Method, device, equipment and storage medium for determining fire hazard level

Abstract

The present disclosure proposes a method, device, equipment and storage medium for determining the fire danger level, which relates to the field of artificial intelligence technology. The specific scheme is: obtaining video data of the fire scene; determining the number of people, flame color, flame trend and scene type at the fire scene based on the video data; determining the fire danger level based on the number of people, flame color, flame trend and scene type. Thus, static video summaries, target detection and scene recognition can be performed on the fire scene based on computer vision methods to determine effective evaluation factors such as the number of people, flame color, flame trend and scene type at the fire scene, and then determine the danger level of the fire scene based on the characteristics of each danger assessment factor, which can accurately and real-time grade the danger level of the fire scene, which is very beneficial to the actual management of decision makers.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to a method, device, computer equipment and storage medium for determining a fire hazard level. Background Art

[0002] When a fire occurs, decision makers often need to make reasonable decisions based on the degree of danger of the fire. The scene type, fire scale, fire location, type of nearby combustibles and number of people are all important factors affecting the degree of danger of the fire. Whether the degree of danger of the fire can be correctly judged in a timely manner often becomes an important factor in whether decision makers can make reasonable decisions.

[0003] In the related art, temperature information and smoke concentration information of the fire scene can be obtained based on temperature and smoke detection technology to alarm the fire. However, this method takes into account too little effective information of the fire scene and therefore cannot accurately determine the degree of danger of the fire scene. Summary of the invention

[0004] The present disclosure aims to solve one of the technical problems in the related art at least to some extent.

[0005] The present disclosure provides a method, device, system and storage medium for determining a fire hazard level.

[0006] According to a first aspect of the present disclosure, a method for determining a fire hazard level is provided, comprising:

[0007] Acquire video data of the fire scene;

[0008] Determine the number of people at the fire scene, flame color, flame trend and scene type based on the video data;

[0009] The danger level of the fire is determined based on the number of personnel, flame color, flame trend and scene type.

[0010] According to a second aspect of the present disclosure, there is provided a device for determining a fire risk level, comprising:

[0011] An acquisition module, used to acquire video data of the fire scene;

[0012] A first determination module is used to determine the number of people, flame color, flame trend and scene type at the fire scene according to the video data;

[0013] The second determination module is used to determine the danger level of the fire according to the number of personnel, flame color, flame trend and scene type.

[0014] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform any method according to the first aspect.

[0018] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any one of the methods of the first aspect.

[0019] The fifth aspect of the present disclosure provides a computer program product. When an instruction processor in the computer program product executes, the method provided in the first aspect of the present disclosure is executed.

[0020] The method, device and apparatus for determining the fire hazard level provided by the present disclosure have at least the following beneficial effects:

[0021] In the disclosed embodiment, the video data of the fire scene is first obtained, and then the number of people, flame color, flame trend and scene type at the fire scene are determined based on the video data, and then the danger level of the fire is determined based on the number of people, flame color, flame trend and scene type. Thus, static video summaries, target detection and scene recognition can be performed on the fire scene based on computer vision methods to determine effective evaluation factors such as the number of people, flame color, flame trend and scene type at the fire scene, and then the danger level of the fire scene can be determined based on the characteristics of each danger assessment factor, so that the danger level of the fire scene can be accurately and real-time graded, which is very beneficial to the actual management of decision makers.

[0022] Additional aspects and advantages of the present disclosure will be given in part in the following description and in part will be obvious from the following description or learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The above and / or additional aspects and advantages of the present disclosure will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0024] Figure 1 is a flow chart of a method for determining a fire hazard level proposed in an embodiment of the present disclosure;

[0025] Figure 2 is a flow chart of a method for determining a fire hazard level proposed in another embodiment of the present disclosure;

[0026] Figure 3 A structural block diagram of a device for determining a fire hazard level provided by the present disclosure;

[0027] Figure 4 The block diagram is a block diagram of an electronic device for implementing the method for determining the fire risk level of the present disclosure. DETAILED DESCRIPTION

[0028] Embodiments of the present disclosure are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present disclosure, and are not to be construed as limitations of the present disclosure. On the contrary, the embodiments of the present disclosure include all changes, modifications, and equivalents that fall within the spirit and connotation of the appended claims.

[0029] Among them, it can be explained that the executor of the method for determining the fire hazard level of this embodiment is a device for determining the fire hazard level, which can be implemented by software and / or hardware. The device can be configured in a computer device, and the computer device can include but is not limited to a terminal, a server, etc. The method for determining the fire hazard level proposed in the present disclosure will be explained below using the device for determining the fire hazard level as the executor, and no limitation is made here.

[0030] Figure 1 It is a flowchart of a method for determining a fire hazard level proposed in an embodiment of the present disclosure.

[0031] like Figure 1 As shown, the method for determining the fire hazard level includes:

[0032] S101, obtaining video data of the fire scene.

[0033] It should be noted that when a fire occurs, the videos and images of the fire scene record a lot of effective on-site information, such as scene type, fire scale, fire location, type of combustibles and number of people, etc. This information is an important factor affecting the decision maker's initial strategy.

[0034] In the present disclosure, video data of a fire may be obtained through a camera device, which may be a video stream, and then the images in the video may be detected and analyzed to extract key effective information, thereby analyzing the degree of danger of the fire.

[0035] S102, determining the number of people at the fire scene, flame color, flame trend and scene type based on the video data.

[0036] Specifically, after acquiring the video data, in the present disclosure, key frames can be extracted from the video data. For example, representative frames of the beginning or end of the scene transition can be predicted from the video frame sequence, so that the device can analyze these representative frames as key frames.

[0037] Specifically, the Long Short-Term Memory (LSTM) model can be used to discover the relationship between the previous and next image samples in the video frame sequence, and then mine the representative frames of the images in the samples as key frames.

[0038] The number of key frames may vary according to different settings, and the present disclosure does not limit the number of key frames.

[0039] It should be noted that since the key frame is a clear fire start picture containing the fire scene, in the present disclosure, the device can parse each key frame to determine the number of people, flame color, flame trend and scene type at the fire scene corresponding to each key frame.

[0040] The flame color may be colorless, white, gray, or black, and is not limited thereto.

[0041] Among them, scene types can be divided into commercial areas, offices, residences, venue areas, street areas, outdoor activity areas and natural environments, etc., which are not limited here.

[0042] Optionally, in the present disclosure, key frames may be detected by a pre-trained neural network model to determine the number of personnel, flame color, scene type, and diagonal length of the flame detection frame corresponding to each key frame.

[0043] For example, the Faster R-CNN algorithm can be used to train images of commercial areas, offices, residences, venues, streets, outdoor activity areas, and natural environments in any scene data set using Resnet-50 as the basic neural network, and then the trained neural network can be used as the scene classification neural network model of the present invention.

[0044] By inputting key frames into the classification neural network model, the scene type label corresponding to each key frame can be determined, such as commercial area, office room and residence, which is not limited here.

[0045] Optionally, a pre-trained target recognition model may be used to detect key frames to output time information, the number of personnel, and the diagonal length of the flame detection frame corresponding to each key frame.

[0046] For example, smoke, flames and people can be detected by the yolov4 algorithm. In the present disclosure, any flame and smoke scene data set can be used as the basis for pre-training. By using a large number of annotated smoke and flame pictures to train the basic neural network model, a target recognition model can be obtained.

[0047] Optionally, the key frames may be input into a target recognition model to output time information, the number of personnel, and the diagonal length of the flame detection frame corresponding to each key frame.

[0048] Specifically, after determining the diagonal length of the flame detection frame corresponding to each key frame, the flame trend corresponding to each key frame may be determined, such as initial stage combustion, development stage combustion, full combustion, and descending stage combustion.

[0049] Specifically, the corresponding adjacent reference key frames, such as the first two frames, can be determined based on each key frame. Based on the diagonal length of the flame detection frame of the current key frame and the adjacent reference key frames, that is, the size and time information of the flame, the device can determine the flame trend corresponding to each key frame.

[0050] S103, determining the danger level of the fire according to the number of people, flame color, flame trend and scene type.

[0051] It is understandable that in order to solve the problem of assessing the danger level of a fire scene, according to the public safety triangle theory, emergencies and carriers jointly determine the danger level of an accident. Fire is an emergency, and the location and people at the location as carriers jointly affect the danger level of the fire. The fire is determined by the flame trend and smoke, and the smoke can be judged based on the color of the smoke at the scene. Therefore, the danger level of a fire can be jointly determined by the color of the smoke, the number of people, the flame trend, and the scene type.

[0052] Specifically, the corresponding danger coefficients can be determined for the number of people, flame color, flame trend and scene type, and then the final danger assessment value can be determined according to the weight corresponding to each indicator. Optionally, the danger level can be reclassified to determine the final danger level of the fire scene, such as general danger, greater danger, major danger, and extremely major danger, which are not limited here.

[0053] In the disclosed embodiment, the video data of the fire scene is first obtained, and then the number of people, flame color, flame trend and scene type at the fire scene are determined based on the video data, and then the danger level of the fire is determined based on the number of people, flame color, flame trend and scene type. Thus, static video summaries, target detection and scene recognition can be performed on the fire scene based on computer vision methods to determine effective evaluation factors such as the number of people, flame color, flame trend and scene type at the fire scene, and then the danger level of the fire scene can be determined based on the characteristics of each danger assessment factor, so that the danger level of the fire scene can be accurately and real-time graded, which is very beneficial to the actual management of decision makers.

[0054] Figure 2 It is a flowchart of a method for determining a fire hazard level proposed in an embodiment of the present disclosure.

[0055] like Figure 2 As shown, the method for determining the fire hazard level includes:

[0056] S201, obtaining video data of the fire scene.

[0057] It should be noted that the specific implementation of step S201 can refer to the above embodiment and will not be described in detail here.

[0058] S202, determining key frames in the video data according to the definition of each frame of the video data, the content contained therein, and the time interval between each frame of the video data.

[0059] It should be noted that the video data may correspond to a video frame sequence, that is, video stream information. In the present disclosure, video static summary technology may be used to filter and screen redundant and blurred frames in the video to obtain valid frames of the video information, that is, key frames.

[0060] Specifically, the key frames in the video frame sequence may be determined according to the clarity and content of each frame, that is, the amount of information, and the time interval between each frame.

[0061] Among them, the key frame, that is, the representative picture, is more conducive to reflecting various factors of the fire scene.

[0062] Preferably, the LSTM algorithm can be used to perform static video summarization on the video stream. In addition to clear, information-rich and less redundant frames, the time information corresponding to the key frames and the number of key frames need to be recorded. In addition, when extracting key frames, a certain threshold should be set for the time interval between key frames, so as to prevent excessive information reduction and affect judgment.

[0063] S203, analyzing the key frames to determine the number of people at the fire scene, flame color, flame trend and scene type.

[0064] Optionally, in the present disclosure, key frames may be detected by a pre-trained neural network model to determine the number of personnel, flame color, scene type, and diagonal length of the flame detection frame corresponding to each key frame.

[0065] For example, the Faster R-CNN algorithm can be used to train images of commercial areas, offices, residences, venues, streets, outdoor activity areas, and natural environments in any scene data set using Resnet-50 as the basic neural network, and then the trained neural network can be used as the scene classification neural network model of the present invention.

[0066] By inputting key frames into the classification neural network model, the scene type label corresponding to each key frame can be determined, such as commercial area, office room and residence, which is not limited here.

[0067] Optionally, a pre-trained target recognition model may be used to detect key frames to output time information, the number of personnel, and the diagonal length of the flame detection frame corresponding to each key frame.

[0068] For example, smoke, flames and people can be detected by the YOLOv4 algorithm. In the present disclosure, any flame and smoke scene data set can be used as the basis for pre-training. By using a large number of annotated smoke and flame pictures to train the basic neural network model, a target recognition model can be obtained.

[0069] Optionally, the key frames may be input into a target recognition model to output time information, the number of personnel, and the diagonal length of the flame detection frame corresponding to each key frame.

[0070] Preferably, each key frame can be parsed to determine the diagonal length corresponding to the flame detection frame contained in each key frame, and then the flame trend can be determined based on the time interval between each key frame and the diagonal length corresponding to the flame detection frame contained in each key frame.

[0071] The flame trend can be determined based on the change in the size of the flame detection frame in the image after the fire starts. For example, the diagonal size of the flame detection frame of the first two key frames of the current frame can be extracted as a reference. For example, if the time corresponding to the current key frame is Ty, the diagonal length of the flame detection frame is Dy, and the time corresponding to the first two frames of the current frame is T y-1 , T y-2 , the diagonal length is D y-1 , D y-2 .

[0072] It should be noted that the initial stage corresponds to a concave function with a low slope and a monotonic interval increase. Due to the low slope, the slope below a certain threshold is regarded as the initial stage of flame combustion; the change in the size of the flame frame in the development stage corresponds to a concave function with a monotonic interval increase; the comprehensive combustion stage corresponds to a convex function or a linear function parallel to the time axis; the descending stage corresponds to a monotonic interval descending function.

[0073] If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the following relationship (1), it can be determined that the flame of the current frame is in the initial stage.

[0074] Wherein, δ is the preset slope threshold.

[0075]

[0076] If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the following relationship (2), it can be determined that the flame in the current frame is in the development stage.

[0077]

[0078] If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the following relationship (3), it can be determined that the flame of the current frame is in the full combustion stage.

[0079]

[0080] If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the following relationship (4), it can be determined that the flame of the current frame is in the descending stage.

[0081] D Y <D Y-1 <D Y-2 (4)

[0082] S204, determining a first risk factor according to the flame color.

[0083] It should be noted that the color of smoke depends on the type of combustible material. The color of smoke can help determine the extent of the fire and the degree of danger at the scene.

[0084] The first risk factor may be a factor determined according to the risk of the flame color.

[0085] Among them, white smoke has the lowest temperature and a small fire, and is set to a general danger factor.

[0086] Among them, gray smoke is the most important thing that should not be underestimated. It is very likely to be smoldering or high-temperature combustion, and is set to a higher risk factor.

[0087] Among them, yellow-green smoke may be the burning of toxic chemicals and is set as a significant hazard factor.

[0088] Among them, black smoke has the highest temperature and usually occurs when the fire burns most fiercely. There are also raging flames in the smoke. It is the most dangerous period in a fire and is set as a very serious risk factor.

[0089] It should be noted that different corresponding values ​​can be determined for different risk factors. For example, a general risk factor can correspond to a value in the range of [0, 0.25]. A larger risk factor can correspond to a value in the range of (0.25, 0.5]. A major risk factor can correspond to a value in the range of (0.5, 0.75). An extremely major risk factor can correspond to a value in the range of (0.75, 1].

[0090] It should be noted that the above examples are only for illustrative purposes and are not intended to be limiting in the present disclosure.

[0091] S205: Determine a second risk factor based on the flame trend.

[0092] It should be noted that the stage of the flame corresponds to different degrees of danger. Usually, the danger level is lower in the initial stage of a fire, but one needs to be vigilant against flame flashover. The temperature at the scene is the highest during the full combustion stage, and it is also the most dangerous stage. In the descending stage, the fire will gradually become smaller, the temperature will gradually decrease, and the degree of danger will also decrease.

[0093] Therefore, the risk factor corresponding to the descending stage of the flame trend can be determined as a general risk factor, the risk factor corresponding to the initial stage of the flame trend can be determined as a larger risk factor, the risk factor corresponding to the development stage of the flame trend can be determined as a major risk factor, and the risk factor corresponding to the comprehensive combustion stage of the flame trend can be determined as an extremely major risk factor.

[0094] It should be noted that different corresponding values ​​can be determined for different risk factors. For example, a general risk factor can correspond to a value in the range of [0, 0.25]. A larger risk factor can correspond to a value in the range of (0.25, 0.5]. A major risk factor can correspond to a value in the range of (0.5, 0.75). An extremely major risk factor can correspond to a value in the range of (0.75, 1].

[0095] S206: Determine a third risk factor according to the scene type.

[0096] The third risk factor may be determined according to the risk corresponding to the scene type.

[0097] More specifically, the third risk coefficient can be set according to the randomness of the distribution and types of combustibles, the randomness of fire sources, and human activity conditions under the standards of the disclosed coefficient system, specifically for commercial areas, offices, residences, venue areas, street areas, natural environments, and outdoor activity areas.

[0098] Among them, commercial areas and offices are classified as Category 1, corresponding to the first-level risk factor; residential areas are classified as Category 1, corresponding to the second-level risk factor; venue areas and street areas are classified as Category 1, corresponding to the third-level risk factor; natural environments and outdoor activity areas are classified as Category 1, corresponding to the fourth-level risk factor.

[0099] It should be noted that different corresponding values ​​can be determined for different risk factors. For example, the first-level risk factor can correspond to a value in the range of [0, 0.25], the second-level risk factor can correspond to a value in the range of (0.25, 0.5], the third-level risk factor can correspond to a value in the range of (0.5, 0.75], and the fourth-level risk factor can correspond to a value in the range of (0.75, 1].

[0100] It should be noted that the above examples are only for illustrative purposes and are not intended to be limiting in the present disclosure.

[0101] S207, determining the fourth risk factor according to the number of personnel.

[0102] It should be noted that a large number of people means a higher risk of casualties, so it is necessary to detect the number of people.

[0103] The fourth risk factor may be a risk factor determined according to the number of personnel.

[0104] In the present disclosure, a corresponding risk factor can be set based on the number of people at the fire scene. For example, after the number of people at the fire scene is determined to be P, if P is less than 10, the risk factor can be determined to be a primary risk factor. If P is in [10,50), the risk factor can be determined to be a medium risk factor. If P is in [50,100), the risk factor can be determined to be a high risk factor. If P is greater than or equal to 100, the risk factor can be determined to be an extremely serious risk factor.

[0105] It should be noted that different corresponding values ​​can be determined for different risk factors. For example, the primary risk factor can correspond to a value in the range of [0, 0.25], the intermediate risk factor can correspond to a value in the range of (0.25, 0.5], the advanced risk factor can correspond to a value in the range of (0.5, 0.75], and the extremely serious risk factor can correspond to a value in the range of (0.75, 1].

[0106] It should be noted that the above examples are only for illustrative purposes and are not intended to be limiting in the present disclosure.

[0107] S208: Determine the risk level according to the first risk factor, the second risk factor, the third risk factor, and the fourth risk factor.

[0108] It is understandable that the process of fire from ignition to peak and then to fire slowing down, as well as its degree of destruction and danger are closely related to the fire scene. Different building structures, fire load conditions, fire spread possibility, heat release rate and other factors in the scene have a great impact on the on-site risk assessment. Compared with natural scenes, artificial environments usually have richer types of combustibles, more diverse combustion characteristics and smaller spaces. Based on this, urban fire scenes are divided into commercial areas, offices, residences, venues, streets, natural environments and outdoor activity areas according to risk hazards. The scene recognition algorithm is used to classify key frames into the above categories and correspond to their scene risk coefficients as the basis for risk assessment.

[0109] Optionally, the risk level may be determined based on a preset reference weight, and a first risk factor, a second risk factor, a third risk factor, and a fourth risk factor.

[0110] It should be noted that for each risk factor, a corresponding reference weight can be preset. For example, for the first risk factor, a reference weight corresponding to the flame color can be set, for the second risk factor, a reference weight corresponding to the flame trend can be set, for the third risk factor, a reference weight corresponding to the number of people can be set, and for the fourth risk factor, a reference weight corresponding to the scene type can be set.

[0111] For example, set the first risk coefficient to A, its corresponding reference weight to a1, the second risk coefficient to B, its corresponding reference weight to a2, the third risk coefficient to C, its corresponding reference weight to a3, the fourth risk coefficient to D, its corresponding reference weight to a4, then the risk assessment value can be determined as S = A*a1+B*a2+C*a3+D*a4.

[0112] After the risk assessment value is determined, the corresponding risk level can be determined according to the range of S, such as general risk, greater risk, major risk, and extremely major risk, which are not limited here.

[0113] In the disclosed embodiment, the video data of the fire scene is first obtained, and then the key frames in the video data are determined according to the clarity of each frame of the video data, the content contained and the time interval between each frame of the video data, and the key frames are parsed to determine the number of people, flame color, flame trend and scene type at the fire scene. Then, the first danger factor is determined according to the flame color, the second danger factor is determined according to the flame trend, the third danger factor is determined according to the scene type, the fourth danger factor is determined according to the number of people, and the danger level is determined according to the first danger factor, the second danger factor, the third danger factor and the fourth danger factor. In this way, the valid frames and valid reference information of the video data of the fire scene can be extracted, and the danger level of the fire scene can be graded according to important factors such as the number of people, flame color, flame trend and scene type, so as to help decision makers make correct decisions in a timely manner.

[0114] like Figure 3 As shown, the fire risk level determination device 300 includes: an acquisition module 310 , a first determination module 320 , and a second determination module 330 .

[0115] An acquisition module, used to acquire video data of the fire scene;

[0116] A first determination module is used to determine the number of people, flame color, flame trend and scene type at the fire scene according to the video data;

[0117] The second determination module is used to determine the danger level of the fire according to the number of personnel, flame color, flame trend and scene type.

[0118] Optionally, the first determining module includes:

[0119] A first determining unit, configured to determine a key frame in the video data according to the definition of each frame image in the video data, the content contained therein, and the time interval between each frame image;

[0120] The parsing unit is used to parse the key frame to determine the number of people at the fire scene, flame color, flame trend and scene type.

[0121] Optionally, the parsing unit is specifically used to:

[0122] Parsing each of the key frames to determine the diagonal length corresponding to the flame detection frame contained in each of the key frames;

[0123] The flame trend is determined according to the time interval between each of the key frames and the diagonal length corresponding to the flame detection frame included in each of the key frames.

[0124] Optionally, the second determining module includes:

[0125] A second determining unit, configured to determine a first risk factor according to the flame color;

[0126] A third determining unit, configured to determine a second risk factor according to the flame trend;

[0127] a fourth determining unit, configured to determine a third risk factor according to the scene type;

[0128] a fifth determining unit, configured to determine a fourth risk factor according to the number of personnel;

[0129] The sixth determining unit is used to determine a risk level according to the first risk factor, the second risk factor, the third risk factor and the fourth risk factor.

[0130] Optionally, the sixth determining unit is specifically configured to:

[0131] The risk level is determined according to a preset reference weight, and the first risk factor, the second risk factor, the third risk factor and the fourth risk factor.

[0132] In the disclosed embodiment, the video data of the fire scene is first obtained, and then the number of people, flame color, flame trend and scene type at the fire scene are determined based on the video data, and then the danger level of the fire is determined based on the number of people, flame color, flame trend and scene type. Thus, static video summaries, target detection and scene recognition can be performed on the fire scene based on computer vision methods to determine effective evaluation factors such as the number of people, flame color, flame trend and scene type at the fire scene, and then the danger level of the fire scene can be determined based on the characteristics of each danger assessment factor, so that the danger level of the fire scene can be accurately and real-time graded, which is very beneficial to the actual management of decision makers.

[0133] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0134] Figure 4A schematic block diagram of an example electronic device 400 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0135] like Figure 4 As shown, the device 400 includes a computing unit 401, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 402 or a computer program loaded from a storage unit 408 into a random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the device 400 can also be stored. The computing unit 401, the ROM 402, and the RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0136] A number of components in the device 400 are connected to the I / O interface 405, including: an input unit 406, such as a keyboard, a mouse, etc.; an output unit 407, such as various types of displays, speakers, etc.; a storage unit 408, such as a disk, an optical disk, etc.; and a communication unit 409, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 409 allows the device 400 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0137] The computing unit 401 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 401 performs the various methods and processes described above, such as the method for determining the fire hazard level. For example, in some embodiments, the method for determining the fire hazard level may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 400 via the ROM 402 and / or the communication unit 409. When the computer program is loaded into the RAM 403 and executed by the computing unit 401, one or more steps of the method for determining the fire hazard level described above may be performed. Alternatively, in other embodiments, the computing unit 401 may be configured to perform the method for determining the fire hazard level in any other appropriate manner (e.g., by means of firmware).

[0138] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs, which may be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general programmable processor, which may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0139] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0140] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0141] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0142] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.

[0143] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0144] In the disclosed embodiment, the video data of the fire scene is first obtained, and then the number of people, flame color, flame trend and scene type at the fire scene are determined based on the video data, and then the danger level of the fire is determined based on the number of people, flame color, flame trend and scene type. Thus, static video summaries, target detection and scene recognition can be performed on the fire scene based on computer vision methods to determine effective evaluation factors such as the number of people, flame color, flame trend and scene type at the fire scene, and then the danger level of the fire scene can be determined based on the characteristics of each danger assessment factor, so that the danger level of the fire scene can be accurately and real-time graded, which is very beneficial to the actual management of decision makers.

[0145] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0146] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for determining a fire hazard level, characterized in that: include: Acquire video data of the fire scene; Determine the key frames in the video data through the long short-term memory model, and parse the key frames through the pre-trained neural network model to determine the number of people at the fire scene, the flame color, the flame trend and the scene type; wherein the flame trend is determined according to the time interval between each of the key frames and the diagonal length corresponding to the flame detection frame contained in each of the key frames, and the flame trend includes an initial stage, a development stage, a full combustion stage and a decline stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (1), it is determined that the flame of the current frame is in the initial stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (2), it is determined that the flame of the current frame is in the development stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (3), it is determined that the flame of the current frame is in the full combustion stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (4), it is determined that the flame of the current frame is in the descending stage; D Y <D Y-1 <D Y-2 (4) Among them, the time corresponding to the current key frame is Ty, the diagonal length of the flame detection frame is Dy, and the time corresponding to the first two frames of the current frame is Ty -1 、Ty -2 , the diagonal length is Dy -1 、Dy -2 , δ is the preset slope threshold; The danger level of the fire is determined based on the number of personnel, flame color, flame trend and scene type.

2. The method according to claim 1, characterized in that Determine the danger level of the fire based on the number of personnel, flame color, flame trend and scene type, including: Determining a first risk factor according to the flame color; Determining a second risk factor according to the flame trend; Determining a third risk factor according to the scenario type; Determine a fourth risk factor based on the number of personnel; A risk level is determined according to the first risk factor, the second risk factor, the third risk factor, and the fourth risk factor.

3. The method according to claim 2, characterized in that Determining the risk level according to the first risk factor, the second risk factor, the third risk factor, and the fourth risk factor includes: The risk level is determined according to a preset reference weight, and the first risk factor, the second risk factor, the third risk factor and the fourth risk factor.

4. A device for determining a fire hazard level, characterized in that: include: An acquisition module, used to acquire video data of the fire scene; The first determination module is used to determine the key frames in the video data through the long short-term memory model, and parse the key frames through the pre-trained neural network model to determine the number of people at the fire scene, the flame color, the flame trend and the scene type; wherein the flame trend is determined according to the time interval between each of the key frames and the diagonal length corresponding to the flame detection frame contained in each of the key frames, and the flame trend includes an initial stage, a development stage, a full combustion stage and a decline stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (1), it is determined that the flame of the current frame is in the initial stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (2), it is determined that the flame of the current frame is in the development stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (3), it is determined that the flame of the current frame is in the full combustion stage; If the diagonal lengths of the flame detection frames of the current frame and the first two key frames and the time corresponding to the key frames satisfy the relationship (4), it is determined that the flame of the current frame is in the descending stage; D Y <D Y-1 <D Y-2 (4) Among them, the time corresponding to the current key frame is Ty, the diagonal length of the flame detection frame is Dy, and the time corresponding to the first two frames of the current frame is Ty -1 、Ty -2 , the diagonal length is Dy -1 、Dy -2 , δ is the preset slope threshold; The second determination module is used to determine the danger level of the fire according to the number of personnel, flame color, flame trend and scene type.

5. The device according to claim 4, characterized in that The second determining module comprises: A second determining unit, configured to determine a first risk factor according to the flame color; A third determining unit, configured to determine a second risk factor according to the flame trend; a fourth determining unit, configured to determine a third risk factor according to the scene type; a fifth determining unit, configured to determine a fourth risk factor according to the number of personnel; The sixth determining unit is used to determine a risk level according to the first risk factor, the second risk factor, the third risk factor and the fourth risk factor.

6. The device according to claim 5, characterized in that The sixth determining unit is specifically configured to: The risk level is determined according to a preset reference weight, and the first risk factor, the second risk factor, the third risk factor and the fourth risk factor.

7. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can implement the method for determining the fire risk level according to any one of claims 1 to 3.

8. A computer-readable storage medium, when the instructions in the computer-readable storage medium are executed by a processor of a server, the server is enabled to execute the method for determining the fire risk level according to any one of claims 1 to 3.

9. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method for determining the fire risk level according to any one of claims 1 to 3 is implemented.

Images