Machine vision-based unmanned aerial vehicle for fire extinguishing and use method

Abstract

The embodiment of the specification provides a machine vision-based unmanned aerial vehicle for fire extinguishing and a use method. The method comprises the following steps: acquiring initial patrol parameters of at least one patrol unmanned aerial vehicle in a region; acquiring fire-fighting data in the region, and determining prediction fire data of at least one predicted fire point within a first time; determining corrected patrol parameters of the patrol unmanned aerial vehicle based on the prediction fire data of the at least one predicted fire point in the region and the initial patrol parameters; acquiring first patrol data based on the corrected patrol parameters; judging whether the first patrol data meets a first preset condition, and in response to meeting the first preset condition, determining a fire rescue template; and determining initial fire extinguishing parameters of at least one fire extinguishing unmanned aerial vehicle based on the fire rescue template.

Description

Technical Field

[0001] This manual relates to the field of remote control, and in particular to a machine vision-based firefighting drone and its usage method. Background Technology

[0002] With the increasing frequency of catastrophic fires such as forest fires and urban fires worldwide, traditional firefighting methods are no longer sufficient to meet the requirements for efficiency and safety. Therefore, the application of drone technology in firefighting has attracted significant attention. Drone firefighting systems combine advanced flight control technology, firefighting devices, and intelligent algorithms, offering significant advantages in firefighting. However, existing drone firefighting systems on the market have limitations in certain aspects, such as low firefighting efficiency and complex operation.

[0003] Therefore, it is desirable to provide a machine vision-based firefighting drone and its usage method. This firefighting drone has intelligent prediction and scheduling capabilities, enabling intelligent prediction of fire probability. Furthermore, the drone firefighting system can quickly determine the fire rescue template and formulate flexible firefighting strategies, which will greatly improve the accuracy, efficiency, and safety of firefighting operations, enhance fire control, and promote the further application and development of drone technology in the field of firefighting. Summary of the Invention

[0004] This manual provides a machine vision-based firefighting drone and its usage method, which can intelligently predict the probability of fires. Furthermore, the drone can quickly determine fire rescue templates and formulate flexible firefighting strategies, which will greatly improve the accuracy, efficiency, and safety of firefighting operations and enhance the control of fires.

[0005] One embodiment of this specification provides a method for using a fire-fighting drone based on machine vision. The method includes: acquiring initial patrol parameters of at least one patrol drone within a region; acquiring fire data within the region and determining predicted fire data for at least one predicted fire point within a first time period; determining corrected patrol parameters of the patrol drone based on the predicted fire data of at least one predicted fire point within the region and the initial patrol parameters; acquiring first patrol data based on the corrected patrol parameters; determining whether the first patrol data meets a first preset condition; and determining a fire rescue template in response to meeting the first preset condition; and determining initial fire-fighting parameters of at least one fire-fighting drone based on the fire rescue template.

[0006] One embodiment of this specification provides a machine vision-based firefighting drone system. The system includes: a first acquisition module for acquiring initial patrol parameters of at least one patrol drone within an area; a second acquisition module for acquiring fire data within the area and determining predicted fire data for at least one predicted fire point within a first time period; a corrected patrol parameter module for determining corrected patrol parameters of the patrol drone based on the predicted fire data of at least one predicted fire point within the area and the initial patrol parameters; a third acquisition module for acquiring first patrol data based on the corrected patrol parameters; a first judgment module for judging whether the first patrol data meets a first preset condition, and determining a fire rescue template in response to meeting the first preset condition; and an initial fire extinguishing module for determining initial fire extinguishing parameters of at least one firefighting drone based on the fire rescue template.

[0007] One embodiment of this specification provides a machine vision-based device for using a fire-fighting drone. The device includes a processor and a memory. The memory stores instructions, which, when executed by the processor, enable the device to implement the fire-fighting drone usage method as described in any of the preceding embodiments.

[0008] One embodiment of this specification provides a computer-readable storage medium, characterized in that the storage medium stores computer instructions, and when a computer reads the computer instructions in the storage medium, the computer executes the method of using a fire-fighting drone as described in any of the above embodiments. Attached Figure Description

[0009] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0010] Figure 1 These are schematic diagrams illustrating application scenarios of firefighting drone systems according to some embodiments of this specification;

[0011] Figure 2 This is a schematic diagram of the modules of a firefighting drone system according to some embodiments of this specification;

[0012] Figure 3 This is an exemplary flowchart illustrating the method of using a firefighting drone according to some embodiments of this specification;

[0013] Figure 4 This is an exemplary flowchart illustrating, according to some embodiments of this specification, a method for determining whether a fire development stage meets a second preset condition. Detailed Implementation

[0014] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0015] It should be understood that the terms “system,” “device,” “unit,” and / or “module” used herein are one way to distinguish different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0016] As indicated in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0017] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0018] Figure 1 These are schematic diagrams illustrating application scenarios of firefighting drone systems according to some embodiments of this specification. For example... Figure 1 As shown, the scenario 100 involved in the firefighting drone system may include a processor 110, a drone 120, a storage device 130, a network 140, and a user terminal 150.

[0019] In some embodiments, processor 110 can be used to process information and / or data related to scenario 100. For example, processor 110 can acquire a network graph based on the data. In some embodiments, processor 110 can be local or remote. For example, processor 110 can access information and / or data stored in storage device 130 and user terminal 150 via network 140. Alternatively, processor 110 can be directly connected to storage device 130 and user terminal 150 to access stored information and / or data.

[0020] Drone 120 may include firefighting drones and patrol drones. In some embodiments, patrol drones may also carry simple firefighting tools, such as dry powder fire extinguishers. Firefighting drones may carry different types of fire extinguishers for different causes of fire. Drones may carry at least one type of sensor that can be used to acquire patrol data, such as image sensors, smoke concentration sensors, smoke type sensors, etc.

[0021] Storage device 130 can be used to store data and / or instructions related to the use of firefighting drones. In some embodiments, storage device 130 can store data obtained / acquired from user terminal 150. In some embodiments, storage device 130 can store data and / or instructions used by processor 110 to execute or use in order to perform the exemplary methods described in this application. In some embodiments, storage device 130 may be implemented on a cloud platform.

[0022] In some embodiments, storage device 130 may be connected to network 140 to communicate with one or more components of scenario 100 (e.g., processor 110, drone 120, user terminal 150). One or more components of scenario 100 may access data or instructions stored in storage device 130 via network 140. In some embodiments, storage device 130 may be directly connected to or communicate with one or more components of scenario 100 (e.g., processor 110, drone 120, user terminal 150). In some embodiments, storage device 130 may be part of processor 110. In some embodiments, storage device 130 may be a separate memory. Storage device 140 may store historical data, such as machine learning models, historical data, etc.

[0023] Network 140 can facilitate the exchange of information and / or data. In some embodiments, one or more components of scenario 100 (e.g., processor 110, drone 120, user terminal 150) can send information and / or data to other components of scenario 100 via network 140.

[0024] User terminal 150 can be a device used by a user and connected to drone 120.

[0025] Figure 2 This is a schematic diagram of a firefighting drone operating system according to some embodiments of this specification.

[0026] like Figure 2 As shown, the firefighting drone system 200 may include a first acquisition module 210, a second acquisition module 220, a patrol parameter correction module 230, a third acquisition module 240, a first judgment module 250, and an initial firefighting module 260.

[0027] The first acquisition module 210 is used to acquire the initial patrol parameters of at least one patrol drone within the area;

[0028] The second acquisition module 220 is used to acquire fire protection data in the area and determine the predicted fire data of at least one predicted fire point in the first time period;

[0029] The modified patrol parameter module 230 is used to determine the modified patrol parameters of the patrol drone based on the predicted fire data of at least one predicted fire point in the area and the initial patrol parameters;

[0030] The third acquisition module 240 is used to acquire the first inspection data based on the corrected inspection parameters;

[0031] The first judgment module 250 is used to judge whether the first inspection data meets the first preset condition, and in response to meeting the first preset condition, to determine the fire rescue template.

[0032] The initial fire suppression module 260 is used to determine the initial fire suppression parameters of at least one fire suppression drone based on the fire rescue template.

[0033] In some embodiments, the fire-fighting drone system further includes: a fourth acquisition module for acquiring second patrol data of the current fire; a stage judgment model for determining the fire development stage based on the second patrol data; a second judgment model for judging whether the fire development stage meets a second preset condition; and a target fire extinguishing module for adjusting the initial fire extinguishing parameters as target fire extinguishing parameters in response to meeting the second preset condition.

[0034] The second judgment module further includes: a single current fire intensity determination model, used to determine the current fire intensity development stage of the at least one ignition point at the current time based on the second patrol data; a single fire intensity prediction model, used to acquire environmental data and predict the fire intensity development stage of the at least one ignition point at a second time based on the environmental data and the current fire intensity development stage of the at least one ignition point; an overall fire intensity prediction model, used to determine the overall fire intensity development stage at the second time based on the fire intensity development stage of the at least one ignition point at the second time; an ideal fire intensity determination model, used to determine the overall ideal fire intensity development stage at the second time based on the current fire intensity development stage of the at least one ignition point and historical successful fire extinguishing statistics; and a difference judgment model, used to determine whether the difference between the predicted overall fire intensity development stage at the second time and the overall ideal fire intensity development stage is greater than a preset difference threshold. If it is greater, the fire intensity development stage meets the second preset condition.

[0035] In some embodiments, the stage of fire development is determined by a stage judgment model based on the second patrol data.

[0036] It should be understood that the above modules are merely simple examples of the main modules involved in this specification, and do not represent a complete representation of all relevant content in this application. Some modules and units are not shown in this module diagram, and will not be listed here. Furthermore, the above modules and units are not entirely independent and may overlap with each other.

[0037] Figure 3 This is an exemplary flowchart illustrating a method of using a firefighting drone according to some embodiments of this specification. Flow 300 can be executed by a processor. Figure 3 As shown, process 300 may include the following steps:

[0038] Step 310: Obtain the initial patrol parameters of at least one patrol drone in the area.

[0039] A region can refer to a city, an administrative division within a city, a neighborhood, a street, etc. Initial patrol parameters refer to the patrol parameters initially set for each patrol drone. These parameters include the initial patrol route, initial patrol frequency, initial patrol time, and initial patrol altitude. The initial route can be determined based on the region's area, the distribution of buildings and people within the region. The initial patrol route can be evenly distributed within the region. Initial patrol parameters can be set by the system.

[0040] Step 320: Obtain fire protection data within the area and determine the predicted fire data for at least one predicted ignition point at the first moment.

[0041] The first time can be a future time, a current time, or a period of time.

[0042] Firefighting data can refer to the historical ignition time, historical weather conditions, number of historical fire points, location of historical fire points, causes of historical fires, types of historical fires, and scale of historical fires in the area. Causes of fire are divided into primary and secondary causes. Primary causes refer to the root cause of the fire. Primary causes can include electrical faults, open flames, etc. Secondary causes refer to factors that promote the spread of fire. Secondary causes can include excessive combustible materials, ventilation conditions, etc. There is a correspondence between causes and factors of fire, and this correspondence can be determined manually. Factors of fire include environmental factors, equipment factors, building factors, and human factors. Factors of fire can refer to factors related to causes of fire that can lead to ignition. The correspondence between causes and factors of fire can be preset. For example, open flames correspond to factors including human factors and environmental factors. Electrical faults correspond to equipment factors and environmental factors. Ventilation conditions can include building factors and environmental factors. Environmental factors include ambient temperature, wind direction, wind speed, and weather conditions.

[0043] Firefighting data can include historical false alarm information. This information can be related to false alarms about historical fire locations. The false alarm information can correspond to specific firefighting data points and can be linked to the falsely reported fire ignition factors.

[0044] Fire prediction data can include the predicted number of fire points in the area, the location of the fire points, the cause of the fire, the type of fire, the scale of the fire, and the corresponding time period.

[0045] In some embodiments, at least one historical ignition factor is determined based on historical ignition causes. Based on the historical ignition factor and the ignition factor at a first time, predicted ignition data for at least one predicted ignition point at a first time is determined. A set of historical ignition factors with the highest similarity to a set of ignition factors at a first time is determined, and the corresponding historical ignition data is used as the predicted ignition data at the first time.

[0046] Step 330: Determine the corrected patrol parameters for the patrol drone based on the predicted ignition data of at least one predicted ignition point in the area and the initial patrol parameters.

[0047] A three-dimensional preset range is determined with the predicted ignition point as the center. It is then determined whether this preset range intersects with the initial patrol route. If the preset range intersects with the initial patrol route, the predicted ignition point is added to the initial patrol route as a corrected patrol route. If the three-dimensional preset range intersects with multiple initial patrol routes, the initial patrol route with the shortest connection between a point on the patrol route and the center (predicted ignition point) is selected, and the predicted ignition point is added to that initial patrol route as a corrected patrol route. In some embodiments, a patrol correction coefficient is determined based on the similarity between the ignition factors at the first time and the corresponding historical ignition factors, and a preset relationship table. The initial patrol parameters are then corrected based on the patrol correction coefficient to obtain the corrected patrol parameters. For example, if the ambient temperature of the predicted ignition point A is higher than the historical ignition temperature corresponding to that ignition point, the patrol parameters can be strengthened, for example, by increasing the patrol frequency or patrol time. Conversely, the patrol parameters can be reduced to ensure the effective utilization of the patrol drone.

[0048] By adjusting patrol parameters to account for potential fires in the future, targeted patrols can be conducted on key fire points, improving the accuracy of early fire detection and reducing potential fire losses.

[0049] Step 340: Obtain the first patrol data based on the corrected patrol parameters.

[0050] The initial survey data can be real-time on-site survey data before a fire is detected. Survey data may include smoke concentration, video and / or image data (including infrared data), etc. This initial survey data can be acquired through the sensors of the survey drone.

[0051] Step 350: Determine whether the first inspection data meets the first preset condition; in response to meeting the first preset condition, determine the fire rescue template.

[0052] In some embodiments, the first preset condition may be that the smoke concentration exceeds a smoke concentration threshold within a preset time period, and / or that the image data within the corresponding preset time period is abnormal. The smoke concentration threshold can be determined manually. The image data may include a frame from a video or an image, etc. Abnormal image data can be determined by the difference data between image data from adjacent time periods and the corresponding preset difference threshold.

[0053] In some embodiments, the first preset condition may include a fire occurrence probability exceeding a fire occurrence probability threshold. The fire occurrence probability can be determined based on the first patrol data using a fire occurrence probability model. The fire occurrence probability model can be a machine learning model, trained using historical data, such as a graph neural network model or a deep neural network model.

[0054] The fire probability model can include a comparison layer and a determination layer. The comparison layer takes at least three sets of first patrol data within a preset time period as input and outputs difference data between the first patrol data at at least two adjacent time points (e.g., the ratio of smoke concentration and the similarity of infrared images). The determination layer further determines the probability of a fire based on this difference data. When the probability of a fire exceeds a fire probability threshold, a fire can be determined to have occurred. A fire refers to the situation or state of a fire. In some embodiments, the similarity in the infrared images can also be determined by the total area change of bright hotspot areas at adjacent time points, the area change of clustered bright hotspot areas in the infrared images, and the movement speed of the center of the clustered bright areas in the infrared images. The center of the clustered bright areas is a connected bright area in the image. In the infrared image of the initial fire, the bright hotspot areas of the crowd may overlap with or be indistinguishable from the bright hotspot areas of the fire point. Based on the positional or area changes of the clustered bright hotspot areas in the infrared images at multiple time points, it can be determined that moving bright hotspot areas with smaller area changes represent the crowd, while stationary or slowly moving bright hotspot areas with larger area changes represent the fire point. The total area change of bright hotspots at adjacent time points, the area change of bright hotspots clustered in infrared images, and the moving speed of the center of bright areas clustered in infrared graphics can be directly used as distinguishing data.

[0055] Determining the probability of a fire based on a single set of patrol data is less accurate. However, using patrol data from adjacent time points allows for clear observation of dynamic changes in the fire's location, and indirectly, by observing crowd movement, the probability of a fire can be determined, thus improving the accuracy of fire probability assessment. For example, a half-extinguished cigarette butt might be put out within seconds, or small flames might be detected and extinguished quickly if someone notices them, without requiring the use of firefighting drones. Machine vision can further enhance the accuracy of fire assessment.

[0056] In some embodiments, a corresponding fire rescue template can be determined based on a preset relationship between differential data, predicted fire data, and fire rescue templates. This preset relationship can be set manually or determined based on historical ideal data. In some embodiments, historical ideal data can be historical data from cases where the fire was successfully controlled before it spread.

[0057] Firefighting templates can include the number of firefighting drones, the firefighting height of at least one firefighting drone above the fire site, the tilt angle, and the type and quantity of firefighting materials it carries.

[0058] In some embodiments, a fire rescue template can be determined based on discriminative data. When only one outlier in the discriminative data (e.g., the total area change of bright hotspots at adjacent time points, the area change of clustered bright hotspots in infrared images, and the moving speed of the center of clustered bright areas in infrared graphics) exceeds the corresponding template threshold, a fire rescue template is determined based on this discriminative data. The template threshold here differs from the discriminative threshold mentioned above. A first fire rescue template corresponding to the outlier is determined from the fire rescue template library. When the number of outliers is not unique, the outlier weight corresponding to the outlier is determined from a preset outlier weight library; outliers and corresponding outlier weights are paired one-to-one to determine the outlier weight distribution; a first descriptive vector of the outlier weight distribution is constructed; the first descriptive vector is matched with a preset second descriptive vector corresponding to any second fire rescue template in the fire rescue template library to obtain the vector matching degree; the fire rescue template corresponding to the maximum vector matching degree is selected. The corresponding threshold can be preset manually.

[0059] The working principle and beneficial effects of the above technical solution are as follows:

[0060] The pre-defined fire rescue template library has two modules: a first fire rescue template, corresponding to a single anomaly; and a second fire rescue template, corresponding to multiple anomalies. When multiple anomalies exist, the anomaly weight distribution is determined, a first descriptive vector is constructed, and the second fire rescue template with the highest vector matching degree to the first descriptive vector is used as the anomaly tracking template. This improves the accuracy and rationality of tracking multiple anomalies collectively.

[0061] Step 360: Based on the fire rescue template, determine the initial firefighting parameters for at least one firefighting drone.

[0062] Initial firefighting parameters refer to the operational parameters for a firefighting drone to begin firefighting operations. The firefighting drone's altitude from the fire site, tilt angle, and the type and quantity of extinguishing materials it carries are used as the initial firefighting parameters for at least one firefighting drone in the fire rescue template.

[0063] Step 370: Obtain the second patrol data of the current fire situation; determine the fire development stage based on the second patrol data; determine whether the fire development stage meets the second preset condition; in response to meeting the second preset condition, adjust the initial fire extinguishing parameters as the target fire extinguishing parameters.

[0064] The second type of patrol data can be real-time patrol data of the scene after a fire has been detected. This data may include smoke concentration, video and / or image data (including infrared data), etc. The second type of patrol data can be acquired via drones.

[0065] The stages of fire development can include the fire development stage at at least one ignition point and the overall fire development stage at all ignition points.

[0066] The stages of fire development can include the initial expansion stage, the free-burning stage, the smoldering stage, and the decay stage. Each stage of fire development can be represented by a numerical range. The larger the value, the closer the fire is to its final stage. The correspondence between fire development stages and numerical ranges can be preset manually.

[0067] In some embodiments, the second preset condition may include a difference between the overall fire development stage and the overall ideal fire development stage that is greater than a preset difference threshold. The second preset condition may also include a difference between the overall fire development stage and the overall ideal fire development stage at a second time that is greater than a preset difference threshold.

[0068] For detailed instructions on how to determine whether the second preset condition is met in the fire development stage, please refer to the instruction manual. Figure 4 Related explanations.

[0069] The adjustment scheme can be determined based on a preset relationship between similarity and adjustment coefficients. This preset relationship between similarity and adjustment coefficients can be determined manually.

[0070] The working principle and beneficial effects of the above technical solution are as follows:

[0071] By predicting the ignition point at a future time, the initial patrol parameters of the patrol drone can be modified, enabling the drone to comprehensively and selectively detect potential fire points in a timely manner. Real-time acquisition of the first patrol data based on machine vision allows for timely assessment of whether a fire has occurred on-site. Upon detection, the corresponding rescue template can be determined promptly, allowing for timely control before the fire escalates. If the initial firefighting parameters of the firefighting drone corresponding to the current fire rescue template are insufficient to control the fire, timely dynamic adjustment of the firefighting parameters increases control over the fire. In some embodiments, after continuously adjusting the firefighting parameters, real-time second patrol data of the fire is acquired until the second patrol data of the fire no longer meets a second preset condition.

[0072] Figure 4 This is an exemplary flowchart illustrating, according to some embodiments of this specification, a method for determining whether a fire development stage meets a second preset condition. Flowchart 400 can be executed by a processor. Figure 4 As shown, process 400 may include the following steps:

[0073] Step 410: Based on the second patrol data, determine the current fire development stage of at least one ignition point at the current time.

[0074] In some embodiments, the fire development stage can be determined based on the second patrol data using a stage judgment model. The stage judgment model can be a machine learning model, acquired through training samples, such as a graph neural network model or a deep neural network. The stage judgment model may include a comparison layer and a stage determination layer, wherein the comparison layer of the stage judgment model can be shared with the comparison layer of the fire probability model. The second patrol data is input into the comparison layer, and the output difference data is input into the stage determination layer. The stage determination layer determines the current fire development stage of at least one ignition point.

[0075] Simply relying on differences in smoke concentration or infrared data from patrol data is insufficient to determine the stage of fire development. For example, during the decay stage, smoke concentration may increase compared to the smoldering stage, or the infrared data of the flames may be the same during the decay stage as during the initial expansion stage. Therefore, using secondary patrol data can improve the accuracy of determining the current stage of fire development.

[0076] The comparison layer of the stage judgment model can be shared with the comparison layer of the fire probability model. A well-trained comparison layer can be obtained by training the stage judgment model, which solves the problem of difficulty in obtaining labels for training data.

[0077] Step 420: Obtain environmental data and predict the fire development stage of at least one ignition point at a second time based on the environmental data and the current fire development stage of at least one ignition point.

[0078] The second time can be a future point in time or a period of time. The second time can be different from the first time.

[0079] Environmental data can include wind speed, wind direction, temperature, humidity, etc. This data can be obtained from third-party information platforms (e.g., weather forecasting platforms). Based on the environmental data and the second patrol data, a fire prediction model is used to predict the fire development stage at each ignition point in the second timeframe.

[0080] Since environmental data can affect fire intensity, and considering the actual deviations caused by the environment in fire development, the difference between the overall fire development stage and the overall ideal fire development stage can be increased to improve the sensitivity to fire extinguishing parameter adjustments and enhance the accuracy and timeliness of fire control.

[0081] Step 430: Based on the fire development stage at at least one ignition point in the second time, determine the overall fire development stage in the second time.

[0082] Given the current fire situation, which may spread, the ignition points are divided into first ignition points and second ignition points based on the cause of the fire (first cause and second cause). The first ignition point is caused by the first cause, and the second ignition point is caused by the second cause. In the same fire, the fire development stages of the first ignition point and the second ignition point are different. At the beginning of the fire, the fire intensity of the first ignition point may be stronger than that of the second ignition point. As time goes by, the fire spreads, the number of second ignition points increases, and as time goes by, the fire intensity of the first ignition point may be weaker than that of the second ignition point. Based on the fire development stage value of each ignition point at the second time and the weighting of the first and second ignition points, the overall fire development stage value is determined by formula (1).

[0083] (1)

[0084] This represents the overall fire development stage at the second time point, where n is the number of ignition points. This represents the numerical value of the fire intensity at each ignition point during the second stage of fire development.

[0085] Step 440: Based on the current fire development stage at at least one ignition point and historical successful fire suppression statistics, determine the overall ideal fire development stage at the second time point.

[0086] In some embodiments, the ideal development stage of at least one fire point at a second time is determined by an ideal model based on the current fire development stage and initial extinguishing parameters of at least one fire point, and the overall ideal fire development stage at the second time is determined by formula (1) based on the ideal development stage of at least one fire point at the second time. Historical successful fire extinguishing statistics can be used to train the ideal model.

[0087] Step 450: Determine whether the difference between the predicted overall fire development stage at the second time and the overall ideal fire development stage is greater than a preset difference threshold. If it is greater, the fire development stage meets the second preset condition.

[0088] The working principle and beneficial effects of the aforementioned technical solution are as follows:

[0089] In some embodiments of this specification, by predicting the development of the fire at future points in time and comparing it with the ideal stage of fire development, it is possible to predict in advance whether the current fire rescue template can control the fire. This can effectively avoid human misjudgment and allow for the adjustment of the initial fire extinguishing parameters obtained in advance to further control the fire.

[0090] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects may be any one or a combination of the above, or any other possible beneficial effects.

[0091] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0092] Furthermore, this specification uses specific terms to describe embodiments thereof. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Moreover, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0093] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods described herein. Although various examples have been discussed in the foregoing disclosure of embodiments that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the substance and scope of the embodiments described herein. For example, while the system components described above can be implemented using hardware devices, they can also be implemented solely using software solutions, such as installing the described system on existing servers or mobile devices.

[0094] Similarly, it should be noted that, in order to simplify the descriptions disclosed herein and thus aid in the understanding of one or more embodiments, the foregoing description of embodiments in this specification sometimes combines multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of the single embodiments disclosed above.

[0095] In some embodiments, numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of embodiments are modified in some examples with the terms "approximately," "approximately," or "generally." Unless otherwise stated, "approximately," "approximately," or "generally" indicates that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may be changed depending on the characteristics required by individual embodiments. In some embodiments, numerical parameters should take into account specified significant digits and employ a general method of digit reservation. Although the numerical ranges and parameters used to confirm their breadth of range in some embodiments of this specification are approximate values, in specific embodiments, such values ​​are set as precisely as feasible.

[0096] For each patent, patent application, patent application publication, and other material such as articles, books, specifications, publications, and documents referenced in this specification, the entire contents of which are incorporated herein by reference. This excludes historical application documents that are inconsistent with or conflict with the content of this specification, as well as documents that limit the broadest scope of the claims in this specification (currently or subsequently appended to this specification). It should be noted that in the event of any inconsistency or conflict between the descriptions, definitions, and / or terminology used in the supplementary materials to this specification and the content of this specification, the descriptions, definitions, and / or terminology used in this specification shall prevail.

[0097] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A method for using a firefighting drone based on machine vision, characterized in that, The method includes: Obtain the initial patrol parameters of at least one patrol drone within the area; Acquire fire protection data within the area and determine the predicted fire data of at least one predicted fire point within a first time period; The corrected patrol parameters of the patrol drone are determined based on the predicted fire data of at least one predicted fire point in the area and the initial patrol parameters. First patrol data is obtained based on the corrected patrol parameters, and the patrol data includes smoke concentration, video and / or image data; Determine whether the first inspection data meets the first preset condition, and in response to meeting the first preset condition, determine the fire rescue template; Based on the fire rescue template, determine the initial firefighting parameters for at least one firefighting drone; The method further includes: Obtain second patrol data on the current fire situation; The stage of fire development was determined based on the second patrol data; Determine whether the fire development stage meets the second preset condition; In response to the fulfillment of the second preset condition, the initial fire extinguishing parameters are adjusted and used as the target fire extinguishing parameters; The second preset condition includes that the difference between the overall fire development stage at the second time and the overall ideal fire development stage is greater than a preset difference threshold. The determination of whether the fire development stage meets the second preset condition further includes: Based on the second patrol data, determine the current fire development stage of the at least one ignition point at the current time; Acquire environmental data, and based on the environmental data and the current fire development stage of the at least one ignition point, predict the fire development stage of the at least one ignition point at a second time. Based on the fire development stage at at least one ignition point in the second time, the overall fire development stage in the second time is determined. Based on the current fire development stage of the at least one ignition point and historical successful fire extinguishing statistics, determine the overall ideal fire development stage at the second time. Determine whether the difference between the predicted overall fire development stage at the second time point and the overall ideal fire development stage is greater than a preset difference threshold. If it is greater, the fire development stage meets the second preset condition.

2. The method of using a firefighting drone according to claim 1, characterized in that, Determining the current fire development stage of at least one ignition point at the current time based on the second patrol data includes: determining the fire development stage based on the second patrol data using a stage judgment model.

3. A machine vision-based unmanned aerial vehicle (UAV) system for firefighting, characterized in that, The system includes: The first acquisition module is used to acquire the initial patrol parameters of at least one patrol drone within the area; The second acquisition module is used to acquire fire protection data within the area and determine the predicted fire data of at least one predicted fire point within a first time period. The modified patrol parameter module is used to determine the modified patrol parameters of the patrol drone based on the predicted fire data of at least one predicted fire point in the area and the initial patrol parameters. The third acquisition module is used to acquire first patrol data based on the corrected patrol parameters, the patrol data including smoke concentration, video and / or image data; The first judgment module is used to determine whether the first inspection data meets the first preset condition, and in response to meeting the first preset condition, to determine the fire rescue template. An initial fire suppression module is used to determine the initial fire suppression parameters of at least one fire suppression drone based on the fire rescue template. The system also includes: The fourth acquisition module is used to acquire the second patrol data of the current fire situation; A stage-based assessment model is used to determine the stage of fire development based on the second patrol data. The second judgment model is used to determine whether the fire development stage meets the second preset condition. The target fire suppression module is used to adjust the initial fire suppression parameters as target fire suppression parameters in response to the fulfillment of the second preset condition. The second preset condition includes that the difference between the overall fire development stage at the second time and the overall ideal fire development stage is greater than a preset difference threshold. The second judgment module also includes: A standalone current fire intensity determination model is used to determine the current fire intensity development stage of the at least one ignition point at the current time based on the second patrol data; A predictive model for individual fires is used to acquire environmental data and, based on the environmental data and the current fire development stage of the at least one ignition point, predict the fire development stage of at least one ignition point at a second time. A predictive model for determining the overall fire intensity is used to determine the overall fire intensity development stage at the second time based on the fire intensity development stage at at least one ignition point at the second time. An ideal fire intensity determination model is used to determine the overall ideal fire intensity development stage at a second time based on the current fire intensity development stage of the at least one ignition point and historical successful fire suppression statistics. The difference judgment model is used to determine whether the difference between the predicted overall fire development stage at the second time and the overall ideal fire development stage is greater than a preset difference threshold. If it is greater, the fire development stage meets the second preset condition.

4. The firefighting drone system according to claim 3, characterized in that, include: The stage of fire development is determined based on the second inspection data using a stage judgment model.

5. A machine vision-based device for using a firefighting drone, the device comprising a processor and a memory; the memory being used to store instructions, characterized in that, When the instructions are executed by the processor, the device implements the method of using a fire-fighting drone as described in any one of claims 1 to 2.

6. A computer-readable storage medium, characterized in that, The storage medium stores computer instructions. When the computer reads the computer instructions in the storage medium, the computer runs the method of using a fire-fighting drone as described in any one of claims 1 to 2.

Images