the loss function is as follows:

[0046] in, This indicates the number of categories for the sample label. Indicates sample label, This represents the predicted value.

[0047] Sample labels refer to the pre-defined hazard level classifications of sample areas, achieved through manual annotation or expert evaluation. For example, a five-level classification method can be used to divide the hazard level into five levels: no danger, low danger, medium danger, high danger, and emergency. Sample labels serve as supervisory signals for neural networks, giving the model learning process a quantifiable optimization objective.

[0048] Sample fusion data refers to standardized data obtained by weighting and fusing raw data collected by the same type of sensors. Specifically, it can be achieved using the average weighting method described in the above embodiments.

[0049] Sample image data refers to visual information of fire scenes acquired through infrared thermal imaging cameras or visible light cameras, and can be stored in the form of a pixel matrix. Sample image data enables neural networks to simultaneously analyze environmental parameters and visual features such as flame morphology and smoke diffusion.

[0050] The initial backpropagation (BP) neural network is a three-layer feedforward network structure consisting of an input layer, hidden layers, and an output layer. The sigmoid function can be used as the activation function for the hidden layers. Minimizing the loss function as the learning objective means iteratively adjusting the network weight parameters through the backpropagation algorithm, using the loss function to measure the difference between the predicted result and the true label.

[0051] During the model training phase, hazard levels were first labeled for different areas in historical fire cases, forming a training dataset containing multiple sets of labels. Multi-source sensor data, including temperature, smoke, gas, and sound, were then fused to generate robust sample fusion data. Simultaneously, flame spread images of the corresponding areas were extracted as sample image data.

[0052] Two types of data are normalized and then input into a neural network. The predicted hazard level is calculated through forward propagation. The difference between the predicted value and the true label is quantified using the cross-entropy function, and the network weights are updated in reverse using the gradient descent algorithm until the loss function converges to a preset threshold. This enables the model to establish a mapping relationship between multimodal data features and hazard levels; for example, it can learn the pattern of high-risk levels corresponding to image features of high temperatures accompanied by dense smoke diffusion.

[0053] This invention constructs sample fusion data by fusing data from multiple sensors and combines it with manually labeled samples, thus providing a clear supervised learning mechanism for the training process. Simultaneously, the joint input of image data and fused data overcomes the limitations of traditional methods that rely on a single data modality, enabling the model to capture multi-dimensional features of the fire scene environment.

[0054] This invention enhances the model's ability to extract features from complex fire scenarios by fusing data from both data sources and image data. The use of a loss function minimization criterion ensures the stability and convergence efficiency of the network parameter optimization process, thereby improving the accuracy and reliability of assessing the hazard level of a fire scene.

[0055] In some embodiments of the present invention, the different types of sensors include: Temperature sensors, smoke sensors, gas sensors, and sound sensors.

[0056] Temperature sensors are used to detect changes in ambient temperature. They can be implemented using thermocouples or infrared sensors. By monitoring changes in temperature gradients, the location of fire sources and their spread trends can be identified.

[0057] Smoke sensors are used to detect the concentration of suspended particulate matter in the air. They can be implemented using photoelectric or ionization sensors. By quantifying the smoke concentration, the type of burning material and the intensity of the fire can be determined.

[0058] Gas sensors are used to identify specific gas components and can be implemented using electrochemical or semiconductor sensors. They assess the level of environmental toxicity risk by detecting the concentrations of carbon monoxide and hydrogen sulfide.

[0059] Sound sensors are used to collect ambient sound wave signals. They can be implemented using piezoelectric or condenser microphones. By analyzing abnormal sound frequency characteristics, the system can determine the precursors to structural collapse or the location of trapped personnel.

[0060] For example, temperature sensors collect real-time temperature data from various areas on-site to create a thermal distribution map to locate the core fire point. Smoke sensors simultaneously monitor the smoke diffusion path and predict the fire's development speed based on the concentration change rate. Gas sensors set detection thresholds for different toxic gases, triggering an early warning when the concentration exceeds the safety limit. Sound sensors continuously receive ambient sound wave signals and identify characteristic frequencies of metal deformation and cracking sounds or cries for help through spectrum analysis. These four types of sensors constitute a multi-dimensional data acquisition system, covering thermodynamics, chemistry, physical structural safety, and vital sign monitoring. Their output data are fused to form a complementary environmental condition description, eliminating assessment blind spots caused by the lack of data dimensions from a single sensor.

[0061] The fire scene data sensing method provided in this invention solves the problem of incomplete sensing of complex disaster factors in existing technologies by adding gas and sound sensing units to construct a complete data chain covering combustion product composition analysis and structural safety assessment.

[0062] To better implement the fire scene data sensing method in the embodiments of the present invention, based on the fire scene data sensing method, correspondingly, as follows: Figure 2As shown, this embodiment of the invention also provides a fire scene data sensing device, the fire scene data sensing device 200 including: The acquisition unit 201 is used to acquire sensor data from multiple different types of sensors at the fire scene and image data acquired by the image acquisition device; The fusion unit 202 is used to fuse sensor data collected by sensors of the same type to obtain fused data corresponding to the same type of sensors; The prediction unit 203 is used to input the image data and the fused data into the trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

[0063] The fire scene data sensing device 200 provided in the above embodiments can realize the technical solutions described in the above fire scene data sensing method embodiments. The specific implementation principles of each module or unit can be found in the corresponding content in the above fire scene data sensing method embodiments, which will not be repeated here.

[0064] like Figure 3 As shown, the present invention also provides an electronic device 300. The electronic device 300 includes a processor 301, a memory 302, and a display 303. Figure 3 Only some components of the electronic device 300 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0065] In some embodiments, processor 301 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 302 or process data, such as the fire scene data sensing method of the present invention.

[0066] In some embodiments, processor 301 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, processor 301 may be local or remote. In some embodiments, processor 301 may be implemented on a cloud platform. In some embodiments, the cloud platform may include a private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, internal cloud, multi-cloud, or any combination thereof.

[0067] In some embodiments, memory 302 may be an internal storage unit of electronic device 300, such as a hard disk or memory of electronic device 300. In other embodiments, memory 302 may also be an external storage device of electronic device 300, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on electronic device 300.

[0068] Furthermore, the memory 302 may include both internal storage units of the electronic device 300 and external storage devices. The memory 302 is used to store application software and various types of data installed on the electronic device 300.

[0069] In some embodiments, display 303 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an organic light-emitting diode (OLED) touchscreen. Display 303 is used to display information from electronic device 300 and to display a visual user interface. Components 301-303 of electronic device 300 communicate with each other via a system bus.

[0070] In one embodiment, when the processor 301 executes the sensing program for fire scene data in the memory 302, the following steps can be implemented: Acquire sensor data from multiple different types of sensors and image data from image acquisition devices at the fire scene; Sensor data collected by the same type of sensors are fused to obtain fused data corresponding to the same type of sensors; The image data and the fused data are input into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

[0071] It should be understood that when the processor 301 executes the sensing program for fire scene data in the memory 302, in addition to the functions mentioned above, it can also perform other functions, as detailed in the description of the corresponding method embodiments above.

[0072] Furthermore, the embodiments of the present invention do not specifically limit the type of electronic device 300 mentioned. Electronic device 300 can be a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, laptop computer, or other portable electronic device. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, electronic device 300 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0073] Accordingly, embodiments of the present invention also provide a computer-readable storage medium for storing computer-readable programs or instructions. When the programs or instructions are executed by a processor, they can implement the steps or functions of the fire scene data sensing methods provided in the above-described method embodiments.

[0074] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to perform the steps or functions of the fire scene data sensing method provided in the above-described method embodiments.

[0075] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0076] The above provides a detailed description of the fire scene data sensing method, device, electronic equipment, and storage medium provided by the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for sensing fire scene data, characterized in that, include: Acquire sensor data from multiple different types of sensors and image data from image acquisition devices at the fire scene; Sensor data collected by the same type of sensors are fused to obtain fused data corresponding to the same type of sensors; The image data and the fused data are input into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

2. The method for sensing fire scene data according to claim 1, characterized in that, The process of fusing sensor data collected by the same type of sensors to obtain fused data corresponding to the same type of sensors includes: Based on the accuracy of the same type of sensor, determine the corresponding weights for the same type of sensor; Based on the weights, sensor data collected by sensors of the same type are fused to obtain the fused data.

3. The method for sensing fire scene data according to claim 2, characterized in that, The expression for the fused data is as follows: in, Indicates data fusion. This indicates the weight corresponding to the high-precision sensor. This indicates the number of high-precision sensors. Indicates the first Data collected by a high-precision sensor This indicates the weight corresponding to the low-precision sensor. This indicates the number of low-precision sensors. Indicates the first Data collected by a low-precision sensor.

4. The method for sensing fire scene data according to claim 1, characterized in that, The BP neural network is trained in the following way: The hazard level of the sample area is marked to obtain the sample label; The initial BP neural network is trained based on the sample fusion data, sample image data, and sample labels of the sample region, with the minimum loss function as the learning objective, to obtain the BP neural network.

5. The method for sensing fire scene data according to claim 4, characterized in that, The expression for the loss function is as follows: in, This indicates the number of categories for the sample label. Indicates sample label, This represents the predicted value.

6. The method for sensing fire scene data according to claim 1, characterized in that, The different types of sensors include: Temperature sensors, smoke sensors, gas sensors, and sound sensors.

7. A sensing device for fire scene data, characterized in that, include: The acquisition unit is used to acquire sensor data from multiple different types of sensors at the fire scene and image data acquired by image acquisition devices; The fusion unit is used to fuse sensor data collected by the same type of sensors to obtain fused data corresponding to the same type of sensors; The prediction unit is used to input the image data and the fused data into a trained BP neural network to obtain the hazard level of the fire scene output by the BP neural network.

8. An electronic device, characterized in that, Including memory and processor, among which, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the fire scene data sensing method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps in the fire scene data sensing method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps in the fire scene data sensing method as described in any one of claims 1 to 6.

Images