Multi-data fusion transformer substation fire alarm method and fire monitoring system

Abstract

The invention provides a multi-data fusion transformer substation fire alarm method and a fire monitoring system. The method comprises the steps that first target data collected by a visible light camera, second target data collected by an infrared camera and third target data collected by a smoke sensor are acquired; and performing fire early warning identification based on the first target data, the second target data and the third target data to obtain a plurality of target alarm results, determining the target alarm result as a to-be-corrected result when any one target alarm result is that a fire exists and other target alarm results are that no fire exists, and correcting the to-be-corrected result according to the to-be-corrected result. Correcting the to-be-corrected result according to the target alarm result except the to-be-corrected result to obtain a corrected alarm result; and performing fire alarm at least according to the corrected alarm result. The method solves the problem that in the prior art, a single fire alarm mode is high in false alarm rate in an extreme environment.

Description

Technical Field

[0001] The present invention relates to the technical field of fire early warning, and in particular to a multi-data fusion substation fire alarm method, an alarm device, a computer-readable storage medium and a fire monitoring system. Background Art

[0002] In modern industrial and residential environments, fire safety has always been one of the most concerned issues. Fire is one of the common safety accidents in substations, and smoke is an important feature of the early stage of fire. Accurate and timely identification of smoke is of great significance for fire prevention and control. With the development of technology, fire monitoring and alarm systems have become more and more intelligent and automated, and can use various sensors and algorithms to improve detection accuracy and response speed. Traditional fire alarm systems mainly rely on a single type of sensor such as smoke detectors or temperature sensors. These systems may have the risk of false alarms or missed alarms in some extreme foggy weather, dark environments or strong light weather. In addition, some fires may occur without obvious smoke or temperature changes. Therefore, a fire alarm system with a single sensor may not meet the monitoring needs of complex environments.

[0003] With the development of computer vision and deep learning technology, image-based fire detection methods have also made significant progress. By analyzing the video images captured by the camera, the system can identify visual features such as smoke and flames produced by the fire. This method has high potential for early fire detection, but there is a large error in large spaces or environments with limited vision. In addition, the fire alarm system also needs to consider the problem of multi-target data fusion in complex environments. For example, visible light images, infrared images, and smoke sensor data may need to be used in combination to improve the accuracy of fire detection. This requires complex data processing and analysis algorithms, as well as a deep understanding of the relationship between different data sources. In the optimization process of the fire alarm system, attention should also be paid to the real-time and reliability of the system. Summary of the invention

[0004] The main purpose of the present application is to provide a multi-data fusion substation fire alarm method, alarm device, computer-readable storage medium and fire monitoring system, so as to at least solve the problem of high false alarm rate of single fire alarm mode in extreme environment in the prior art.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a multi-data fusion substation fire alarm method is provided, including: obtaining first target data, second target data and third target data, the first target data being image data collected by a visible light camera, the second target data being image data collected by an infrared camera, and the third target data being concentration data collected by a smoke sensor; performing fire warning identification based on the first target data, the second target data and the third target data respectively, to obtain multiple target alarm results, the target alarm results including a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data and a third target alarm result corresponding to the third target data; in the case where any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; and a fire alarm is performed at least according to the corrected alarm result.

[0006] Optionally, fire warning identification is performed based on the first target data, the second target data and the third target data respectively to obtain multiple target alarm results, including: inputting the first target data into a first target model to obtain the first target alarm result, the first target model is a fire warning model based on visible light images trained based on a support vector machine algorithm or a random forest algorithm; inputting the second target data into a second target model to obtain the second target alarm result, the second target model is a fire warning model based on infrared images trained based on a support vector machine algorithm, a decision tree algorithm or a random forest algorithm; determining the current smoke concentration based on the third target data, performing fire warning identification according to the current smoke concentration and a first threshold value, and obtaining the third target alarm result.

[0007] Optionally, the first target data is input into a first target model to obtain the first target alarm result, including: performing feature recognition based on the first target data to obtain a first target feature, the first target feature including at least one or more of color, texture and shape; converting the first target feature into a numerical feature vector to obtain a first target vector; performing classification recognition based on the first target vector to obtain the first target alarm result.

[0008] Optionally, the first target data is input into a second target model to obtain a second target alarm result, including: performing feature recognition based on the second target data to obtain a second target feature; converting the second target feature into a numerical feature vector to obtain a second target vector; performing classification recognition based on the second target vector to obtain the second target alarm result.

[0009] Optionally, when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result, including: when the first target alarm result is that there is a fire, the first target alarm result is determined as the result to be corrected; heat source identification is performed based on the second target data to obtain a heat source identification result; smoke identification is performed based on the current smoke concentration and a second threshold to obtain a smoke identification result; when the heat source identification result satisfies a first preset condition and the smoke identification result satisfies a second preset condition, the corrected alarm result is determined to be that there is a fire, the first preset condition being that there is at least one heat source, and the second preset condition being that the current smoke concentration is greater than or equal to a second threshold; when the heat source identification result does not satisfy the first preset condition and / or the smoke identification result does not satisfy the second preset condition, the corrected alarm result is determined to be that there is no fire.

[0010] Optionally, when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result, including: when the second target alarm result is that there is a fire, the second target alarm result is determined as the result to be corrected; based on the first target alarm result, a corresponding confidence score is determined to obtain a first confidence score, and based on the third target alarm result, a corresponding confidence score is determined to obtain a second confidence score; based on the first confidence score and the second confidence score, a weighted calculation is performed to obtain a target confidence score; when the target confidence score is less than or equal to a third threshold, the corrected alarm result is determined to be that there is a fire; when the target confidence score is greater than the third threshold, the corrected alarm result is determined to be that there is no fire.

[0011] Optionally, when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result, including: when the third target alarm result is that there is a fire, the second target alarm result is determined as the result to be corrected; the moment when the current smoke concentration in the third target data is equal to the first threshold is recorded to obtain the first moment; the second target data corresponding to the first moment and the second moment are obtained to obtain the first target image and the second target image, the mean of the grayscale value difference corresponding to the first target image and the second target image is calculated to obtain the target value, the second moment is separated from the first moment by a preset time length and the second moment is after the first moment; the second moment and the second moment are obtained The corresponding a target data, obtaining a third target image and a fourth target image, inputting the current smoke density, the third target image and the fourth target image into a third target model to obtain a fifth target image, inputting the fifth target image into the first target model to obtain an updated first target alarm result, wherein the third target model is used to remove the noise generated by smoke in the third target image according to the current smoke density and the fourth target image; when the target value satisfies a third preset condition and the updated first target alarm result satisfies a fourth preset condition, determining that the modified alarm result is that there is a fire, the third preset condition is that the target value is greater than a fourth threshold value, and the fourth preset condition is that the updated first target alarm is that there is a fire; when the target value does not satisfy the third preset condition and / or the updated first target alarm result does not satisfy the fourth preset condition, determining that the modified alarm result is that there is no fire.

[0012] According to another aspect of the present application, a multi-data fusion substation fire alarm device is provided, the device comprising: an acquisition unit, used to acquire first target data, second target data and third target data, the first target data being image data collected by a visible light camera, the second target data being image data collected by an infrared camera, and the third target data being concentration data collected by a smoke sensor; an identification unit, used to perform fire warning identification based on the first target data, the second target data and the third target data, respectively, to obtain multiple target alarm results, the target alarm results comprising a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data and a third target alarm result corresponding to the third target data; a correction unit, used to determine the target alarm result as a result to be corrected when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, and to correct the result to be corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; and an alarm unit, at least according to the corrected alarm result, performs a fire alarm.

[0013] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the method described in any one of the devices where the computer-readable storage medium is located is controlled.

[0014] According to another aspect of the present application, a fire monitoring system is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of the methods described.

[0015] Applying the technical solution of the present application, in the above-mentioned multi-data fusion substation fire alarm method, first, the first target data, the second target data and the third target data are obtained, the first target data is the image data collected by the visible light camera, the second target data is the image data collected by the infrared camera, and the third target data is the concentration data collected by the smoke sensor; then, fire warning identification is performed based on the first target data, the second target data and the third target data, respectively, to obtain multiple target alarm results, the target alarm results include the first target alarm result corresponding to the first target data, the second target alarm result corresponding to the second target data and the third target alarm result corresponding to the third target data; thereafter, when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; finally, a fire alarm is performed at least according to the corrected alarm result. This application adds infrared image recognition and smoke concentration recognition to the traditional visible light image-based fire warning. When any one of the recognition methods triggers an alarm, the alarm information is corrected through the other two alarm methods, and a fire alarm is issued based on the corrected alarm information. This method solves the problem of high false alarm rate of a single fire alarm method in extreme environments in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A hardware structure block diagram of a mobile terminal of a substation fire alarm method with multi-data fusion provided in an embodiment of the present application is shown;

[0017] Figure 2 A schematic flow chart of a substation fire alarm method with multi-data fusion provided according to an embodiment of the present application is shown;

[0018] Figure 3 A structural block diagram of a substation fire alarm device with multi-data fusion provided according to an embodiment of the present application is shown.

[0019] The above drawings include the following reference numerals:

[0020] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0021] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] As introduced in the background technology, traditional smoke recognition methods in the prior art often rely on a single visible light, infrared sensor or smoke sensor device. In order to solve the problem of high false alarm rate of a single fire alarm method in the prior art under extreme environments, the embodiments of the present application provide a multi-data fusion substation fire alarm method, alarm device, computer-readable storage medium and fire monitoring system.

[0025] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0026] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a multi-data fusion substation fire alarm method according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / 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 mobile terminal. Figure 1More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0027] 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 display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. 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 can be connected to the mobile 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 a combination thereof. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile 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 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0028] In this embodiment, a multi-data fusion substation fire alarm method running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings 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 an order different from that here.

[0029] Figure 2 FIG. 1 is a flow chart of a substation fire alarm method using multi-data fusion according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0030] Step S201, acquiring first target data, second target data and third target data, wherein the first target data is image data collected by a visible light camera, the second target data is image data collected by an infrared camera, and the third target data is concentration data collected by a smoke sensor;

[0031] Specifically, the fire alarm system uses visible light cameras and infrared cameras to collect image data, smoke sensors to collect smoke concentration information, and multiple data and multi-data fusion methods to comprehensively analyze the fire situation.

[0032] Step S202, performing fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain multiple target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data;

[0033] Specifically, based on the above-mentioned first target data, image processing technology and computer vision algorithms are used to analyze the images captured by the visible light camera, identify visual features such as flames, smoke or other fire-related features in the image, and obtain the first target alarm result. Based on the above-mentioned second target data, thermal imaging analysis technology is used to identify abnormal heat source areas in the infrared camera image data to obtain the second target alarm result. Based on the above-mentioned third target data, the data on smoke concentration or other particulate matter in the air collected by the smoke sensor is analyzed, and the third target alarm result is obtained through the detected smoke concentration.

[0034] Step S203, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result;

[0035] Specifically, in a fire alarm system, different alarm results may be obtained based on different target data. In this case, if there is at least one result indicating the existence of a fire, and at least one result indicating the absence of a fire, the target alarm result indicating the existence of a fire will be used as the result to be corrected, and the alarm results of other target data will be analyzed, and the result to be corrected will be evaluated to obtain a corrected alarm result.

[0036] Step S204, issuing a fire alarm at least based on the above-mentioned corrected alarm result.

[0037] Specifically, the fire alarm system analyzes the corrected alarm result, determines whether there is a fire risk, and issues a fire alarm.

[0038] Through this embodiment, first, the first target data, the second target data and the third target data are obtained, the first target data being image data collected by a visible light camera, the second target data being image data collected by an infrared camera, and the third target data being concentration data collected by a smoke sensor; then, fire warning identification is performed based on the first target data, the second target data and the third target data, respectively, to obtain multiple target alarm results, the target alarm results including a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data and a third target alarm result corresponding to the third target data; thereafter, in the case where any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; finally, a fire alarm is performed at least according to the corrected alarm result. This application adds infrared image recognition and smoke concentration recognition to the traditional visible light image-based fire warning. When any one of the recognition methods triggers an alarm, the alarm information is corrected through the other two alarm methods, and a fire alarm is issued based on the corrected alarm information. This method solves the problem of high false alarm rate of a single fire alarm method in extreme environments in the prior art.

[0039] In order to obtain multiple alarm results, in an optional implementation, the above step S202 includes:

[0040] Step S2021, inputting the first target data into a first target model to obtain the first target alarm result, wherein the first target model is a fire warning model based on visible light images obtained by training based on a support vector machine algorithm or a random forest algorithm;

[0041] Specifically, the real-time image data collected by the visible light camera is input into the first target model deployed by the fire alarm system. The first target model is based on a support vector machine algorithm or a random forest algorithm, and is trained using image data of normal conditions and fire conditions. The support vector machine algorithm uses the extracted features to train a classifier that can identify fire and non-fire scenes in the image data. The random forest algorithm constructs a forest consisting of multiple decision trees that work together to improve the accuracy and robustness of the classification. The first target data is input into the trained first target model, and the model evaluates the input image and gives a probability or label of a fire occurrence to obtain the above-mentioned first target alarm result.

[0042] Step S2022, inputting the second target data into a second target model to obtain the second target alarm result, wherein the second target model is a fire warning model based on infrared images obtained by training based on a support vector machine algorithm, a decision tree algorithm or a random forest algorithm;

[0043] Specifically, the real-time image data collected by the infrared camera is input into the second target model deployed in the fire alarm system. The second target model is based on a support vector machine algorithm, a decision tree algorithm or a random forest algorithm. The selected algorithm is trained using labeled historical normal and fire infrared image data. The evaluated model is deployed into the real-time monitoring system, and the second target data is input into the second target model. The model will analyze the input infrared image and evaluate the possibility of a fire to obtain the above-mentioned second target alarm result.

[0044] Step S2023, determining the current smoke density based on the third target data, performing fire warning identification according to the current smoke density and the first threshold value, and obtaining the third target alarm result.

[0045] Specifically, the fire alarm system also installs smoke sensors at key locations to collect smoke concentration data. If the current smoke concentration exceeds the first threshold, the system identifies that there may be a fire, triggers the early warning program, and generates a third target alarm result.

[0046] In order to obtain the first target alarm result, in an optional implementation, the above step S2021 includes:

[0047] Step S20211, performing feature recognition based on the first target data to obtain a first target feature, where the first target feature at least includes one or more of color, texture and shape;

[0048] Specifically, feature recognition is performed based on image data collected by a visible light camera in a monitored area. The image data is first preprocessed. The preprocessing step may include adjusting contrast, brightness, and applying filters to remove noise, etc. At least color features are extracted from the preprocessed image data to identify color features unique to fires, such as abnormal red or yellow tones, which may be related to flames or hot areas. Texture features are extracted to identify irregular or smoke patterns generated by fires and distinguish flames and smoke from other ordinary backgrounds. Shape features are extracted to identify specific shapes in the image, such as linear or fan-shaped outlines of flames, or other visual features related to fires. A first target feature is obtained based on one or more of the above features.

[0049] Step S20212, converting the first target feature into a numerical feature vector to obtain a first target vector;

[0050] Specifically, the color, texture, shape and other features in the above-mentioned first target features are converted into numerical form. Commonly used methods include histograms, gradient descriptors, geometric features, etc. In order to eliminate the dimensional influence between different features, the feature vectors are normalized so that all features have the same scale. All quantized features are combined into a unified feature vector. Each eigenvalue is an element in the vector, and the first target vector is constructed according to the eigenvalues.

[0051] Step S20213, performing classification and identification based on the above-mentioned first target vector to obtain the above-mentioned first target alarm result.

[0052] Specifically, the first target vector is classified and identified using classification algorithms such as support vector machines, random forests, and neural networks. According to the fire classification label output by the model, it is decided whether to trigger an alarm to obtain the first target alarm result.

[0053] In order to obtain the second target alarm result, in an optional implementation, the above step S2022 includes:

[0054] Step S20221, performing feature recognition based on the second target data to obtain a second target feature;

[0055] Specifically, based on the image data collected by the infrared camera in the monitoring area, the image processing software is used to analyze the heat source in the image, and the size, position, shape and other features of the heat source in the image are extracted to obtain the second target feature.

[0056] Step S20222, converting the second target feature into a numerical feature vector to obtain a second target vector;

[0057] Specifically, for each identified feature, a quantization method is determined and converted into a numerical form, the feature values ​​are normalized to a uniform range, and all quantized feature values ​​are combined into a single vector to obtain a second target vector.

[0058] Step S20223, classify and identify based on the second target vector to obtain the second target alarm result.

[0059] Specifically, a suitable classification and recognition algorithm is used to classify and recognize the second target vector, and according to the output of the model, it is determined whether to trigger an alarm to obtain a second target alarm result.

[0060] In order to obtain a corrected alarm result, in an optional implementation, the above step S203 includes:

[0061] Step S20301, when the first target alarm result is that there is a fire, determining the first target alarm result as the result to be corrected;

[0062] Specifically, when the first target alarm result preliminarily determines that there is a fire based on the feature recognition and classification model of the visible light image, the correction process will be triggered and the first target alarm result will be determined as the result to be corrected for further verification.

[0063] Step S20302, performing heat source identification based on the second target data to obtain a heat source identification result;

[0064] Specifically, an infrared camera is used to capture infrared images within the monitored area, and image processing technology is used to detect high-temperature areas in the image that may indicate the presence of heat sources. The distribution, size, shape, and temperature of the hot spots are analyzed to generate heat source identification results.

[0065] Step S20303, performing smoke recognition based on the current smoke concentration and the second threshold value to obtain a smoke recognition result;

[0066] Specifically, the current smoke concentration data measured by the smoke sensor is collected, and according to the second threshold value set by the system, it is determined whether the smoke concentration has reached a level causing an early warning, and if the smoke concentration exceeds the threshold value, it may indicate the existence of a potential fire.

[0067] Step S20404, when the heat source identification result satisfies a first preset condition and the smoke identification result satisfies a second preset condition, determining that the modified alarm result is a fire, the first preset condition being that there is at least one heat source, and the second preset condition being that the current smoke concentration is greater than or equal to a second threshold;

[0068] Specifically, according to the heat source identification result, the existence of a fire can be determined only when at least one heat source is detected and the smoke concentration in the monitored air also meets the second preset condition, and the revised alarm result is determined to be the existence of a fire.

[0069] Step S20405: When the heat source identification result does not satisfy the first preset condition and / or the smoke identification result does not satisfy the second preset condition, determining that the corrected alarm result is that there is no fire.

[0070] Specifically, if the heat source identification result shows that no heat source is detected, or the smoke identification result shows that the smoke concentration is lower than the fire alarm threshold, these situations indicate that there may be no fire, and the revised alarm result is determined to be no fire.

[0071] In order to obtain a corrected alarm result, in an optional implementation, the above step S203 includes:

[0072] Step S20311, when the second target alarm result is that there is a fire, determining the second target alarm result as the result to be corrected;

[0073] Specifically, when the second target alarm result determines that there is a fire based on the infrared image data, the second target alarm result is determined as a result to be corrected, awaiting further verification.

[0074] Step S20312, determining a corresponding confidence score based on the first target alarm result to obtain a first confidence score, and determining a corresponding confidence score based on the third target alarm result to obtain a second confidence score;

[0075] Specifically, in the fire alarm system, based on a pre-set confidence quantification standard, the first target alarm result obtained based on the visible light image data is evaluated, and the corresponding confidence score is determined to obtain a first confidence score. If there is a third target alarm result, the confidence score of the third target alarm result is determined at the same time to obtain a second confidence score.

[0076] Step S20313, performing weighted calculation based on the first confidence score and the second confidence score to obtain a target confidence score;

[0077] Specifically, weights are assigned to the first confidence score and the second confidence score according to their importance to the final judgment result, and a weighted average formula is used to calculate the target confidence score.

[0078] Step S20314, when the target confidence score is less than or equal to a third threshold, determining that the modified alarm result is a fire;

[0079] Specifically, a third threshold is pre-set in the fire alarm system to determine the possibility of the existence of a fire. The calculated target confidence score is compared with the third threshold. If the target confidence score is less than or equal to the third threshold, the system determines that a fire exists.

[0080] Step S20315: When the target confidence score is greater than a third threshold, determine that the revised alarm result is that there is no fire.

[0081] Specifically, the target confidence score is judged in the fire alarm system, and if the target confidence score is greater than a third threshold, the system determines that there is no fire.

[0082] In order to obtain a corrected alarm result, in an optional implementation, the above step S203 includes:

[0083] Step S20321, when the third target alarm result is that there is a fire, determining the third target alarm result as the result to be corrected;

[0084] Specifically, when the third target alarm result determines that there is a fire based on the smoke density data,

[0085] Step S20322, recording the moment when the current smoke density in the third target data is equal to the first threshold, to obtain a first moment;

[0086] Specifically, in the fire alarm system, a preset value of smoke concentration is pre-set as a first threshold value, which is used as a trigger point for fire warning. When the smoke concentration reaches the first threshold value, this moment is recorded to obtain the first moment.

[0087] Step S20323, obtaining the second target data corresponding to the first moment and the second moment, obtaining the first target image and the second target image, calculating the average of the gray value difference between the first target image and the second target image, and obtaining the target value, wherein the second moment is separated from the first moment by a preset time and the second moment is after the first moment;

[0088] Specifically, a moment at a preset time interval after the first moment is determined as the second moment, and at the first moment and the second moment, two images at corresponding moments are acquired from the infrared camera and determined as the first target image and the second target image, and corresponding grayscale values ​​are extracted therefrom, and the difference in grayscale values ​​of corresponding pixel points of the two images is calculated, and the average of all grayscale value differences is calculated to obtain a target value, and the target value is analyzed to evaluate the degree of change between the images at the two moments.

[0089] Step S20324, obtaining the first target data corresponding to the first moment and the second moment, obtaining a third target image and a fourth target image, inputting the current smoke concentration, the third target image and the fourth target image into a third target model, obtaining a fifth target image, inputting the fifth target image into the first target model, obtaining an updated alarm result of the first target, and the third target model is used to remove the noise generated by smoke in the third target image according to the current smoke concentration and the fourth target image;

[0090] Specifically, at the first moment and the second moment, the first target data collected by the visible light camera is collected to obtain the third target image and the fourth target image respectively, the current smoke concentration data is obtained from the smoke sensor, and the smoke concentration, the third target image and the fourth target image are input into the third target model deployed in the fire alarm system. After the third target model is processed, a fifth target image with smoke noise removed from the image is generated based on the third target image, and the fifth target image is used as input and input into the first target model. An updated first target alarm result is generated according to the fifth target image to determine whether a fire has occurred.

[0091] Step S20325, when the target value satisfies a third preset condition and the updated first target alarm result satisfies a fourth preset condition, determining that the revised alarm result is a fire, the third preset condition being that the target value is greater than a fourth threshold, and the fourth preset condition being that the updated first target alarm is a fire;

[0092] Specifically, a fourth threshold is preset in the fire alarm system, which is used to determine whether the updated first target alarm result indicates the existence of a fire. If the target value meets the third preset condition and the updated alarm result meets the fourth preset condition, the revised alarm result is determined to be the existence of a fire.

[0093] Step S20326: When the target value does not satisfy the third preset condition and / or the updated first target alarm result does not satisfy the fourth preset condition, determine that the revised alarm result is that there is no fire.

[0094] Specifically, in the fire alarm system, the calculated target value does not satisfy the third preset condition, that is, the target value is less than or equal to the fourth threshold value, and the revised alarm result is determined to be that there is no fire.

[0095] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the substation fire alarm method of the present application with multi-data fusion will be described in detail below in combination with specific embodiments.

[0096] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0097] The embodiment of the present application also provides a substation fire alarm device with multi-data fusion. It should be noted that the substation fire alarm device with multi-data fusion in the embodiment of the present application can be used to execute the substation fire alarm method for multi-data fusion provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0098] The following is an introduction to the substation fire alarm device with multi-data fusion provided in an embodiment of the present application.

[0099] Figure 3 1 is a structural block diagram of a substation fire alarm device with multi-data fusion according to an embodiment of the present application. Figure 3 As shown, the device comprises:

[0100] The acquisition unit 10 is used to acquire first target data, second target data and third target data, wherein the first target data is image data acquired by a visible light camera, the second target data is image data acquired by an infrared camera, and the third target data is concentration data acquired by a smoke sensor;

[0101] Specifically, the fire alarm system uses visible light cameras and infrared cameras to collect image data, smoke sensors to collect smoke concentration information, and multiple data and multi-data fusion methods to comprehensively analyze the fire situation.

[0102] an identification unit 20, configured to perform fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain a plurality of target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data;

[0103] Specifically, based on the above-mentioned first target data, image processing technology and computer vision algorithms are used to analyze the images captured by the visible light camera, identify visual features such as flames, smoke or other fire-related visual features in the image, and obtain the first target alarm result. Based on the above-mentioned second target data, thermal imaging analysis technology is used to identify hot spots and temperature abnormality areas in the infrared camera image data to obtain the second target alarm result. Based on the above-mentioned third target data, the data on smoke concentration or other particulate matter in the air collected by the smoke sensor is analyzed, and the third target alarm result is obtained through the detected smoke concentration.

[0104] The correction unit 30 is used to determine the target alarm result as a result to be corrected when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, and correct the result to be corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result;

[0105] Specifically, in a fire alarm system, different alarm results may be obtained based on different target data. In this case, if there is at least one result indicating the existence of a fire, and at least one result indicating the absence of a fire, the target alarm result indicating the existence of a fire will be used as the result to be corrected, and the alarm results of other target data will be analyzed, and the result to be corrected will be evaluated to obtain a corrected alarm result.

[0106] The alarm unit 40 is used to issue a fire alarm at least according to the above-mentioned corrected alarm result.

[0107] Specifically, the fire alarm system analyzes the corrected alarm result, determines whether there is a fire risk, and issues a fire alarm.

[0108] Through this embodiment, the acquisition unit acquires the first target data, the second target data and the third target data, the first target data is the image data collected by the visible light camera, the second target data is the image data collected by the infrared camera, and the third target data is the concentration data collected by the smoke sensor; the identification unit performs fire warning identification based on the first target data, the second target data and the third target data, respectively, to obtain multiple target alarm results, the target alarm results include the first target alarm result corresponding to the first target data, the second target alarm result corresponding to the second target data and the third target alarm result corresponding to the third target data; the correction unit determines the target alarm result as the result to be corrected when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, and corrects the result to be corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; the alarm unit performs a fire alarm at least according to the corrected alarm result. This application adds infrared image recognition and smoke concentration recognition to the traditional visible light image-based fire warning. When any one of the recognition methods triggers an alarm, the alarm information is corrected through the other two alarm methods, and a fire alarm is issued based on the corrected alarm information. This method solves the problem of high false alarm rate of a single fire alarm method in extreme environments in the prior art.

[0109] In order to obtain multiple alarm results, in an optional implementation, the identification unit includes:

[0110] A first identification subunit is used to input the first target data into a first target model to obtain the first target alarm result, wherein the first target model is a fire warning model based on visible light images obtained by training based on a support vector machine algorithm or a random forest algorithm;

[0111] Specifically, the real-time image data collected by the visible light camera is input into the first target model deployed by the fire alarm system. The first target model is based on a support vector machine algorithm or a random forest algorithm, and is trained using image data of normal conditions and fire conditions. The support vector machine algorithm uses the extracted features to train a classifier that can identify fire and non-fire scenes in the image data. The random forest algorithm constructs a forest consisting of multiple decision trees that work together to improve the accuracy and robustness of the classification. The first target data is input into the trained first target model, and the model evaluates the input image and gives a probability or label of a fire occurrence to obtain the above-mentioned first target alarm result.

[0112] A second identification subunit is used to input the second target data into a second target model to obtain the second target alarm result, wherein the second target model is a fire warning model based on infrared images obtained by training based on a support vector machine algorithm, a decision tree algorithm or a random forest algorithm;

[0113] Specifically, the real-time image data collected by the infrared camera is input into the second target model deployed in the fire alarm system. The second target model is based on a support vector machine algorithm, a decision tree algorithm or a random forest algorithm. The selected algorithm is trained using labeled historical normal and fire infrared image data. The evaluated model is deployed into the real-time monitoring system, and the second target data is input into the second target model. The model will analyze the input infrared image and evaluate the possibility of a fire to obtain the above-mentioned second target alarm result.

[0114] The third identification subunit is used to determine the current smoke concentration based on the above third target data, perform fire warning identification according to the above current smoke concentration and the first threshold value, and obtain the above third target alarm result.

[0115] Specifically, the fire alarm system also installs smoke sensors at key locations to collect smoke concentration data. If the current smoke concentration exceeds the first threshold, the system identifies that there may be a fire, triggers the early warning program, and generates a third target alarm result.

[0116] In order to obtain the first target alarm result, in an optional implementation manner, the first identification subunit of the above steps includes:

[0117] A fourth recognition subunit is used to perform feature recognition based on the first target data to obtain a first target feature, where the first target feature includes at least one or more of color, texture and shape;

[0118] Specifically, feature recognition is performed based on image data collected by a visible light camera in a monitored area. The image data is first preprocessed. The preprocessing step may include adjusting contrast, brightness, and applying filters to remove noise, etc. At least color features are extracted from the preprocessed image data to identify color features unique to fires, such as abnormal red or yellow tones, which may be related to flames or hot areas. Texture features are extracted to identify irregular or smoke patterns generated by fires and distinguish flames and smoke from other ordinary backgrounds. Shape features are extracted to identify specific shapes in the image, such as linear or fan-shaped outlines of flames, or other visual features related to fires. A first target feature is obtained based on one or more of the above features.

[0119] A first calculation subunit is used to convert the first target feature into a numerical feature vector to obtain a first target vector;

[0120] Specifically, the color, texture, shape and other features in the above-mentioned first target features are converted into numerical form. Commonly used methods include histograms, gradient descriptors, geometric features, etc. In order to eliminate the dimensional influence between different features, the feature vectors are normalized so that all features have the same scale. All quantized features are combined into a unified feature vector. Each eigenvalue is an element in the vector, and the first target vector is constructed according to the eigenvalues.

[0121] The fifth identification subunit is used to perform classification and identification based on the above-mentioned first target vector to obtain the above-mentioned first target alarm result.

[0122] Specifically, the first target vector is classified and identified using classification algorithms such as support vector machines, random forests, and neural networks. According to the fire classification label output by the model, it is decided whether to trigger an alarm to obtain the first target alarm result.

[0123] In order to obtain the second target alarm result, in an optional implementation, the second identification subunit includes:

[0124] a sixth identification subunit, configured to perform feature recognition based on the second target data to obtain a second target feature;

[0125] Specifically, based on the image data collected by the infrared camera in the monitoring area, the image processing software is used to analyze the heat source in the image, and the size, position, shape and other features of the heat source in the image are extracted to obtain the second target feature.

[0126] A second calculation subunit is used to convert the second target feature into a numerical feature vector to obtain a second target vector;

[0127] Specifically, for each identified feature, a quantization method is determined and converted into a numerical form, the feature values ​​are normalized to a uniform range, and all quantized feature values ​​are combined into a single vector to obtain a second target vector.

[0128] The seventh identification subunit is used to perform classification and identification based on the above-mentioned second target vector to obtain the above-mentioned second target alarm result.

[0129] Specifically, a suitable classification and recognition algorithm is used to classify and recognize the second target vector, and according to the output of the model, it is determined whether to trigger an alarm to obtain a second target alarm result.

[0130] In order to obtain a corrected alarm result, in an optional implementation, the identification unit includes:

[0131] A first determining subunit is used to determine the first target alarm result as the result to be corrected when the first target alarm result is that there is a fire;

[0132] Specifically, when the first target alarm result preliminarily determines that there is a fire based on the feature recognition and classification model of the visible light image, the correction process will be triggered and the first target alarm result will be determined as the result to be corrected for further verification.

[0133] an eighth identification subunit, configured to perform heat source identification based on the second target data to obtain a heat source identification result;

[0134] Specifically, an infrared camera is used to capture infrared images within the monitored area, and image processing technology is used to detect high-temperature areas in the image that may indicate the presence of heat sources. The distribution, size, shape, and temperature of the hot spots are analyzed to generate heat source identification results.

[0135] A ninth identification subunit, configured to perform smoke identification based on the current smoke concentration and the second threshold value to obtain a smoke identification result;

[0136] Specifically, the current smoke concentration data measured by the smoke sensor is collected, and according to the second threshold value set by the system, it is determined whether the smoke concentration has reached a level causing an early warning, and if the smoke concentration exceeds the threshold value, it may indicate the existence of a potential fire.

[0137] A second determination subunit is used to determine that the modified alarm result is that there is a fire if the heat source identification result satisfies a first preset condition and the smoke identification result satisfies a second preset condition, the first preset condition being that there is at least one heat source, and the second preset condition being that the current smoke concentration is greater than or equal to a second threshold value;

[0138] Specifically, according to the heat source identification result, the existence of a fire can be determined only when at least one heat source is detected and the smoke concentration in the monitored air also meets the second preset condition, and the revised alarm result is determined to be the existence of a fire.

[0139] The third determination subunit is used to determine that the modified alarm result is that there is no fire when the heat source identification result does not meet the first preset condition and / or the smoke identification result does not meet the second preset condition.

[0140] Specifically, if the heat source identification result shows that no heat source is detected, or the smoke identification result shows that the smoke concentration is lower than the fire alarm threshold, these situations indicate that there may be no fire, and the revised alarm result is determined to be no fire.

[0141] In order to obtain a corrected alarm result, in an optional implementation manner, the correction unit includes:

[0142] A first correction subunit is used to determine the second target alarm result as the result to be corrected when the second target alarm result is that there is a fire;

[0143] Specifically, when the second target alarm result determines that there is a fire based on the infrared image data, the second target alarm result is determined as a result to be corrected, awaiting further verification.

[0144] A fourth determination subunit is used to determine a corresponding confidence score based on the first target alarm result to obtain a first confidence score, and to determine a corresponding confidence score based on the third target alarm result to obtain a second confidence score;

[0145] Specifically, in the fire alarm system, based on a pre-set confidence quantification standard, the first target alarm result obtained based on the visible light image data is evaluated, and the corresponding confidence score is determined to obtain a first confidence score. If there is a third target alarm result, the confidence score of the third target alarm result is determined at the same time to obtain a second confidence score.

[0146] A third calculation subunit, configured to perform weighted calculation based on the first confidence score and the second confidence score to obtain a target confidence score;

[0147] Specifically, weights are assigned to the first confidence score and the second confidence score according to their importance to the final judgment result, and a weighted average formula is used to calculate the target confidence score.

[0148] A second correction subunit is used to determine that the corrected alarm result is the presence of a fire when the target confidence score is less than or equal to a third threshold;

[0149] Specifically, a third threshold is pre-set in the fire alarm system to determine the possibility of the existence of a fire. The calculated target confidence score is compared with the third threshold. If the target confidence score is less than or equal to the third threshold, the system determines that a fire exists.

[0150] The third correction subunit is used to determine that the corrected alarm result is that there is no fire when the target confidence score is greater than a third threshold.

[0151] Specifically, the target confidence score is judged in the fire alarm system, and if the target confidence score is greater than a third threshold, the system determines that there is no fire.

[0152] In order to obtain a corrected alarm result, in an optional implementation manner, the correction unit includes:

[0153] A fourth correction subunit is used for determining the third target alarm result as the result to be corrected when the third target alarm result is that there is a fire;

[0154] Specifically, when the third target alarm result determines that there is a fire based on the smoke density data,

[0155] A first acquisition subunit is used to record the moment when the current smoke concentration in the third target data is equal to the first threshold value to obtain a first moment;

[0156] Specifically, in the fire alarm system, a preset value of smoke concentration is pre-set as a first threshold value, which is used as a trigger point for fire warning. When the smoke concentration reaches the first threshold value, this moment is recorded to obtain the first moment.

[0157] A second acquisition subunit is used to acquire the second target data corresponding to the first moment and the second moment, obtain the first target image and the second target image, calculate the average of the gray value difference between the first target image and the second target image, and obtain the target value, wherein the second moment is separated from the first moment by a preset time and the second moment is after the first moment;

[0158] Specifically, a moment at a preset time interval after the first moment is determined as the second moment, and at the first moment and the second moment, two images at corresponding moments are acquired from the infrared camera and determined as the first target image and the second target image, and corresponding grayscale values ​​are extracted therefrom, and the difference in grayscale values ​​of corresponding pixel points of the two images is calculated, and the average of all grayscale value differences is calculated to obtain a target value, and the target value is analyzed to evaluate the degree of change between the images at the two moments.

[0159] a third acquisition subunit, for acquiring the first target data corresponding to the first moment and the second moment, obtaining a third target image and a fourth target image, inputting the current smoke concentration, the third target image and the fourth target image into a third target model, obtaining a fifth target image, inputting the fifth target image into the first target model, obtaining an updated alarm result of the first target, and the third target model is used to remove noise generated by smoke in the third target image according to the current smoke concentration and the fourth target image;

[0160] Specifically, at the first moment and the second moment, the first target data collected by the visible light camera is collected to obtain the third target image and the fourth target image respectively, the current smoke concentration data is obtained from the smoke sensor, and the smoke concentration, the third target image and the fourth target image are input into the third target model deployed in the fire alarm system. After the third target model is processed, a fifth target image with smoke noise removed from the image is generated based on the third target image, and the fifth target image is used as input and input into the first target model. An updated first target alarm result is generated according to the fifth target image to determine whether a fire has occurred.

[0161] A fifth correction subunit is used to determine that the corrected alarm result is that there is a fire if the target value satisfies a third preset condition and the updated first target alarm result satisfies a fourth preset condition, the third preset condition being that the target value is greater than a fourth threshold value, and the fourth preset condition being that the updated first target alarm is that there is a fire;

[0162] Specifically, a fourth threshold is preset in the fire alarm system, which is used to determine whether the updated first target alarm result indicates the existence of a fire. If the target value meets the third preset condition and the updated alarm result meets the fourth preset condition, the revised alarm result is determined to be the existence of a fire.

[0163] The sixth correction subunit is used to determine that the corrected alarm result is that there is no fire when the target value does not meet the third preset condition and / or the updated first target alarm result does not meet the fourth preset condition.

[0164] Specifically, in the fire alarm system, the calculated target value does not satisfy the third preset condition, that is, the target value is less than or equal to the fourth threshold value, and the revised alarm result is determined to be that there is no fire.

[0165] The above-mentioned multi-data fusion substation fire alarm device includes a processor and a memory. The above-mentioned acquisition unit, identification unit, correction unit, alarm unit, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in any combination.

[0166] The processor includes a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the accuracy of fire alarm can be improved by adjusting the kernel parameters.

[0167] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0168] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the substation fire alarm method with multi-data fusion.

[0169] Specifically, the multi-data fusion substation fire alarm method includes:

[0170] Step S201, acquiring first target data, second target data and third target data, wherein the first target data is image data collected by a visible light camera, the second target data is image data collected by an infrared camera, and the third target data is concentration data collected by a smoke sensor;

[0171] Specifically, the fire alarm system uses visible light cameras and infrared cameras to collect image data, smoke sensors to collect smoke concentration information, and multiple data and multi-data fusion methods to comprehensively analyze the fire situation.

[0172] Step S202, performing fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain multiple target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data;

[0173] Specifically, based on the above-mentioned first target data, image processing technology and computer vision algorithms are used to analyze the images captured by the visible light camera, identify visual features such as flames, smoke or other fire-related visual features in the image, and obtain the first target alarm result. Based on the above-mentioned second target data, thermal imaging analysis technology is used to identify hot spots and temperature abnormality areas in the infrared camera image data to obtain the second target alarm result. Based on the above-mentioned third target data, the data on smoke concentration or other particulate matter in the air collected by the smoke sensor is analyzed, and the third target alarm result is obtained through the detected smoke concentration.

[0174] Step S203, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result;

[0175] Specifically, in a fire alarm system, different alarm results may be obtained based on different target data. In this case, if there is at least one result indicating the existence of a fire, and at least one result indicating the absence of a fire, the target alarm result indicating the existence of a fire will be used as the result to be corrected, and the alarm results of other target data will be analyzed, and the result to be corrected will be evaluated to obtain a corrected alarm result.

[0176] Step S204, issuing a fire alarm at least based on the above-mentioned corrected alarm result.

[0177] Specifically, the fire alarm system analyzes the corrected alarm result, determines whether there is a fire risk, and issues a fire alarm.

[0178] An embodiment of the present invention provides a processor, which is used to run a program, wherein the substation fire alarm method of multi-data fusion is executed when the program is running.

[0179] Specifically, the multi-data fusion substation fire alarm method includes:

[0180] Step S201, acquiring first target data, second target data and third target data, wherein the first target data is image data collected by a visible light camera, the second target data is image data collected by an infrared camera, and the third target data is concentration data collected by a smoke sensor;

[0181] Specifically, the fire alarm system uses visible light cameras and infrared cameras to collect image data, smoke sensors to collect smoke concentration information, and multiple data and multi-data fusion methods to comprehensively analyze the fire situation.

[0182] Step S202, performing fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain multiple target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data;

[0183] Specifically, based on the above-mentioned first target data, image processing technology and computer vision algorithms are used to analyze the images captured by the visible light camera, identify visual features such as flames, smoke or other fire-related visual features in the image, and obtain the first target alarm result. Based on the above-mentioned second target data, thermal imaging analysis technology is used to identify hot spots and temperature abnormality areas in the infrared camera image data to obtain the second target alarm result. Based on the above-mentioned third target data, the data on smoke concentration or other particulate matter in the air collected by the smoke sensor is analyzed, and the third target alarm result is obtained through the detected smoke concentration.

[0184] Step S203, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result;

[0185] Specifically, in a fire alarm system, different alarm results may be obtained based on different target data. In this case, if there is at least one result indicating the existence of a fire, and at least one result indicating the absence of a fire, the target alarm result indicating the existence of a fire will be used as the result to be corrected, and the alarm results of other target data will be analyzed, and the result to be corrected will be evaluated to obtain a corrected alarm result.

[0186] Step S204, issuing a fire alarm at least based on the above-mentioned corrected alarm result.

[0187] Specifically, the fire alarm system analyzes the corrected alarm result, determines whether there is a fire risk, and issues a fire alarm.

[0188] An embodiment of the present invention provides a communication system, the communication system comprising a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the following steps are implemented:

[0189] Step S201, acquiring first target data, second target data and third target data, wherein the first target data is image data collected by a visible light camera, the second target data is image data collected by an infrared camera, and the third target data is concentration data collected by a smoke sensor;

[0190] Specifically, the fire alarm system uses visible light cameras and infrared cameras to collect image data, smoke sensors to collect smoke concentration information, and multiple data and multi-data fusion methods to comprehensively analyze the fire situation.

[0191] Step S202, performing fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain multiple target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data;

[0192] Specifically, based on the above-mentioned first target data, image processing technology and computer vision algorithms are used to analyze the images captured by the visible light camera, identify visual features such as flames, smoke or other fire-related visual features in the image, and obtain the first target alarm result. Based on the above-mentioned second target data, thermal imaging analysis technology is used to identify hot spots and temperature abnormality areas in the infrared camera image data to obtain the second target alarm result. Based on the above-mentioned third target data, the data on smoke concentration or other particulate matter in the air collected by the smoke sensor is analyzed, and the third target alarm result is obtained through the detected smoke concentration.

[0193] Step S203, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result;

[0194] Specifically, in a fire alarm system, different alarm results may be obtained based on different target data. In this case, if there is at least one result indicating the existence of a fire, and at least one result indicating the absence of a fire, the target alarm result indicating the existence of a fire will be used as the result to be corrected, and the alarm results of other target data will be analyzed, and the result to be corrected will be evaluated to obtain a corrected alarm result.

[0195] Step S204, issuing a fire alarm at least based on the above-mentioned corrected alarm result.

[0196] Specifically, the fire alarm system analyzes the corrected alarm result, determines whether there is a fire risk, and issues a fire alarm.

[0197] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing at least the following method steps:

[0198] Step S201, acquiring first target data, second target data and third target data, wherein the first target data is image data collected by a visible light camera, the second target data is image data collected by an infrared camera, and the third target data is concentration data collected by a smoke sensor;

[0199] Specifically, the fire alarm system uses visible light cameras and infrared cameras to collect image data, smoke sensors to collect smoke concentration information, and multiple data and multi-data fusion methods to comprehensively analyze the fire situation.

[0200] Step S202, performing fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain multiple target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data;

[0201] Specifically, based on the above-mentioned first target data, image processing technology and computer vision algorithms are used to analyze the images captured by the visible light camera, identify visual features such as flames, smoke or other fire-related visual features in the image, and obtain the first target alarm result. Based on the above-mentioned second target data, thermal imaging analysis technology is used to identify hot spots and temperature abnormality areas in the infrared camera image data to obtain the second target alarm result. Based on the above-mentioned third target data, the data on smoke concentration or other particulate matter in the air collected by the smoke sensor is analyzed, and the third target alarm result is obtained through the detected smoke concentration.

[0202] Step S203, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result;

[0203] Specifically, in a fire alarm system, different alarm results may be obtained based on different target data. In this case, if there is at least one result indicating the existence of a fire, and at least one result indicating the absence of a fire, the target alarm result indicating the existence of a fire will be used as the result to be corrected, and the alarm results of other target data will be analyzed, and the result to be corrected will be evaluated to obtain a corrected alarm result.

[0204] Step S204, issuing a fire alarm at least based on the above-mentioned corrected alarm result.

[0205] Specifically, the fire alarm system analyzes the corrected alarm result, determines whether there is a fire risk, and issues a fire alarm.

[0206] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0207] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0208] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0209] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0210] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0211] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0212] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0213] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0214] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0215] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0216] 1) The multi-data fusion substation fire alarm method of the present application first obtains the first target data, the second target data and the third target data, wherein the first target data is the image data collected by the visible light camera, the second target data is the image data collected by the infrared camera, and the third target data is the concentration data collected by the smoke sensor; then, fire warning identification is performed based on the first target data, the second target data and the third target data, respectively, to obtain multiple target alarm results, wherein the target alarm results include the first target alarm result corresponding to the first target data, the second target alarm result corresponding to the second target data and the third target alarm result corresponding to the third target data; thereafter, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; finally, a fire alarm is performed at least according to the corrected alarm result. This application adds infrared image recognition and smoke concentration recognition to the traditional visible light image-based fire warning. When any one of the recognition methods triggers an alarm, the alarm information is corrected through the other two alarm methods, and a fire alarm is issued based on the corrected alarm information. This method solves the problem of high false alarm rate of a single fire alarm method in extreme environments in the prior art.

[0217] 2) In the multi-data fusion substation fire alarm device of the present application, the acquisition unit acquires the first target data, the second target data and the third target data, the first target data is the image data collected by the visible light camera, the second target data is the image data collected by the infrared camera, and the third target data is the concentration data collected by the smoke sensor; the identification unit performs fire warning identification based on the first target data, the second target data and the third target data, respectively, to obtain multiple target alarm results, the target alarm results include the first target alarm result corresponding to the first target data, the second target alarm result corresponding to the second target data and the third target alarm result corresponding to the third target data; the correction unit determines the target alarm result as the result to be corrected when any one of the target alarm results is that there is a fire and the other target alarm results are that there is no fire, and corrects the result to be corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; the alarm unit performs a fire alarm at least according to the corrected alarm result. This application adds infrared image recognition and smoke concentration recognition to the traditional visible light image-based fire warning. When any one of the recognition methods triggers an alarm, the alarm information is corrected through the other two alarm methods, and a fire alarm is issued based on the corrected alarm information. This method solves the problem of high false alarm rate of a single fire alarm method in extreme environments in the prior art.

[0218] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A multi-data fusion substation fire alarm method, characterized in that: include: Acquire first target data, second target data and third target data, wherein the first target data is image data collected by a visible light camera, the second target data is image data collected by an infrared camera, and the third target data is concentration data collected by a smoke sensor; Fire warning identification is performed based on the first target data, the second target data, and the third target data respectively, to obtain multiple target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data; In the case where any one of the target alarm results indicates that there is a fire and the other target alarm results indicate that there is no fire, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; A fire alarm is issued at least based on the corrected alarm result.

2. The method according to claim 1, characterized in that Fire warning identification is performed based on the first target data, the second target data, and the third target data respectively, to obtain multiple target alarm results, including: Inputting the first target data into a first target model to obtain the first target alarm result, wherein the first target model is a fire warning model based on visible light images obtained by training based on a support vector machine algorithm or a random forest algorithm; Inputting the second target data into a second target model to obtain the second target alarm result, wherein the second target model is a fire warning model based on infrared images obtained by training based on a support vector machine algorithm, a decision tree algorithm or a random forest algorithm; The current smoke concentration is determined based on the third target data, and fire warning identification is performed according to the current smoke concentration and the first threshold value to obtain the third target alarm result.

3. The method according to claim 2, characterized in that Inputting the first target data into a first target model to obtain the first target alarm result includes: Performing feature recognition based on the first target data to obtain a first target feature, where the first target feature includes at least one or more of color, texture, and shape; Converting the first target feature into a numerical feature vector to obtain a first target vector; Classification and identification are performed based on the first target vector to obtain the first target alarm result.

4. The method according to claim 2, characterized in that: Inputting the first target data into a second target model to obtain a second target alarm result includes: Performing feature recognition based on the second target data to obtain a second target feature; Converting the second target feature into a numerical feature vector to obtain a second target vector; Classification and identification are performed based on the second target vector to obtain the second target alarm result.

5. The method according to claim 2, characterized in that: In the case where any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result, including: When the first target alarm result is that there is a fire, determining the first target alarm result as the result to be corrected; Perform heat source identification based on the second target data to obtain a heat source identification result; Perform smoke recognition based on the current smoke concentration and the second threshold value to obtain a smoke recognition result; When the heat source identification result satisfies a first preset condition and the smoke identification result satisfies a second preset condition, determining that the modified alarm result is that there is a fire, the first preset condition being that there is at least one heat source, and the second preset condition being that the current smoke concentration is greater than or equal to a second threshold; When the heat source identification result does not satisfy the first preset condition and / or the smoke identification result does not satisfy the second preset condition, the modified alarm result is determined to be that there is no fire.

6. The method according to claim 2, characterized in that In the case where any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result, including: When the second target alarm result is that there is a fire, determining the second target alarm result as the result to be corrected; Determine a corresponding confidence score based on the first target alarm result to obtain a first confidence score, and determine a corresponding confidence score based on the third target alarm result to obtain a second confidence score; Performing a weighted calculation based on the first confidence score and the second confidence score to obtain a target confidence score; When the target confidence score is less than or equal to a third threshold, determining that the modified alarm result is that there is a fire; When the target confidence score is greater than a third threshold, the modified alarm result is determined to be that no fire exists.

7. The method according to claim 2, characterized in that In the case where any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, the target alarm result is determined as a result to be corrected, and the result to be corrected is corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result, including: When the third target alarm result is that there is a fire, determining the third target alarm result as the result to be corrected; Record the moment when the current smoke density in the third target data is equal to the first threshold to obtain a first moment; Acquire the second target data corresponding to the first moment and the second moment, obtain a first target image and a second target image, calculate the average of the gray value difference corresponding to the first target image and the second target image, and obtain a target value, wherein the second moment is separated from the first moment by a preset time length and the second moment is after the first moment; Acquire the first target data corresponding to the first moment and the second moment to obtain a third target image and a fourth target image, input the current smoke density, the third target image and the fourth target image into a third target model to obtain a fifth target image, input the fifth target image into the first target model to obtain an updated alarm result of the first target, and the third target model is used to remove noise generated by smoke in the third target image according to the current smoke density and the fourth target image; When the target value satisfies a third preset condition and the updated first target alarm result satisfies a fourth preset condition, determining that the modified alarm result is that there is a fire, the third preset condition is that the target value is greater than a fourth threshold, and the fourth preset condition is that the updated first target alarm is that there is a fire; When the target value does not satisfy the third preset condition and / or the updated first target alarm result does not satisfy the fourth preset condition, the revised alarm result is determined to be that there is no fire.

8. A multi-data fusion substation fire alarm device, characterized in that: The device comprises: an acquisition unit, configured to acquire first target data, second target data, and third target data, wherein the first target data is image data acquired by a visible light camera, the second target data is image data acquired by an infrared camera, and the third target data is concentration data acquired by a smoke sensor; an identification unit, configured to perform fire warning identification based on the first target data, the second target data, and the third target data, respectively, to obtain a plurality of target alarm results, wherein the target alarm results include a first target alarm result corresponding to the first target data, a second target alarm result corresponding to the second target data, and a third target alarm result corresponding to the third target data; a correction unit, configured to, when any one of the target alarm results indicates that a fire exists and the other target alarm results indicate that a fire does not exist, determine the target alarm result as a result to be corrected, and correct the result to be corrected according to the target alarm results other than the result to be corrected to obtain a corrected alarm result; The alarm unit issues a fire alarm at least according to the corrected alarm result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A fire monitoring system, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of claims 1 to 7.

Images