Safety warning methods and related products for hot work

Abstract

The present application discloses a safety warning method for hot work and related products, which includes: obtaining target basic data; arranging m control balls in a target operation area according to the target basic data and a preset arrangement method; collecting target monitoring data of the target operation area; analyzing the target monitoring video to determine the start time of the flame in the flame area; intercepting multiple video frames of the flame area in the target monitoring video; inputting the multiple video frames into a target hot work recognition model to obtain at least one recognition result and at least one confidence level; when the target recognition result includes the presence of hot work, determining a first risk probability based on the flame area and the target confidence level; determining a second risk probability corresponding to the flame area based on the target sensor data; and controlling an alarm module to execute a target alarm operation when the first risk probability and the second risk probability meet preset conditions. The embodiments of the present application improve the safety of hot work.

Description

Technical Field

[0001] The present application relates to the field of fire safety technology, and in particular to a safety warning method for hot work operations and related products. Background Art

[0002] In the power industry, hot work is a common process. Due to the involvement of high temperature, open flames and other factors, there is a high risk of fire and explosion.

[0003] Traditional hot work risk analysis relies primarily on manual inspections and empirical judgment, which has certain limitations and subjectivity, leading to lower safety levels in hot work. Therefore, improving the safety of hot work has become an urgent issue that needs to be addressed. Summary of the Invention

[0004] The embodiments of the present application provide a safety warning method for hot work and related products, which improve the safety of hot work.

[0005] In a first aspect, an embodiment of the present application provides a safety warning method for hot work, which is applied to a control module in a hot work safety detection system. The hot work safety detection system further includes: m control balls, a cloud data analysis platform, and an alarm module, where m is a positive integer. The method includes:

[0006] Obtain target basic data of target operation area;

[0007] Arranging the m control balls within the target operation area according to the target basic data and a preset arrangement method;

[0008] During a first preset time period, the m control balls are used to collect data in the target operation area to obtain target monitoring data; the target monitoring data includes target monitoring video and target sensor data;

[0009] Analyzing the target surveillance video using the cloud data analysis platform based on a preset flame recognition algorithm to determine whether there is a flame area in the target surveillance video;

[0010] When the flame area exists in the target monitoring video, determining the starting time of the flame in the flame area;

[0011] Intercepting video frames of the flame area in the target monitoring video within a second preset time period to obtain multiple video frames; the starting time is the middle time of the second preset time period;

[0012] Inputting the plurality of video frames into a target hot work identification model in the cloud data analysis platform to obtain at least one identification result and at least one confidence level; each identification result includes whether a hot work operation exists or does not exist; and the identification results and the confidence levels correspond one to one;

[0013] When the target recognition result in the at least one recognition result includes the presence of hot work, determining a first risk probability according to the flame area and the target confidence corresponding to the target recognition result; the target recognition result is any recognition result in the at least one recognition result;

[0014] Determining a second risk probability corresponding to the flame area based on the target sensor data through the cloud data analysis platform;

[0015] When the first risk probability and the second risk probability meet a preset condition, the control alarm module performs a target alarm operation according to the target recognition result, the flame area, the first risk probability and the second risk probability.

[0016] In a second aspect, an embodiment of the present application provides a safety warning device for hot work, which is applied to a control module in a hot work safety detection system. The hot work safety detection system further includes: m control balls, a cloud data analysis platform, and an alarm module, where m is a positive integer; the device includes: an acquisition unit, a control unit, and an early warning unit, wherein:

[0017] The acquisition unit is used to acquire target basic data of the target operation area;

[0018] The control unit is configured to arrange the m control balls within the target operation area according to the target basic data and a preset arrangement method;

[0019] The acquisition unit is further configured to collect data in the target operation area through the m control balls within a first preset time period to obtain target monitoring data; the target monitoring data includes target monitoring video and target sensor data;

[0020] The control unit is further configured to analyze the target monitoring video based on a preset flame recognition algorithm through the cloud data analysis platform to determine whether there is a flame area in the target monitoring video; when the flame area exists in the target monitoring video, determine the starting time of the flame in the flame area; intercept video frames of the flame area in the target monitoring video within a second preset time period to obtain multiple video frames; the starting time is the middle time of the second preset time period; input the multiple video frames into the target hot work recognition model in the cloud data analysis platform to obtain at least one recognition result and at least one confidence level; each recognition result includes the presence or absence of hot work; the recognition result and the confidence level correspond one to one; when the target recognition result in the at least one recognition result includes the presence of hot work, determine a first risk probability according to the flame area and the target confidence level corresponding to the target recognition result; the target recognition result is any recognition result among the at least one recognition result; determine a second risk probability corresponding to the flame area according to the target sensor data through the cloud data analysis platform;

[0021] The early warning unit is configured to control the alarm module to perform a target alarm operation according to the target recognition result, the flame area, the first risk probability, and the second risk probability when the first risk probability and the second risk probability meet preset conditions.

[0022] In a third aspect, the present application provides an electronic device comprising: a processor and a memory, wherein the memory is used to store one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for executing the steps in the first aspect of the present application.

[0023] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute some or all of the steps described in the first aspect of the present application.

[0024] In a fifth aspect, the present application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program, wherein the computer program is operable to cause a computer to perform some or all of the steps described in the first aspect of the present application. The computer program product may be a software installation package.

[0025] The implementation of this application has the following beneficial effects:

[0026] It can be seen that the safety warning method and related products for hot work operations described in this application achieve all-round monitoring of the operation area and reduce monitoring blind spots by arranging multiple control balls in the target operation area. Whether in indoor or outdoor operation areas, potential hot work operation risks can be discovered in a timely manner. Moreover, through the cloud data analysis platform combined with the preset flame recognition algorithm and hot work recognition model, the flame area and hot work operation can be detected quickly and accurately. Once an abnormal situation is found, an alarm is triggered immediately to gain valuable time for taking emergency measures, thereby effectively improving the safety of hot work operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0028] Figure 1 This is a structural diagram of a hot work safety detection system provided in an embodiment of the present application;

[0029] Figure 2 This is a flow chart of a safety warning method for hot work provided in an embodiment of the present application;

[0030] Figure 3 This is a block diagram of the functional units of a safety warning device for hot work provided by an embodiment of the present application;

[0031] Figure 4 This is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] The terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0034] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0035] The electronic devices described in the embodiments of the present application may include smart phones (such as Android phones, iOS phones, Windows Phone phones, etc.), tablet computers, PDAs, laptops, video matrices, monitoring platforms, mobile Internet devices (MIDs) or wearable devices, etc. The above are only examples and not exhaustive, including but not limited to the above devices. Of course, the above electronic devices can also be servers, for example, cloud servers.

[0036] The following is an explanation of some professional terms involved in this application:

[0037] Hot work: refers to non-routine work that may produce flames, sparks, or hot surfaces in a fire-free zone outside of process facilities that directly or indirectly produce open flames, including electric welding, gas welding (cutting), blowtorches, electric drills, grinding wheels, sandblasting machines, etc.

[0038] A surveillance camera is a portable monitoring device equipped with high-definition video capture, wireless transmission, and remote control. It can also be equipped with various sensors. Spherical or hemispherical in shape, it is relatively compact and easy to carry and install. It typically comes with an adjustable stand, allowing for flexible placement in various locations, such as on a tabletop, tripod, or wall. At power construction sites and substations, surveillance cameras can monitor the progress of construction, equipment operation, and safety. They can promptly detect operational violations and equipment failures, ensuring the safety and stability of power production.

[0039] See also Figure 1 , Figure 1 This is a structural diagram of a hot work safety detection system provided in an embodiment of the present application. It can be seen that the hot work safety detection system includes: m control balls, a control module, a cloud data analysis platform, and an alarm module, wherein:

[0040] Each of the m control balls mentioned above is equipped with a high-definition camera that can capture video images of the hot work site in real time. By arranging control balls in different locations, the entire work area can be covered to ensure that there are no blind spots in monitoring. For example, in the hot work area of ​​a factory, multiple control balls can be installed in different locations to comprehensively record the operating process of the operators, the surrounding environment, and possible fire sources. The m control balls mentioned above can also be integrated with sensors, such as temperature sensors, smoke sensors, etc., to collect data such as temperature and smoke concentration at the site, providing more basis for judging whether there is a fire risk. The m control balls can transmit the collected video images and sensor data to the cloud data analysis platform and / or control module through wireless communication technology (such as 4G / 5G, Wi-Fi, etc.), realizing the function of remote monitoring. Managers can understand the operation status without having to visit the site in person and discover potential safety hazards in a timely manner.

[0041] The control module is responsible for coordinating the operations of various modules within the hot work safety detection system. It ensures that the hot work dome is properly collecting data and transmitting it to the cloud-based data analysis platform. It also receives commands from the cloud-based data analysis platform to control other modules. For example, the control module can receive video and sensor data from each hot work dome, convert the video into a bitstream, and transmit it to the cloud-based data analysis platform. It can also receive control commands from the cloud-based data analysis platform and support sending these commands to individual hot work domes. These commands can include operations such as zooming in, zooming out, and performing 360-degree rotations on the dome.

[0042] The cloud-based data analysis platform processes and analyzes video and sensor data transmitted by surveillance cameras. For example, it can determine whether the on-site temperature is too high and outside the safe range based on temperature sensor data; and determine smoke concentration and fire hazards based on smoke sensor data. Combining video analysis results with sensor data allows for a comprehensive assessment of the safety risk of hot work.

[0043] The alarm module is used to detect safety risks during hot work and, based on the control module's instructions, initiate an alarm. Alarms can include audible and visual alarms, text messages, emails, and mobile apps to draw the attention of on-site workers and managers.

[0044] See also Figure 2 , Figure 2 This is a flow chart of a safety warning method for hot work provided in an embodiment of the present application. The method is applied to a control module in a hot work safety detection system. The hot work safety detection system further includes: m control balls, a cloud data analysis platform, and an alarm module, where m is a positive integer. The method includes but is not limited to the following steps:

[0045] S201: Obtain target basic data of the target operation area.

[0046] In an embodiment of the present application, the target operation area may include one of the following: a power plant, a substation, a machinery manufacturing workshop, etc., which are not limited here; the target basic data may include at least one of the following: regional map data, regional item data, lighting condition data, historical data of hot work, etc., which are not limited here.

[0047] In a specific embodiment, the target basic data of the target operation area can be obtained through the control module. Specifically, the target basic data may include regional map data and regional item data. The control module can access the target database of the target operation area and obtain the target basic data from the target database. Alternatively, the target basic data can be manually input into the control module by the staff of the target operation area.

[0048] S202: Arrange the m control balls in the target operation area according to the target basic data and a preset arrangement method.

[0049] In the embodiment of the present application, the preset arrangement method can be preset or defaulted in advance.

[0050] In a specific embodiment, m positions can be selected in the target operation area according to the target basic data and the preset arrangement method, and then m control balls can be arranged at the m positions, with each position corresponding to a control ball.

[0051] It should be explained that the m control balls can be of the same type, and all parameters of the m control balls can be the same.

[0052] Optionally, step S202, arranging the m control balls in the target operation area according to the target basic data and a preset arrangement method, may include the following steps:

[0053] S21, obtaining target device specification parameters of the m control balls;

[0054] S22. Determine a first monitoring coverage area corresponding to the specification parameters of the target device;

[0055] S23, obtaining the target light intensity in the target operation area;

[0056] S24, determining a target interference factor corresponding to the target light intensity;

[0057] S25. Adjust the first monitoring coverage area according to the target interference factor to obtain a second monitoring coverage area;

[0058] S26. Divide the target operation area into n first areas according to the second monitoring coverage area, wherein an area of ​​each of the n first areas is smaller than the second monitoring coverage area, and n is a positive integer less than or equal to m.

[0059] S27, determining a regional function of each of the n first regions according to the target basic data, to obtain n regional functions;

[0060] S28. Determine the area functions corresponding to the hot work operation among the n area functions to obtain i area functions; i is a positive integer less than or equal to n;

[0061] S29. Determine first areas corresponding to the i area functions in the n first areas, to obtain i first areas;

[0062] S210, dividing each of the i first areas into j second areas according to a preset area, where j is a positive integer greater than or equal to i, and the sum of j and ni is equal to m;

[0063] S211. Set a control ball in each of the j second areas and the ni first areas according to the preset arrangement method; the ni first areas are the first areas in the n first areas except the i first areas.

[0064] In the embodiment of the present application, the preset area can be preset or defaulted in advance; the target device specification parameters may include at least one of the following: viewing angle parameters, focal length parameters, resolution parameters, etc., which are not limited here.

[0065] In a specific embodiment, target device specification parameters of m control balls are obtained. Specifically, the target device specification parameters may be viewing angle parameters. Configuration software of the m control balls may be obtained. Generally speaking, the control balls are equipped with special configuration software. Various parameters of the control balls may be viewed and adjusted through the configuration software. The target device specification parameters may be viewed and obtained in the configuration software, or the target device specification parameters may be obtained from the user manual or technical documentation of the control ball. Then, a first monitoring coverage area corresponding to the target device specification parameters may be determined. Specifically, a mapping relationship between preset device specification parameters and monitoring coverage areas may be pre-stored, and the first monitoring coverage area corresponding to the target device specification parameters may be determined based on the mapping relationship. Then, the target lighting in the target operating area may be obtained. Intensity. Specifically, a light sensor can be installed at a representative location in the target operation area to obtain the target light intensity through the light sensor. For example, assuming that the target operation area is a factory workshop, light sensors can be installed in different work areas, passages, and near key equipment to fully understand the lighting conditions of the entire operation area and obtain the target light intensity. Then, the target interference factor corresponding to the target light intensity can be determined. Specifically, a preset mapping relationship between light intensity and interference factor can be pre-stored, and the target interference factor corresponding to the target light intensity can be determined based on the mapping relationship. The value range of the target interference factor can be -0.3 to 0.3. Then, the first monitoring coverage area can be adjusted according to the target interference factor. The specific calculation formula is as follows:

[0066] Second monitoring coverage area = first monitoring coverage area × (1 + target interference factor);

[0067] According to the above formula, the second monitoring coverage area can be obtained; then, the target operation area can be divided into regions according to the second monitoring coverage area to obtain n first regions. Specifically, a regular division method can be used to divide the target operation area into multiple regions of equal area to obtain n first regions. For example, assuming that the target operation area is a rectangle, it can be divided into n rectangular regions of equal size, and the area of ​​each region is less than or equal to the second monitoring coverage area; then, the regional function of each of the n first regions can be determined according to the target basic data to obtain n regional functions. Specifically, the target basic data can be regional map data. According to the features on the regional map data, the n first regions can be divided into different types according to their usage functions. Common types include production areas, storage areas, office areas, channel areas, equipment areas, etc., to obtain n regional functions; it should be explained that the same regional function can exist in the n regional functions, which is not limited here.

[0068] Then, the area functions corresponding to hot work operations among the n area functions can be determined to obtain i area functions. Specifically, the n area functions can be screened to determine those area functions that may be related to hot work operations, and these area functions can be marked as potential hot work area functions, that is, i area functions. For example, from multiple area functions of a factory, area functions related to hot work operations such as "welding workshop", "metal processing area", and "equipment maintenance area" can be screened out; then, the first area corresponding to the i area functions can be selected from the n first areas to obtain i first areas; then, the i first areas can be further divided into areas according to a preset area to obtain j second areas. Specifically, the i first areas can be divided into j areas with a preset area by equal area to obtain j areas; finally, a control ball can be set in each of the j second areas and the ni first areas according to a preset layout method until all m control balls are arranged. For example, the preset layout method can be a center layout method, and a control ball is arranged at the center position of each of the j second areas and the ni first areas.

[0069] In this way, by obtaining the target equipment specification parameters of the control ball to determine the first monitoring coverage area, the monitoring range of each control ball under ideal conditions can be clarified, which provides basic data for subsequent area division and control ball layout, making the monitoring layout more scientific and reasonable, and realizing the scientific allocation and efficient use of monitoring resources. On the other hand, through precise monitoring coverage and area division, safety hazards in the process of hot work operations can be discovered in time. The control ball can collect images and data of the target operation area in real time. Once dangerous factors such as flames, smoke, abnormal temperatures, etc. are found, an alarm can be issued immediately to remind relevant personnel to take measures, thereby improving the safety of hot work operations.

[0070] S203. Within a first preset time period, data in the target operation area is collected by the m control balls to obtain target monitoring data; the target monitoring data includes target monitoring video and target sensor data.

[0071] In the embodiment of the present application, the first preset time period may be preset in advance or default.

[0072] In a specific embodiment, within a first preset time period, data in a target operation area may be collected by m control balls to obtain target monitoring data, and the target monitoring data may be transmitted to a control module.

[0073] Optionally, in step S203, each of the m control balls includes a camera module and a sensor module, and collecting data in the target operation area by the m control balls to obtain target monitoring data may include the following steps:

[0074] S31, obtaining a target control ball; the target control ball includes a target camera module and a target sensor module; the target control ball is any one of the m control balls;

[0075] S32, determining the target monitoring area corresponding to the target control ball;

[0076] S33, obtaining initial monitoring parameters of the target control ball;

[0077] S34. Within a third preset time period, the target monitoring area is monitored by the target control ball using initial monitoring parameters to obtain a first monitoring video; the first preset time period includes the third preset time period;

[0078] S35. Determine a first definition corresponding to the first surveillance video;

[0079] S36. When the first definition is less than a preset definition, obtaining a difference between the first definition and the preset definition to obtain a target definition difference;

[0080] S37, determining a target adjustment parameter corresponding to the target clarity difference;

[0081] S38. Adjust the initial monitoring parameters according to the target adjustment parameters to obtain target monitoring parameters;

[0082] S39. During a fourth preset time period, the target monitoring area is monitored by the target control ball using the target monitoring parameters to obtain a second monitoring video; the start time of the fourth preset time period is the end time of the third preset time period; and the first preset time period includes the fourth preset time period.

[0083] S310: Determine the target surveillance video according to the first surveillance video and the second surveillance video;

[0084] S311. Within the first preset time period, the target control ball reads data from the target sensor module at preset time intervals to obtain the target sensor data.

[0085] In an embodiment of the present application, the monitoring parameters may include at least one of the following: resolution, frame rate, bit rate, data transmission rate, etc., which are not limited here; the preset clarity and preset time interval can be preset or defaulted in advance.

[0086] In a specific embodiment, a control ball can be randomly obtained from m control balls and used as the target control ball; then, the target monitoring area corresponding to the target control ball can be determined. Specifically, the target device unique identifier of the target control ball can be obtained first, and the target monitoring area of ​​the target control ball can be determined based on the target device unique identifier. The mapping relationship between the preset device unique identifier and the monitoring area can be pre-stored, and the target monitoring area corresponding to the target device unique identifier can be determined based on the mapping relationship; then, the initial monitoring parameters of the target control ball can be obtained. Specifically, the initial monitoring parameters can be resolution. The configuration software of the target control ball can be started, and the configuration software will display relevant information and parameter setting options of the target control ball, thereby obtaining the initial monitoring parameters. For example, there will be a "device parameters" or "image settings" menu in the software interface of the configuration software. Click to enter the menu to view and modify parameters such as resolution. Of course, you can also refer to the instruction manual or technical manual of the target control ball to obtain the above-mentioned initial monitoring parameters.

[0087] During the third preset time period, the target monitoring area can be monitored using the target control ball with the initial monitoring parameters to obtain a first monitoring video. Then, the first monitoring video can be analyzed using a preset clarity analysis method (e.g., a resolution measurement method) to obtain a first clarity. When the first clarity is less than the preset clarity, it indicates that the clarity of the first monitoring video is insufficient, and it is necessary to obtain the difference between the first clarity and the preset clarity. The specific calculation formula is as follows:

[0088] Target clarity difference = |first clarity - preset clarity|;

[0089] According to the above formula, the target clarity difference can be obtained; then, the target adjustment parameter corresponding to the target clarity difference can be determined. Specifically, a preset mapping relationship between the clarity difference and the adjustment parameter can be pre-stored, and the target adjustment parameter corresponding to the target clarity difference can be determined based on the mapping relationship. The value range of the target adjustment parameter can be -0.2 to 0.2; then, the initial monitoring parameter can be adjusted according to the target adjustment parameter. The specific calculation formula is as follows:

[0090] Target monitoring parameter = initial monitoring parameter × (1 + target adjustment parameter);

[0091] The target monitoring parameters can be obtained according to the above formula; within the fourth preset time period, the target monitoring area can be monitored by the target monitoring parameters through the target control ball to obtain a second monitoring video; then, the target monitoring video is determined based on the first monitoring video and the second monitoring video. For example, the first monitoring video and the second monitoring video can be spliced ​​in chronological order to obtain the target monitoring video; in addition, within the first preset time period, the target control ball can also read the data of the target sensor module at a preset time interval, thereby obtaining the target sensor data.

[0092] In this way, by determining the first clarity of the first surveillance video and comparing it with the preset clarity, when the first clarity is less than the preset clarity, the corresponding target adjustment parameters are determined by calculating the clarity difference, and then the initial monitoring parameters are adjusted to obtain the target monitoring parameters, thereby improving the clarity of the overall surveillance video.

[0093] S204: Analyze the target surveillance video using the cloud data analysis platform based on a preset flame recognition algorithm to determine whether there is a flame area in the target surveillance video.

[0094] In the embodiment of the present application, the preset flame recognition algorithm can be preset in advance or defaulted.

[0095] In a specific embodiment, the cloud data analysis platform may use a preset flame recognition algorithm (eg, a flame recognition algorithm based on a color model and edge detection) to analyze the target surveillance video to determine whether there is a flame area in the target surveillance video.

[0096] If there is no flame area in the target surveillance video, it means that there is no hot work in the target operation area and there is no safety risk, so there is no need to perform the following steps.

[0097] S205 : When the flame area exists in the target monitoring video, determine the starting time of the flame in the flame area.

[0098] In an embodiment of the present application, when a flame region exists in a target surveillance video, the start time of the flame in the flame region is determined. Specifically, starting from the starting frame (starting video frame) of the target surveillance video, frame by frame analysis is performed. For each frame, a preset flame recognition algorithm is used to determine whether the frame contains a flame region. When a frame containing a flame region is first detected, the timestamp corresponding to the frame is the start time of the flame. To avoid misjudgment, a judgment threshold can also be set. For example, when several consecutive frames (such as 3 to 5 frames) are judged to contain a flame region, it is determined that the flame has actually appeared. At this time, the timestamp of the first frame in this group of consecutive frames is taken as the start time of the flame. This is because in actual situations, there may be a misjudgment of one frame (such as due to factors such as sudden changes in light). Confirming multiple frames can improve the accuracy of the judgment.

[0099] S206 , intercepting video frames of the flame area in the target monitoring video within a second preset time period to obtain multiple video frames; the starting time is the middle time of the second preset time period.

[0100] In the embodiment of the present application, the second preset time period may be preset in advance or default.

[0101] In a specific embodiment, video frames of the flame area in the target monitoring video within the second preset time period may be intercepted in chronological order, thereby obtaining a plurality of video frames.

[0102] Optionally, step S206, intercepting the video frames of the flame area in the target monitoring video within the second preset time period to obtain multiple video frames, may include the following steps:

[0103] S61, intercepting the surveillance video of the target surveillance video within the second preset time period to obtain a third surveillance video;

[0104] S62: intercepting the surveillance video of the flame area in the third surveillance video to obtain a fourth surveillance video;

[0105] S63: Determine a target duration of the second preset time period;

[0106] S64, obtaining the target accuracy corresponding to the target dynamic fire recognition model;

[0107] S65, determining a first interception time interval according to the target accuracy;

[0108] S66. Determine a target optimization factor corresponding to the target duration;

[0109] S67: Optimize the first interception time interval according to the target optimization factor to obtain a second interception time interval;

[0110] S68. Intercept video frames of the fourth surveillance video according to the second interception time interval to obtain the multiple video frames.

[0111] In an embodiment of the present application, a video editing tool can be used to capture the video of the target surveillance video within the second preset time period to obtain a third surveillance video; similarly, a video editing tool can be used to capture the surveillance video of the flame area in the third surveillance video to obtain a fourth surveillance video; then, the time length of the second preset time period can be obtained to obtain the target duration; then, the target accuracy rate corresponding to the target dynamic fire recognition model can be obtained. Specifically, the mapping relationship between the preset dynamic fire recognition model and the accuracy rate can be pre-stored, and the target accuracy rate corresponding to the target dynamic fire recognition model can be determined based on the mapping relationship.

[0112] Then, the first interception time interval can be determined according to the target accuracy. For example, a mapping relationship between a preset accuracy and an interception time interval can be pre-stored, and the first interception time interval corresponding to the target accuracy can be determined based on the mapping relationship. Then, a target optimization factor corresponding to the target duration can be determined. Similarly, a mapping relationship between a preset duration and an optimization factor can be pre-stored, and the target optimization factor corresponding to the target duration can be determined based on the mapping relationship. The value range of the target optimization factor can be -0.12 to 0.12. The first interception time interval is optimized according to the target optimization factor, and the calculation formula is as follows:

[0113] Second interception time interval = first interception time interval × (1 + target optimization factor);

[0114] According to the above formula, the second interception time interval can be obtained; finally, the video frames of the fourth surveillance video can be intercepted according to the second interception time interval to obtain multiple video frames. Specifically, the frame rate of the fourth surveillance video can be obtained first, and the second interception time interval can be converted into the corresponding frame interval according to the obtained frame rate. For example, if the frame rate is 30fps and the second interception time interval is 1 second, then the frame interval corresponding to the second interception time interval is 30 frames. If the second interception time interval is 0.5 seconds, the frame interval is 30×0.5=15 frames. The video frames in the fourth surveillance video are intercepted according to the calculated frame interval to obtain multiple video frames. For example, if the frame interval is 15 frames, one frame is intercepted every 15 frames read.

[0115] In this way, the third surveillance video is obtained by intercepting the surveillance video of the target surveillance video within the second preset time period, and then the surveillance video of the flame area in the third surveillance video is further intercepted to obtain the fourth surveillance video. These two steps can focus the analysis on the situation in the flame area within a specific time period, greatly reducing the amount of data that needs to be processed, thereby improving data processing efficiency. In addition, the second interception time interval is determined according to the target accuracy and the target optimization factor, and the video frames of the fourth surveillance video are intercepted according to the second interception time interval. This can ensure that the video frames that are most suitable for analyzing the flame situation are obtained. This is because different dynamic fire recognition models have different accuracies. Adjusting the video frame interception time interval according to the accuracy can avoid intercepting too many unnecessary video frames (when the model accuracy is high) or intercepting too few video frames to cause information loss (when the model accuracy is low) while ensuring the effectiveness of the analysis, thereby making the output results of the model more accurate.

[0116] S207. Input the multiple video frames into the target hot work identification model in the cloud data analysis platform to obtain at least one identification result and at least one confidence level; each identification result includes whether hot work exists or does not exist; the identification results and the confidence levels correspond one to one.

[0117] In an embodiment of the present application, multiple video frames are sequentially input into the target fire recognition model, and the target fire recognition model analyzes these multiple video frames, thereby obtaining at least one recognition result and at least one confidence level. It should be explained that one recognition result corresponds to one video frame.

[0118] Optionally, the method may further include the following steps:

[0119] S71, obtaining a first hot work recognition model, a training set, and a test set;

[0120] S72: training the first hot work recognition model according to the training set to obtain a second hot work recognition model;

[0121] S73, testing the second hot work recognition model using the test set to obtain a first accuracy rate;

[0122] S74: When the first accuracy rate is greater than or equal to a preset accuracy rate, determine the target hot work identification model according to the second hot work identification model.

[0123] In the embodiment of the present application, the fire identification model can be at least one of the following: a deep learning model, a convolutional neural network model, a decision tree model, etc., which are not limited here; the preset accuracy rate can be preset or defaulted in advance.

[0124] In a specific embodiment, a first dynamic fire recognition model, a training set and a test set can be obtained first. Specifically, the first dynamic fire recognition model can be a convolutional neural network model, and the training set and the test set can be preset or defaulted in advance; then, the first dynamic fire recognition model can be trained according to the training set to obtain a second dynamic fire recognition model; then, the second dynamic fire recognition model can be tested by the test set to obtain a first accuracy rate. The specific training and testing processes are routine operations and will not be repeated here; when the first accuracy rate is greater than or equal to the preset accuracy rate, it means that the second dynamic fire recognition model has reached the preset performance standard, which means that the model has sufficient accuracy in identifying dynamic fire situations and can meet the needs of actual applications, so the second dynamic fire recognition model can be set as the target dynamic fire recognition model.

[0125] When the first accuracy is lower than the preset accuracy, it indicates that the second dynamic fire recognition model has not reached the preset performance standard, and the second dynamic fire recognition model needs to be further trained until the accuracy of the second dynamic fire recognition model is greater than or equal to the preset accuracy, and it is used as the target dynamic fire recognition model.

[0126] S208. When the target recognition result in the at least one recognition result includes the presence of hot work, determine a first risk probability based on the flame area and the target confidence corresponding to the target recognition result; the target recognition result is any recognition result in the at least one recognition result.

[0127] In an embodiment of the present application, when the target recognition result in at least one recognition result includes the presence of hot work, it indicates that someone is performing hot work in the target operation area, and the hot work safety detection system needs to determine the first risk probability based on the flame area and the target confidence corresponding to the target recognition result.

[0128] Optionally, in step S208, the target basic data includes regional map data and regional object data, and determining the first risk probability based on the target confidence corresponding to the flame area and the target recognition result may include the following steps:

[0129] S81, determining the position coordinates of the flame area in the target operation area according to the area map data to obtain the starting fire position coordinates;

[0130] S82, determining the items within the preset range of the hot work position coordinates based on the area item data, and obtaining p items; p is a natural number;

[0131] S83. Determine the dangerous items among the p items to obtain q dangerous items, where q is a positive integer less than or equal to p.

[0132] S84, determining the target fire intensity corresponding to the ignition position coordinates;

[0133] S85. Determine a first reference risk probability corresponding to the target fire severity;

[0134] S86. Determine the risk factor corresponding to each of the q dangerous goods to obtain q risk factors;

[0135] S87. Determine a second reference risk probability based on the q risk coefficients and the first reference risk probability;

[0136] S88. Determine a target influence coefficient corresponding to the target confidence level;

[0137] S89. Adjust the second reference risk probability according to the target impact coefficient to obtain the first risk probability.

[0138] In the embodiment of the present application, the preset range can be preset in advance or defaulted.

[0139] In a specific embodiment, the position coordinates of the flame area in the target operation area can be determined based on the regional map data to obtain the hot fire position coordinates; then, the items within the preset range of the hot fire position coordinates can be determined based on the regional item data to obtain p items. Specifically, the distance between each item and the hot fire position coordinates can be obtained based on the regional item data to obtain multiple distances. The items within the preset range can be screened out based on these multiple distances to obtain p items. Then, the dangerous items among the p items can be determined to obtain q dangerous items. Specifically, the hot fire operation safety detection system can store a dangerous item database, and the items among the p items that are in the dangerous item database can be determined to obtain q dangerous items.

[0140] Then, the target fire intensity corresponding to the coordinates of the hot work position can be determined. Specifically, the temperature data corresponding to the target sensor data of the coordinates of the hot work position can be obtained, the maximum temperature in the temperature data can be obtained, and the target fire intensity can be determined based on the maximum temperature. The mapping relationship between the preset temperature value and the fire intensity can be pre-stored, and the target fire intensity corresponding to the maximum temperature can be determined based on the mapping relationship. The greater the fire intensity, the higher the safety risk of the hot work operation. Then, the first reference risk probability corresponding to the target fire intensity can be determined. Specifically, the mapping relationship between the preset fire intensity and the reference risk probability can be pre-stored, and the first reference risk probability corresponding to the target fire intensity can be determined based on the mapping relationship. Then, the hazard coefficient corresponding to each of the q dangerous items can be determined to obtain q hazard coefficients. For example, the mapping relationship between the preset dangerous items and the hazard coefficient can be pre-stored, and the q hazard coefficients corresponding to the q dangerous items can be determined based on the mapping relationship, wherein the value range of the hazard coefficient is 0 to 0.5. Further, the second reference risk probability is determined based on the q hazard coefficients and the first reference risk probability. The specific calculation formula is as follows:

[0141] Second reference risk probability = first reference risk probability × (1 + risk factor a) × (1 + risk factor r);

[0142] Among them, the risk coefficient a is the first risk coefficient among the q risk coefficients; the risk coefficient r is the qth risk coefficient among the q risk coefficients, that is, the last risk coefficient; according to the above formula, the second reference risk probability can be obtained; then, the target influence coefficient corresponding to the target confidence level can be determined. Specifically, a mapping relationship between a preset confidence level and an influence coefficient can be pre-stored, and the target influence coefficient corresponding to the target confidence level can be determined based on the mapping relationship, wherein the value range of the influence coefficient is -0.4 to 0.4; finally, the second reference risk probability can be adjusted according to the target influence coefficient. The specific calculation formula is as follows:

[0143] First risk probability = second reference risk probability × (1 + target impact coefficient);

[0144] The first risk probability can be obtained according to the above formula.

[0145] In this way, by determining the coordinates of the hot work location based on the regional map data, and then finding the objects within the preset range of the coordinates, and further determining the dangerous objects therein, this step-by-step positioning method can accurately focus on the dangerous areas and objects directly related to the hot work operation, which helps to accurately identify the source of risk in a complex working environment. In addition, the risk probability corresponding to the degree of fire at the hot work location is considered, and the risk coefficient of the dangerous objects is combined to determine the final risk probability. This process integrates the impact of multiple factors on the risk, realizes the quantitative assessment of the risk, and provides more convincing data support for subsequent safety management or alarm operations.

[0146] S209: Determine a second risk probability corresponding to the flame area according to the target sensor data through the cloud data analysis platform.

[0147] In the embodiment of the present application, the cloud data analysis platform may analyze the target sensor data to determine the second risk probability corresponding to the flame area.

[0148] Optionally, in step S209, the target sensor data includes: first temperature data and first smoke data, and determining the second risk probability corresponding to the flame area based on the target sensor data through the cloud data analysis platform may include the following steps:

[0149] S91, acquiring sensor data within the flame area from the target sensor data to obtain second sensor data; the second sensor data includes: second temperature data and second smoke data;

[0150] S92. Sampling the second temperature data to obtain multiple temperatures and multiple first sampling times; each temperature corresponds to a first sampling time;

[0151] S93, performing straight line fitting according to the multiple temperatures and the multiple first sampling times to obtain a first straight line;

[0152] S94. Determine a first slope corresponding to the first straight line;

[0153] S95. Obtain a maximum temperature value in the second temperature data;

[0154] S96. Determine a reference second risk probability corresponding to the maximum temperature value;

[0155] S97, sampling the second smoke data to obtain multiple smoke concentrations and multiple second sampling times, wherein each smoke concentration corresponds to a second sampling time;

[0156] S98, performing straight line fitting according to the multiple smoke concentrations and the multiple second sampling times to obtain a second straight line;

[0157] S99. Determine a second slope corresponding to the second straight line;

[0158] S910: Determine a first adjustment coefficient corresponding to the first slope;

[0159] S911. Determine a second adjustment coefficient corresponding to the second slope;

[0160] S912. Adjust the reference second risk probability according to the first adjustment coefficient and the second adjustment coefficient to obtain the second risk probability.

[0161] In an embodiment of the present application, the sensor data within the flame area in the target sensor data can be first obtained to obtain the second sensor data. Specifically, the control ball that monitors the flame area among the m control balls can be first obtained to obtain at least one control ball. Then, the device unique identifier corresponding to each control ball in the at least one control ball is obtained to obtain at least one device unique identifier. The corresponding sensor data is extracted from the target sensor data based on the at least one device unique identifier to obtain the second sensor data; then, the second temperature data can be equidistantly sampled to obtain multiple temperatures and multiple first sampling times; then, the multiple temperatures and the corresponding first sampling times in the multiple first sampling times can be combined to obtain multiple first coordinate points, and the least squares method can be used to perform straight line fitting on the multiple first coordinate points to obtain a first straight line; then, the first slope corresponding to the first straight line can be determined. Specifically, the linear equation of the first straight line y=ax+b can be first obtained, where y is the temperature, x is the time point, a is the first slope, and b is the intercept. The first slope can be obtained according to the linear equation.

[0162] Then, the maximum temperature value in the second temperature data can be found; further, the reference second risk probability corresponding to the maximum temperature value can be determined. Specifically, a mapping relationship between a preset temperature value and anomaly probability can be pre-stored, and the reference second risk probability corresponding to the maximum temperature value can be determined based on the mapping relationship. Then, the second smoke data can be equidistantly sampled to obtain multiple smoke concentrations and multiple second sampling times. Then, a straight line fitting can be performed based on the multiple smoke concentrations and the multiple second sampling times to obtain a second straight line. The second slope corresponding to the second straight line is determined. Specifically, the fitting method for the second straight line can be the same as the fitting method for the first straight line, and the method for obtaining the second slope can also be the same as the method for obtaining the first slope. Then, a first adjustment coefficient corresponding to the first slope can be determined. A mapping relationship between a preset slope and the adjustment coefficient can be pre-stored, and the first adjustment coefficient corresponding to the first slope can be determined based on the mapping relationship. The second adjustment coefficient corresponding to the second slope can also be determined based on the mapping relationship. The value range of the adjustment coefficient can be -0.5 to 0.5. Finally, the reference second risk probability can be adjusted according to the first adjustment coefficient and the second adjustment coefficient. The specific calculation formula is as follows:

[0163] Second risk probability = reference second risk probability × (1 + first adjustment coefficient) × (1 + second adjustment coefficient);

[0164] The second risk probability can be obtained according to the above formula.

[0165] In this way, by sampling the second temperature data and the second smoke data respectively, and performing straight-line fitting based on the sampled data to obtain the first and second straight lines, this method can extract representative trend information from continuous sensor data, thereby providing more accurate data support for subsequent risk assessment. The final obtained second risk probability provides a quantitative basis for safety management decisions. In the management of hot work operations, decision makers decide whether to continue the operation, whether to increase fire-fighting measures or adjust personnel evacuation plans based on the second risk probability, thereby improving the safety of hot work operations.

[0166] S2010: When the first risk probability and the second risk probability meet a preset condition, the alarm module is controlled to perform a target alarm operation according to the target recognition result, the flame area, the first risk probability, and the second risk probability.

[0167] In an embodiment of the present application, the preset condition can be preset in advance or defaulted. For example, the preset condition can be: the first risk probability is greater than or equal to the first preset probability, and / or the second risk probability is greater than or equal to the first preset probability, and / or the average value of the first risk probability and the second risk probability is greater than or equal to the first preset probability; wherein, the first preset probability can be preset in advance or defaulted; the alarm operation can include at least one of the following: SMS alarm, email alarm, telephone alarm, broadcast alarm, sound and light alarm, etc., which are not limited here.

[0168] In a specific embodiment, when the first risk probability and the second risk probability meet preset conditions, the alarm module is controlled to perform a target alarm operation according to the target recognition result, the flame area, the first risk probability and the second risk probability. Specifically, the first risk probability and the second risk probability can be calculated according to a preset probability calculation formula. The preset probability calculation formula is as follows:

[0169] Target risk probability = first risk probability × 0.5 + second risk probability × 0.5;

[0170] The target risk probability is obtained according to the above formula, and the target risk level is determined according to the size of the target risk probability. The corresponding target alarm operation is performed according to the target risk level through the alarm module. Specifically, a mapping relationship between preset risk levels and alarm operations can be pre-stored, and the target alarm operation corresponding to the target risk level is determined based on the mapping relationship. Then, the target alarm operation is performed through the alarm module to notify the staff in the target operation area of ​​the abnormal situation of the hot work operation, so that the staff can prepare in advance.

[0171] For example, assuming that the target risk level is the highest level, the target alarm operation can include all feasible operations such as SMS alarm, email alarm, telephone alarm, broadcast alarm, sound and light alarm, etc. Specifically, a high-frequency harsh sound and red flashing light can be emitted by a powerful sound and light alarm in the target operation area, and a loud announcement of "High risk, emergency of hot work, please stop the work immediately and evacuate" can be made through the broadcasting system. In addition, detailed alarm notifications (for example, alarm emails, alarm calls, etc.) can be pushed to all personnel in the target operation area (including personnel who are not in the target operation area, such as emergency team members, etc.), including the precise coordinates of the hot work location, detailed sensor data (for example, temperature 300℃, smoke concentration 0.8mg / m 3 ), high-definition fire pictures taken by surveillance cameras and real-time video streams of fire operations, and can also provide emergency response guidelines (such as evacuation route maps, the location of nearby fire-fighting equipment, etc.).

[0172] The implementation of this application has the following beneficial effects:

[0173] It can be seen that the safety warning method and related products for hot work operations described in this application achieve all-round monitoring of the operation area and reduce monitoring blind spots by arranging multiple control balls in the target operation area. Whether in indoor or outdoor operation areas, potential hot work operation risks can be discovered in a timely manner. Moreover, through the cloud data analysis platform combined with the preset flame recognition algorithm and hot work recognition model, the flame area and hot work operation can be detected quickly and accurately. Once an abnormal situation is found, an alarm is triggered immediately to gain valuable time for taking emergency measures, thereby effectively improving the safety of hot work operations.

[0174] See also Figure 3 , Figure 3 This is a functional unit block diagram of a safety warning device 300 for hot work provided in an embodiment of the present application, which is applied to a control module in a hot work safety detection system. The hot work safety detection system further includes: m control balls, a cloud data analysis platform, and an alarm module, where m is a positive integer. The safety warning device 300 for hot work includes: an acquisition unit 301, a control unit 302, and an alarm unit 303, wherein:

[0175] The acquisition unit 301 is used to acquire target basic data of the target operation area;

[0176] The control unit 302 is configured to arrange the m control balls within the target operation area according to the target basic data and a preset arrangement method;

[0177] The acquisition unit 301 is further configured to collect data in the target operation area through the m control balls within a first preset time period to obtain target monitoring data; the target monitoring data includes target monitoring video and target sensor data;

[0178] The control unit 302 is further configured to analyze the target monitoring video based on a preset flame recognition algorithm through the cloud data analysis platform to determine whether there is a flame area in the target monitoring video; when the flame area exists in the target monitoring video, determine the starting time of the flame in the flame area; intercept video frames of the flame area in the target monitoring video within a second preset time period to obtain multiple video frames; the starting time is the middle time of the second preset time period; input the multiple video frames into the target hot work recognition model in the cloud data analysis platform to obtain at least one recognition result and at least one confidence level; each recognition result includes the presence or absence of hot work; the recognition result and the confidence level correspond one to one; when the target recognition result in the at least one recognition result includes the presence of hot work, determine a first risk probability based on the flame area and the target confidence level corresponding to the target recognition result; the target recognition result is any recognition result among the at least one recognition result; determine a second risk probability corresponding to the flame area based on the target sensor data through the cloud data analysis platform;

[0179] The early warning unit 303 is configured to control the alarm module to perform a target alarm operation according to the target recognition result, the flame area, the first risk probability, and the second risk probability when the first risk probability and the second risk probability meet preset conditions.

[0180] Optionally, in arranging the m control balls in the target operation area according to the target basic data and a preset arrangement method, the control unit 302 is specifically configured to:

[0181] Obtain target device specification parameters of the m control balls;

[0182] Determining a first monitoring coverage area corresponding to the specification parameters of the target device;

[0183] Acquire target light intensity in the target operating area;

[0184] Determining a target interference factor corresponding to the target light intensity;

[0185] Adjusting the first monitoring coverage area according to the target interference factor to obtain a second monitoring coverage area;

[0186] The target operation area is divided into n first areas according to the second monitoring coverage area; the area of ​​each of the n first areas is smaller than the second monitoring coverage area; and n is a positive integer less than or equal to m;

[0187] determining a regional function of each of the n first regions according to the target basic data to obtain n regional functions;

[0188] Determine the area function corresponding to the hot work among the n area functions to obtain i area functions, where i is a positive integer less than or equal to n;

[0189] Determine first areas corresponding to the i area functions in the n first areas, to obtain i first areas;

[0190] Divide each of the i first areas into j second areas according to a preset area, wherein j is a positive integer greater than or equal to i, and the sum of j and ni is equal to m;

[0191] According to the preset arrangement method, a control ball is set in each of the j second areas and the ni first areas; the ni first areas are the first areas of the n first areas except the i first areas.

[0192] Optionally, each of the m control balls includes a camera module and a sensor module. In terms of collecting data in the target operation area through the m control balls to obtain target monitoring data, the acquisition unit 301 is specifically configured to:

[0193] Acquire a target control ball; the target control ball includes a target camera module and a target sensor module; the target control ball is any one of the m control balls;

[0194] Determine the target monitoring area corresponding to the target control ball;

[0195] Obtaining initial monitoring parameters of the target control ball;

[0196] During a third preset time period, the target monitoring area is monitored by the target control ball with initial monitoring parameters to obtain a first monitoring video; the first preset time period includes the third preset time period;

[0197] Determining a first definition corresponding to the first surveillance video;

[0198] When the first definition is less than a preset definition, obtaining a difference between the first definition and the preset definition to obtain a target definition difference;

[0199] Determining a target adjustment parameter corresponding to the target clarity difference;

[0200] Adjusting the initial monitoring parameters according to the target adjustment parameters to obtain target monitoring parameters;

[0201] During a fourth preset time period, the target monitoring area is monitored by the target control ball using the target monitoring parameters to obtain a second monitoring video; the start time of the fourth preset time period is the end time of the third preset time period; and the first preset time period includes the fourth preset time period;

[0202] Determine the target surveillance video according to the first surveillance video and the second surveillance video;

[0203] During the first preset time period, the target control ball reads the data of the target sensor module at a preset time interval to obtain the target sensor data.

[0204] Optionally, the safety warning device 300 for hot work is further specifically used for:

[0205] Obtain the first dynamic fire recognition model, training set and test set;

[0206] Training the first dynamic fire recognition model according to the training set to obtain a second dynamic fire recognition model;

[0207] Testing the second hot work recognition model using the test set to obtain a first accuracy rate;

[0208] When the first accuracy rate is greater than or equal to a preset accuracy rate, the target hot work identification model is determined according to the second hot work identification model.

[0209] Optionally, in the aspect of intercepting the video frame of the flame area in the target monitoring video within the second preset time period to obtain the multiple video frames, the control unit 302 is specifically configured to:

[0210] intercepting the target surveillance video within the second preset time period to obtain a third surveillance video;

[0211] intercepting a surveillance video of the flame area in the third surveillance video to obtain a fourth surveillance video;

[0212] Determining a target duration of the second preset time period;

[0213] Obtaining the target accuracy corresponding to the target dynamic fire recognition model;

[0214] determining a first interception time interval according to the target accuracy;

[0215] Determining a target optimization factor corresponding to the target duration;

[0216] Optimizing the first interception time interval according to the target optimization factor to obtain a second interception time interval;

[0217] Video frames of the fourth surveillance video are intercepted according to the second interception time interval to obtain the multiple video frames.

[0218] Optionally, the target basic data includes regional map data and regional object data. In determining the first risk probability according to the target confidence level corresponding to the flame area and the target recognition result, the control unit 302 is specifically configured to:

[0219] Determine the position coordinates of the flame area in the target operation area according to the area map data to obtain the hot work position coordinates;

[0220] Determine the items within the preset range of the hot work position coordinates according to the area item data, and obtain p items; p is a natural number;

[0221] Determine the dangerous items among the p items to obtain q dangerous items, where q is a positive integer less than or equal to p;

[0222] Determine the target fire intensity corresponding to the ignition position coordinates;

[0223] Determining a first reference risk probability corresponding to the target fire severity;

[0224] Determining a hazard factor corresponding to each of the q dangerous goods to obtain q hazard factors;

[0225] Determining a second reference risk probability based on the q risk coefficients and the first reference risk probability;

[0226] Determining a target influence coefficient corresponding to the target confidence;

[0227] The second reference risk probability is adjusted according to the target impact coefficient to obtain the first risk probability.

[0228] Optionally, the target sensor data includes: first temperature data and first smoke data. In determining the second risk probability corresponding to the flame area based on the target sensor data through the cloud data analysis platform, the control unit 302 is specifically configured to:

[0229] Acquire sensor data within the flame area from the target sensor data to obtain second sensor data; the second sensor data includes: second temperature data and second smoke data;

[0230] Sampling the second temperature data to obtain a plurality of temperatures and a plurality of first sampling times, wherein each temperature corresponds to a first sampling time;

[0231] Performing straight line fitting according to the multiple temperatures and the multiple first sampling times to obtain a first straight line;

[0232] determining a first slope corresponding to the first straight line;

[0233] Obtaining a maximum temperature value in the second temperature data;

[0234] determining a reference second risk probability corresponding to the maximum temperature value;

[0235] Sampling the second smoke data to obtain a plurality of smoke concentrations and a plurality of second sampling times, wherein each smoke concentration corresponds to a second sampling time;

[0236] Performing straight line fitting according to the multiple smoke concentrations and the multiple second sampling times to obtain a second straight line;

[0237] determining a second slope corresponding to the second straight line;

[0238] determining a first adjustment coefficient corresponding to the first slope;

[0239] determining a second adjustment coefficient corresponding to the second slope;

[0240] The reference second risk probability is adjusted according to the first adjustment coefficient and the second adjustment coefficient to obtain the second risk probability.

[0241] In a specific implementation, the safety warning device 300 for hot work described in the embodiment of the present invention can also execute other implementations described in the safety warning method for hot work provided in the above embodiment of the present invention, which will not be repeated here.

[0242] See also Figure 4 , Figure 4 This is a structural diagram of an electronic device provided in an embodiment of the present application, which includes a processor, a memory, a communication interface and one or more programs. The processor, memory and communication interface are interconnected through a bus. The above one or more programs are stored in the above memory and are configured to be executed by the above processor. The one or more programs include instructions for executing other implementation methods described in the safety warning method for hot work operations provided in the above embodiment of the present invention, which will not be repeated here.

[0243] It should be noted that for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by this application.

[0244] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0245] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the above-mentioned units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0246] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0247] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0248] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes a number of instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the above-mentioned methods of each embodiment of the present application. The aforementioned memory includes: various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0249] The above is a detailed introduction to the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for those skilled in the art, according to the idea of ​​the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A safety warning method for hot work, characterized in that: A control module applied to a hot work safety detection system, wherein the hot work safety detection system further comprises: m control balls, a cloud data analysis platform, and an alarm module, where m is a positive integer; and the method comprises: Obtain target basic data of target operation area; Arranging the m control balls within the target operation area according to the target basic data and a preset arrangement method; During a first preset time period, the m control balls are used to collect data in the target operation area to obtain target monitoring data; the target monitoring data includes target monitoring video and target sensor data; Analyzing the target surveillance video using the cloud data analysis platform based on a preset flame recognition algorithm to determine whether there is a flame area in the target surveillance video; When the flame area exists in the target monitoring video, determining the starting time of the flame in the flame area; Intercepting video frames of the flame area in the target monitoring video within a second preset time period to obtain multiple video frames; the starting time is the middle time of the second preset time period; Inputting the plurality of video frames into a target hot work identification model in the cloud data analysis platform to obtain at least one identification result and at least one confidence level; each identification result includes whether a hot work operation exists or does not exist; and the identification results and the confidence levels correspond one to one; When the target recognition result in the at least one recognition result includes the presence of hot work, determining a first risk probability according to the flame area and the target confidence corresponding to the target recognition result; the target recognition result is any recognition result in the at least one recognition result; Determining a second risk probability corresponding to the flame area based on the target sensor data through the cloud data analysis platform; When the first risk probability and the second risk probability meet a preset condition, the control alarm module performs a target alarm operation according to the target recognition result, the flame area, the first risk probability and the second risk probability; The target basic data includes regional map data and regional item data, and determining the first risk probability according to the target confidence corresponding to the flame area and the target recognition result includes: Determine the position coordinates of the flame area in the target operation area according to the area map data to obtain the hot work position coordinates; Determine the items within the preset range of the hot work position coordinates according to the area item data, and obtain p items; p is a natural number; Determine the dangerous items among the p items to obtain q dangerous items, where q is a positive integer less than or equal to p; Determine the target fire intensity corresponding to the ignition position coordinates; Determining a first reference risk probability corresponding to the target fire severity; Determining a hazard factor corresponding to each of the q dangerous goods to obtain q hazard factors; Determining a second reference risk probability based on the q risk coefficients and the first reference risk probability; Determining a target influence coefficient corresponding to the target confidence; The second reference risk probability is adjusted according to the target impact coefficient to obtain the first risk probability.

2. The method according to claim 1, wherein The arranging of the m control balls in the target operation area according to the target basic data and a preset arrangement method includes: Obtain target device specification parameters of the m control balls; Determining a first monitoring coverage area corresponding to the specification parameters of the target device; Acquire target light intensity in the target operating area; Determining a target interference factor corresponding to the target light intensity; Adjusting the first monitoring coverage area according to the target interference factor to obtain a second monitoring coverage area; The target operation area is divided into n first areas according to the second monitoring coverage area; the area of ​​each of the n first areas is smaller than the second monitoring coverage area; and n is a positive integer less than or equal to m; determining a regional function of each of the n first regions according to the target basic data to obtain n regional functions; Determine the area function corresponding to the hot work among the n area functions to obtain i area functions, where i is a positive integer less than or equal to n; Determine first areas corresponding to the i area functions in the n first areas, to obtain i first areas; Divide each of the i first areas into j second areas according to a preset area, wherein j is a positive integer greater than or equal to i, and the sum of j and ni is equal to m; According to the preset arrangement method, a control ball is set in each of the j second areas and the ni first areas; the ni first areas are the first areas of the n first areas except the i first areas.

3. The method according to claim 1, wherein Each of the m control balls includes: a camera module and a sensor module. The m control balls are used to collect data in the target operation area to obtain target monitoring data, including: Acquire a target control ball; the target control ball includes a target camera module and a target sensor module; the target control ball is any one of the m control balls; Determine the target monitoring area corresponding to the target control ball; Obtaining initial monitoring parameters of the target control ball; During a third preset time period, the target monitoring area is monitored by the target control ball with initial monitoring parameters to obtain a first monitoring video; the first preset time period includes the third preset time period; Determining a first definition corresponding to the first surveillance video; When the first definition is less than a preset definition, obtaining a difference between the first definition and the preset definition to obtain a target definition difference; Determining a target adjustment parameter corresponding to the target clarity difference; Adjusting the initial monitoring parameters according to the target adjustment parameters to obtain target monitoring parameters; During a fourth preset time period, the target monitoring area is monitored by the target control ball using the target monitoring parameters to obtain a second monitoring video; the start time of the fourth preset time period is the end time of the third preset time period; and the first preset time period includes the fourth preset time period; Determine the target surveillance video according to the first surveillance video and the second surveillance video; During the first preset time period, the target control ball reads the data of the target sensor module at a preset time interval to obtain the target sensor data.

4. The method according to any one of claims 1 to 3, wherein The method further comprises: Obtain the first dynamic fire recognition model, training set and test set; Training the first dynamic fire recognition model according to the training set to obtain a second dynamic fire recognition model; Testing the second hot work recognition model using the test set to obtain a first accuracy rate; When the first accuracy rate is greater than or equal to a preset accuracy rate, the target hot work identification model is determined according to the second hot work identification model.

5. The method according to any one of claims 1 to 3, wherein The intercepting of the video frame of the flame area in the target monitoring video within the second preset time period to obtain multiple video frames includes: intercepting the target surveillance video within the second preset time period to obtain a third surveillance video; intercepting a surveillance video of the flame area in the third surveillance video to obtain a fourth surveillance video; Determining a target duration of the second preset time period; Obtaining the target accuracy corresponding to the target dynamic fire recognition model; determining a first interception time interval according to the target accuracy; Determining a target optimization factor corresponding to the target duration; Optimizing the first interception time interval according to the target optimization factor to obtain a second interception time interval; Video frames of the fourth surveillance video are intercepted according to the second interception time interval to obtain the multiple video frames.

6. The method according to any one of claims 1 to 3, wherein: The target sensor data includes: first temperature data and first smoke data. The determining, by the cloud data analysis platform, of the second risk probability corresponding to the flame area based on the target sensor data includes: Acquire sensor data within the flame area from the target sensor data to obtain second sensor data; the second sensor data includes: second temperature data and second smoke data; Sampling the second temperature data to obtain a plurality of temperatures and a plurality of first sampling times, wherein each temperature corresponds to a first sampling time; Performing straight line fitting according to the multiple temperatures and the multiple first sampling times to obtain a first straight line; determining a first slope corresponding to the first straight line; Obtaining a maximum temperature value in the second temperature data; determining a reference second risk probability corresponding to the maximum temperature value; Sampling the second smoke data to obtain a plurality of smoke concentrations and a plurality of second sampling times, wherein each smoke concentration corresponds to a second sampling time; Performing straight line fitting according to the multiple smoke concentrations and the multiple second sampling times to obtain a second straight line; determining a second slope corresponding to the second straight line; determining a first adjustment coefficient corresponding to the first slope; determining a second adjustment coefficient corresponding to the second slope; The reference second risk probability is adjusted according to the first adjustment coefficient and the second adjustment coefficient to obtain the second risk probability.

7. A safety warning device for hot work, characterized in that: A control module used in a hot work safety detection system, wherein the hot work safety detection system further comprises: m control balls, a cloud data analysis platform, and an alarm module, where m is a positive integer; the device comprises: an acquisition unit, a control unit, and an early warning unit, wherein: The acquisition unit is used to acquire target basic data of the target operation area; The control unit is configured to arrange the m control balls within the target operation area according to the target basic data and a preset arrangement method; The acquisition unit is further configured to collect data in the target operation area through the m control balls within a first preset time period to obtain target monitoring data; the target monitoring data includes target monitoring video and target sensor data; The control unit is further configured to analyze the target monitoring video based on a preset flame recognition algorithm through the cloud data analysis platform to determine whether there is a flame area in the target monitoring video; when the flame area exists in the target monitoring video, determine the starting time of the flame in the flame area; intercept video frames of the flame area in the target monitoring video within a second preset time period to obtain multiple video frames; the starting time is the middle time of the second preset time period; input the multiple video frames into the target hot work recognition model in the cloud data analysis platform to obtain at least one recognition result and at least one confidence level; each recognition result includes the presence or absence of hot work; the recognition result and the confidence level correspond one to one; when the target recognition result in the at least one recognition result includes the presence of hot work, determine a first risk probability according to the flame area and the target confidence level corresponding to the target recognition result; the target recognition result is any recognition result among the at least one recognition result; determine a second risk probability corresponding to the flame area according to the target sensor data through the cloud data analysis platform; The early warning unit is configured to control the alarm module to perform a target alarm operation according to the target recognition result, the flame area, the first risk probability, and the second risk probability when the first risk probability and the second risk probability meet a preset condition; The target basic data includes regional map data and regional item data. In determining the first risk probability based on the target confidence level corresponding to the flame area and the target recognition result, the control unit is specifically configured to: Determine the position coordinates of the flame area in the target operation area according to the area map data to obtain the hot work position coordinates; Determine the items within the preset range of the hot work position coordinates according to the area item data, and obtain p items; p is a natural number; Determine the dangerous items among the p items to obtain q dangerous items, where q is a positive integer less than or equal to p; Determine the target fire intensity corresponding to the ignition position coordinates; Determining a first reference risk probability corresponding to the target fire severity; Determining a hazard factor corresponding to each of the q dangerous goods to obtain q hazard factors; Determining a second reference risk probability based on the q risk coefficients and the first reference risk probability; Determining a target influence coefficient corresponding to the target confidence; The second reference risk probability is adjusted according to the target impact coefficient to obtain the first risk probability.

8. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store one or more programs and is configured to be executed by the processor, wherein the programs include instructions for executing the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that A computer program for electronic data exchange is stored, wherein the computer program enables a computer to execute the method according to any one of claims 1 to 6.

Images