Forest fire early warning method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of forest fire early warning, in particular to a forest fire early warning method and device, electronic equipment and a storage medium, which are used for improving the accuracy and timeliness of forest fire detection and improving the accuracy of fire position information and the timeliness of reporting. The method comprises the steps that the installation position of a dome camera set in a target forest is determined, the monitoring range of the dome camera set covers the target forest, and the dome camera set comprises at least one dome camera; video frames collected by a target dome camera are received, fire detection is carried out on the video frames by adopting an image detection model, and the target dome camera is any one dome camera in a dome camera set; and when it is detected that a fire exists in the video frame, fire position information is determined based on the installation position of the target dome camera and holder information carried by the video frame, and the fire position information is reported to the target object.

Description

Technical Field

[0001] The present application relates to the technical field of forest fire warning, and in particular to a forest fire warning method, device, electronic equipment and storage medium. Background Art

[0002] Forest fire refers to a forest fire that loses human control, spreads and expands freely in the forest, causing certain harm and losses to the forest, forest ecosystem and humans. Forest fire is a natural disaster that is highly sudden, destructive and difficult to handle and rescue.

[0003] In the prior art, in the early stage of a forest fire, the forest fire is often detected by manual patrol, fire detection using an observation tower, or fire detection using a satellite.

[0004] However, the manual patrol method has the disadvantages of limited patrol area, large detection errors, and high human resource consumption; the method of using observation towers for fire detection still relies on the experience of the observers, with large deviations and low accuracy; the method of using satellites for fire detection faces the problem that the detection range is affected by the satellite operation cycle. When the timing is not right, it is easy to fail to detect the fire, causing the fire to spread.

[0005] In the prior art, if a fire occurs in a forest, the fire location information is usually reported through manual analysis. However, manual analysis and reporting rely on the experience of patrollers, and the fire location reporting may have a large deviation, thus causing delays in fire fighting.

[0006] Therefore, how to detect forest fires more timely and accurately, and report the fire location information accurately and in a timely manner, is a technical problem that needs to be solved at present. Summary of the invention

[0007] The embodiments of the present application provide a forest fire warning method, device, electronic device and storage medium to improve the accuracy and timeliness of forest fire detection, and improve the accuracy of fire location information and the timeliness of reporting.

[0008] The specific technical solutions provided by the embodiments of this application are as follows: In a first aspect, a forest fire early warning method is provided, the method comprising: Determine an installation location of a ball camera set in a target forest, wherein a monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera; Receive video frames captured by a target dome camera, and use an image detection model to perform fire detection on the video frames, wherein the target dome camera is any dome camera in the dome camera set; When a fire is detected in a video frame, the fire location information is determined based on the installation position of the target dome camera and the pan / tilt information carried in the video frame, and the fire location information is reported to the target object.

[0009] In a possible embodiment, determining the installation location of the ball camera set in the target forest includes: Obtain the location information of each low point of the target forest; Based on the location information of each low point, each low point is clustered to obtain multiple cluster centers; Based on multiple cluster centers and the visible distance of the ball cameras, the installation location of the ball camera collection in the target forest is determined.

[0010] Through the above method, combined with the mountain topographic map, it is possible to maximize the utilization of resources and avoid waste of resources.

[0011] In a possible embodiment, determining the installation location of the ball camera set in the target forest based on multiple cluster centers and ball camera visibility distances includes: For multiple cluster centers, perform the following operations respectively: If the distance from a cluster center to a target low point is less than the visible distance of the ball camera, then a cluster center is used as an installation location of the ball camera set in the target forest, where a target low point is any low point contained in the cluster corresponding to a cluster center; If the distance from a cluster center to a target low point is greater than or equal to the visible distance of the ball camera, then a cluster center is used as an installation location information of the ball camera set in the target forest, and the point whose distance from a target low point is the visible distance of the ball camera is used as an installation location of the ball camera set in the target forest.

[0012] Through the above method, combined with the mountain topographic map of the target forest and the characteristics of the ball camera, the number of deployments was reduced, and the optimal coverage of different target forests was achieved while installing fewer shooting devices, ensuring the maximum range of forest fire monitoring and faster and more accurate fire detection.

[0013] In a possible embodiment, using an image detection model to perform fire detection on a video frame includes: Obtain a forest fire data set, and train an image detection model to be trained based on the forest fire data set to obtain a trained image detection model, wherein the forest fire data set includes: fire image data of a target forest; Preprocess the video frames; The preprocessed video frames are input into the trained image detection model to obtain the fire detection results.

[0014] Through the above method, fires of different sizes and characteristics can be identified in the image, which improves the accuracy and real-time performance of fire detection and meets the need for reporting forest fires as soon as possible.

[0015] In a possible embodiment, the gimbal information includes: a current shooting magnification and a current rotation angle; Based on the installation location of the target ball camera and the pan / tilt information carried by the video frame, the fire location information is determined, including: Obtain a relationship table, and query the relationship table to obtain the target distance corresponding to the current shooting magnification, wherein the relationship table is used to: record the candidate distances corresponding to each candidate shooting magnification, and the candidate distances are: the straight-line distance between the center point position of the shooting picture and the installation position of the target ball camera under the corresponding candidate shooting magnification; Based on the current rotation angle, determine the direction deviation between the current shooting image and the true north direction; Based on the target distance, direction deviation and the installation position of the target ball camera, the center point position of the current shooting picture is determined, and the center point position of the current shooting picture is used as the fire position information.

[0016] Through the above method, the fire location information can be accurately and quickly determined, providing effective location guidance for subsequent fire fighting.

[0017] In a possible embodiment, reporting the fire location information to the target object includes: Based on the fire location information, video frames and the installation location of the target dome camera, early warning information is generated and reported to the target object using a preset early warning method.

[0018] Through the above method, rich warning information can be quickly reported to the target object, achieving real-time warning and solving the problem of untimely human warning leading to further spread of fire.

[0019] In a possible embodiment, the ball camera set has a scheduled patrol task, and the monitoring ranges of different ball cameras in the ball camera set cover the same area in different time periods.

[0020] In a second aspect, the present application provides a forest fire warning device, the device comprising: A determination module, used to determine the installation location of the ball camera set in the target forest, wherein the monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera; A detection module is used to receive video frames collected by a target dome camera and perform fire detection on the video frames using an image detection model, wherein the target dome camera is any dome camera in the dome camera set; The early warning module is used to determine the fire location information based on the installation position of the target ball camera and the pan / tilt information carried by the video frame when a fire is detected in the video frame, and report the fire location information to the target object.

[0021] In a possible embodiment, when determining the installation location of the ball camera set in the target forest, the determination module is further used to: Obtain the location information of each low point of the target forest; Based on the location information of each low point, each low point is clustered to obtain multiple cluster centers; Based on multiple cluster centers and the visible distance of the ball cameras, the installation location of the ball camera collection in the target forest is determined.

[0022] In a possible embodiment, when determining the installation location of the ball camera set in the target forest based on the multiple cluster centers and the ball camera visibility distance, the determination module is further used to: For multiple cluster centers, perform the following operations respectively: If the distance from a cluster center to a target low point is less than the visible distance of the ball camera, then a cluster center is used as an installation location of the ball camera set in the target forest, where a target low point is any low point contained in the cluster corresponding to a cluster center; If the distance from a cluster center to a target low point is greater than or equal to the visible distance of the ball camera, then a cluster center is used as an installation location information of the ball camera set in the target forest, and the point whose distance from a target low point is the visible distance of the ball camera is used as an installation location of the ball camera set in the target forest.

[0023] In a possible embodiment, when the image detection model is used to perform fire detection on the video frame, the detection module is further used to: Obtain a forest fire data set, and train an image detection model to be trained based on the forest fire data set to obtain a trained image detection model, wherein the forest fire data set includes: fire image data of a target forest; Preprocess the video frames; The preprocessed video frames are input into the trained image detection model to obtain the fire detection results.

[0024] In a possible embodiment, the gimbal information includes: a current shooting magnification and a current rotation angle; Based on the installation position of the target ball camera and the PTZ information carried by the video frame, when determining the fire location information, the early warning module is also used to: Obtain a relationship table, and query the relationship table to obtain the target distance corresponding to the current shooting magnification, wherein the relationship table is used to: record the candidate distances corresponding to each candidate shooting magnification, and the candidate distances are: the straight-line distance between the center point position of the shooting picture and the installation position of the target ball camera under the corresponding candidate shooting magnification; Based on the current rotation angle, determine the direction deviation between the current shooting image and the true north direction; Based on the target distance, direction deviation and the installation position of the target ball camera, the center point position of the current shooting picture is determined, and the center point position of the current shooting picture is used as the fire position information.

[0025] In a possible embodiment, when reporting the fire location information to the target object, the early warning module is further used to: Based on the fire location information, video frames and the installation location of the target dome camera, early warning information is generated and reported to the target object using a preset early warning method.

[0026] In a possible embodiment, the ball camera set has a scheduled patrol task, and the monitoring ranges of different ball cameras in the ball camera set cover the same area in different time periods.

[0027] According to a third aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any one of the methods described in the first aspect are implemented.

[0028] According to a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in any one of the first aspects are implemented.

[0029] In a fifth aspect, the present application provides a computer program product, comprising: a computer program code, when the computer program code is run on a computer, the computer executes any one of the methods described in the first aspect.

[0030] In the embodiment of the present application, the installation position of the ball camera set in the target forest is first determined, wherein the monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera. After the ball camera set is installed, the video frame collected by the target ball camera is received, and the image detection model is used to detect fire in the video frame, wherein the target ball camera is any ball camera in the ball camera set. If a fire is detected in the video frame, the fire location information is determined based on the installation position of the target ball camera and the pan / tilt information carried by the video frame, and the fire location information is reported to the target object. In this way, the installation of the ball camera set realizes the maximum range and most effective monitoring of the target forest, solves the problem that manual inspections, satellites and other methods cannot monitor all the time and in all directions. At the same time, more accurate and more real-time detection is achieved through the image detection model, which improves the accuracy and timeliness of forest fire detection. In addition, through the installation position of the target ball camera and the pan / tilt information carried by the video frame, the fire location information can be determined in a timely and accurate manner, and a quick and accurate early warning is achieved, which avoids the further spread of the fire caused by untimely early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a system architecture diagram of the forest fire early warning system in the embodiment of the present application; Figure 2 This is a flowchart of an implementation method of a forest fire early warning method provided in an embodiment of the present application; Figure 3A A first schematic diagram of the installation of a ball camera in an embodiment of the present application; Figure 3B A second schematic diagram of the installation of a ball camera in an embodiment of the present application; Figure 4 A schematic diagram of a process for determining the installation location of a ball camera set in a target forest in an embodiment of the present application; Figure 5 A schematic diagram of distance calculation in an embodiment of the present application; Figure 6 This is a schematic diagram of the structure of a forest fire warning device provided in an embodiment of the present application; Figure 7 Schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme in the embodiment of the present application will be clearly and completely described below in conjunction with the drawings in the embodiment of the present application. Obviously, the described embodiment is only a part of the embodiment of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application. In the absence of conflict, the embodiments in the present application and the features in the embodiments can be arbitrarily combined with each other. In addition, although the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in an order different from that here.

[0033] The terms "first" and "second" in the specification and claims of the present application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the term "comprising" and any of their variations are intended to cover non-exclusive protection. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices. "Multiple" in the present application can mean at least two, for example, two, three or more, and the embodiments of the present application are not limited.

[0034] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description. It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, which should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0035] In the technical solution of this application, the acquisition, transmission, storage, and use of data are in compliance with the requirements of relevant national laws and regulations.

[0036] To facilitate understanding of the technical solutions provided in the embodiments of the present application, some key terms used in the embodiments of the present application are explained here: Ball camera: The full name is spherical camera, which is a camera device that can rotate 360 ​​degrees to achieve monitoring without blind spots.

[0037] Pan / Tilt / Zoom (PTZ): stands for PTZ all-round (left / right / up / down) movement and lens zoom / zoom control.

[0038] The following is a brief introduction to the design concept of the embodiment of the present application: In recent years, forest fires have occurred frequently. Forest fires refer to forest fires that are out of human control, spread freely and expand in the forest, causing certain harm and losses to the forest, forest ecosystem and humans. Forest fires are a natural disaster that is highly sudden, destructive and difficult to deal with and rescue.

[0039] In the prior art, in the early stage of a forest fire, the forest fire is often detected by manual patrol, fire detection using an observation tower, or fire detection using a satellite.

[0040] However, the manual patrol method has the disadvantages of limited patrol area, large detection errors, and high human resource consumption; the method of using observation towers for fire detection still relies on the experience of the observers, with large deviations and low accuracy; the method of using satellites for fire detection faces the problem that the detection range is affected by the satellite operation cycle. When the timing is not right, it is easy to fail to detect the fire, causing the fire to spread.

[0041] In the prior art, if a fire occurs in a forest, the fire location information (i.e., the physical location information of the fire point) is usually reported through manual analysis. However, manual analysis and reporting rely on the experience of patrollers, and the fire location reporting may have a large deviation, thus causing delays in fire fighting.

[0042] Therefore, how to detect forest fires more timely and accurately, and report the fire location information accurately and in a timely manner, is a technical problem that needs to be solved at present.

[0043] In view of this, in the embodiment of the present application, a forest fire early warning method, device, electronic device and storage medium are provided. First, the installation position of the ball camera set in the target forest is determined, wherein the monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera. After the ball camera set is installed, the video frame collected by the target ball camera is received, and the video frame is detected by using an image detection model. The target ball camera is any ball camera in the ball camera set. If a fire is detected in the video frame, the fire location information is determined based on the installation position of the target ball camera and the pan / tilt information carried by the video frame, and the fire location information is reported to the target object. In this way, the installation of the ball camera set realizes the maximum range and most effective monitoring of the target forest, solves the problem that human inspection, satellite and other methods cannot monitor all the time and in all directions. At the same time, more accurate and more real-time detection is realized through the image detection model, which improves the accuracy and timeliness of forest fire detection. In addition, through the installation position of the target ball camera and the pan / tilt information carried by the video frame, the fire location information can be determined in a timely and accurate manner, and a quick and accurate early warning is realized, so as to avoid the further spread of the fire caused by untimely early warning.

[0044] See also Figure 1 , which is a system architecture diagram of a forest fire early warning system in an embodiment of the present application. The forest fire early warning system includes a ball camera 110 and a ground station 120 , and the ground station 120 is deployed with an image detection model 1201 and an operation and maintenance platform 1202 .

[0045] The ball camera 110 is used to collect video frames and transmit the video frames to the ground station 120. The ball camera 110 itself has implemented 5G transmission technology. By inserting an Internet of Things card, the video frame signals collected by the ball camera 110 can be quickly and accurately transmitted to the ground station 120. The power supply method of the ball camera 110 includes but is not limited to: a combination of solar panels and lithium batteries. The built-in GPS module of the ball camera 110 will provide the location information of the ball camera 110. The number of ball cameras 110 in the embodiment of the present application is not limited.

[0046] The image detection model 1201 is used to perform fire detection on the video frames collected by the ball camera 110. The image detection model may be a YOLOv8 detection model.

[0047] The operation and maintenance platform 1202 is used to report the warning information to the target object.

[0048] See also Figure 2 As shown, it is an implementation flow chart of a forest fire early warning method provided in an embodiment of the present application. The specific implementation process of the method is as follows: Step 20: Determine the installation location of the dome camera collection in the target forest.

[0049] The monitoring range of the ball camera set covers the target forest. The ball camera set includes at least one ball camera, and each ball camera corresponds to an installation location.

[0050] In the embodiment of the present application, the installation location of the ball camera set in the target forest is determined based on the mountain topographic map of the target forest and the visible distance of the ball camera, which is convenient for subsequent installation.

[0051] Optionally, in an embodiment of the present application, a possible embodiment is provided to determine the installation location of the ball camera set in the target forest, and the specific steps are as follows: Step 200: Obtain the location information of each low point of the target forest.

[0052] In the embodiment of the present application, the location information of each low point of the target forest is obtained based on the mountain topographic map of the target forest.

[0053] Among them, the low points are the concave points in the mountain topographic map.

[0054] Step 201: Based on the position information of each low point, cluster each low point to obtain multiple cluster centers.

[0055] In the embodiment of the present application, clustering is performed on each low point based on the position information of each low point to obtain multiple cluster centers.

[0056] The clustering method may be a k-means clustering algorithm, etc., which is not limited in the embodiments of the present application.

[0057] Step 202: Based on multiple cluster centers and the visible distance of the ball cameras, determine the installation location of the ball camera set in the target forest.

[0058] In the embodiment of the present application, a plurality of cluster centers are used as installation locations of the ball camera set in the target forest, and then based on the visible distance of the ball camera, the installation location of the ball camera set in the target forest is added.

[0059] In this way, combined with mountain topographic maps, it is possible to maximize the utilization of resources and avoid waste of resources.

[0060] Optionally, in an embodiment of the present application, a possible embodiment is provided to determine the installation location of the ball camera set in the target forest based on multiple cluster centers and the ball camera visual distance. The specific process is as follows. For multiple cluster centers, the following operations are performed respectively: Step 2020: Determine whether the distance from a cluster center to a target low point is less than the visible distance of the ball camera. If so, execute step 2021; otherwise, execute step 2022.

[0061] Among them, a target low point is any low point contained in the cluster corresponding to a cluster center.

[0062] For example, the cluster corresponding to the cluster center H1 includes low point M1 and low point M2, and the target low point is low point M1 or low point M2.

[0063] Step 221: A cluster center is used as an installation location of the ball camera set in the target forest. In the embodiment of the present application, if the distance from a cluster center to a target low point is less than the visible distance of the ball camera, a cluster center is used as an installation position of the ball camera collection in the target forest.

[0064] For example, see Figure 3A As shown, it is a first schematic diagram of the ball camera installation in the embodiment of the present application, the low point M1 and the low point M2 are a cluster, H1 is the cluster center of M1 and M2, and the distance d from the cluster center H1 to the low point M1 is （H1，M1） Less than the ball camera's visual distance d 球机可视距离 , take the cluster center H1 as an installation location of the ball camera collection in the target forest, install a ball camera at H1, and further, the distance d from the cluster center H1 to the low point M2 （H1，M2） Less than the ball camera's visual distance d 球机可视距离 , the cluster center H1 is taken as an installation location of the ball camera collection in the target forest. In summary, only one ball camera needs to be installed at H1.

[0065] Step 2022: A cluster center is used as an installation location information of the ball camera set in the target forest, and a point whose distance from a target low point is the visible distance of the ball camera is used as an installation location of the ball camera set in the target forest.

[0066] In an embodiment of the present application, if the distance from a cluster center to a target low point is greater than or equal to the visible distance of the ball camera, then the cluster center is used as an installation location information of the ball camera collection in the target forest, and the point in the path from the cluster center to the target low point whose distance from a target low point is the visible distance of the ball camera is used as an installation location of the ball camera collection in the target forest.

[0067] For example, see Figure 3B As shown, it is a second schematic diagram of the ball camera installation in the embodiment of the present application, the low point M1 and the low point M2 are a cluster, H1 is the cluster center of M1 and M2, and the distance d from the cluster center H1 to the low point M1 is （H1，M1） Greater than or equal to the ball camera's visual distance d 球机可视距离, then the cluster center H1 is used as an installation location of the ball camera set in the target forest, and the point H2 whose distance from the low point M1 is the visible distance of the ball camera is used as an installation location of the ball camera set in the target forest, and a ball camera is installed at H1 and H2 respectively. （H1，M2） Greater than or equal to the ball camera's visual distance d 球机可视距离 , the cluster center H1 is used as an installation location of the ball camera set in the target forest, and the point H2 whose distance from the low point M2 is the visible distance of the ball camera is used as an installation location of the ball camera set in the target forest. In summary, a ball camera needs to be installed at H1, H2 and H3 respectively.

[0068] In this way, combined with the mountainous terrain map of the target forest and the characteristics of the ball camera, the number of deployments was reduced, and the optimal coverage of different target forests was achieved while installing fewer shooting devices. This ensured the maximum range of forest fire monitoring, faster and more accurate fire detection, and effectively reduced the use of equipment resources while effectively collecting forest video frames.

[0069] Based on the above embodiments, see Figure 4 FIG. 1 is a flow chart of determining the installation location of the ball camera set in the target forest in an embodiment of the present application, which specifically includes: Step 40: Analyze the environmental terrain information of the target forest and obtain the location information of each low point.

[0070] Step 41: Based on the position information of each low point, cluster each low point to obtain multiple cluster centers.

[0071] Step 42: Use multiple cluster centers as installation locations of the ball camera collection in the target forest.

[0072] Step 43: For multiple cluster centers, perform the following operations respectively: When the distance from a cluster center to a target low point is greater than or equal to the visible distance of the ball camera, a new installation position is added, and the point whose distance from a target low point is the visible distance of the ball camera is used as the new installation position.

[0073] Among them, a target low point is any low point contained in the cluster corresponding to the cluster center.

[0074] Furthermore, in the embodiment of the present application, a scheduled patrol task is set for the dome camera set, and the monitoring ranges of different dome cameras in the dome camera set are set to cover the same area in different time periods, that is, an area in the forest is not covered by the monitoring ranges of multiple dome cameras in a time period. In this way, cross coverage is achieved, which is more conducive to timely detection of fire.

[0075] In addition, it is worth noting that in the embodiment of the present application, the power supply method of the ball camera set includes but is not limited to: a combination of solar panels and lithium batteries, which ensures the self-sufficiency of the power supply of the ball camera, and the built-in GPS of each ball camera will provide the current location information of the ball camera. Since the ball camera itself has its own characteristics, including: it can be powered by solar energy, the Internet of Things network, and the GPS carried by itself can provide device location information, etc., therefore, the use of the ball camera can reduce the cost of supporting facilities caused by the deployment of equipment in the forest environment.

[0076] Step 21: Receive the video frames captured by the target ball camera, and use the image detection model to perform fire detection on the video frames.

[0077] Among them, the image detection model can be a YOLOv8 detection model, which is not limited in the embodiments of the present application.

[0078] The target ball camera is any ball camera in the ball camera set.

[0079] In the embodiment of the present application, the 5G transmission of the ball camera can realize the fast and accurate transmission of the video frames collected by the ball camera to the ground station by inserting an IoT card. There is no need for manpower to regularly obtain the storage data of the ball camera's SD card, and data loss caused by excessive SD card data is avoided.

[0080] Optionally, in the embodiment of the present application, a possible embodiment is provided for using an image detection model to perform fire detection on a video frame, and the specific steps are as follows: Step 210: Obtain a forest fire data set, and train an image detection model to be trained based on the forest fire data set to obtain a trained image detection model.

[0081] The forest fire dataset includes: fire image data of the target forest, and fire image data of other forests, such as a large number of forest fire pictures currently available on the Internet.

[0082] In an embodiment of the present application, the forest fire data set is preprocessed and divided into a training set and a test set according to a certain ratio. The training set is used to train the image detection model to be trained, and the test set is used to perform performance evaluation, thereby obtaining a trained image detection model.

[0083] Step 211: Preprocess the video frame.

[0084] Among them, preprocessing includes: image scaling, normalization, data enhancement, denoising and standardization.

[0085] Step 212: Input the preprocessed video frames into the trained image detection model to obtain fire detection results.

[0086] In the embodiment of the present application, the preprocessed video frames are input into a trained image detection model to compare whether there is fire information and obtain fire detection results.

[0087] In this way, the image detection model (YOLOv8 detection model) has high real-time and high accuracy compared to the traditional convolutional neural network and Transform framework, and can achieve faster and more accurate detection under lower hardware configuration. Its multi-scale prediction, adaptive anchor frame, and stronger feature extractor ensure that it can identify fires of different sizes and characteristics in the image, improve the accuracy and real-time performance of fire detection, and meet the need to report forest fires as soon as possible.

[0088] Step 22: When a fire is detected in the video frame, the fire location information is determined based on the installation position of the target dome camera and the pan / tilt information carried in the video frame, and the fire location information is reported to the target object.

[0089] The pan / tilt information is PTZ information, including the current shooting magnification and the current rotation angle. The current rotation angle includes the current horizontal rotation angle and the current vertical rotation angle. The characteristics of the ball camera itself support the video frame to carry the pan / tilt information.

[0090] In an embodiment of the present application, after obtaining the fire detection result of the video frame, if the detection result is that a fire exists, the center point position of the current shooting picture is determined based on the installation position of the target camera and the pan / tilt information carried by the video frame, and the center point position is used as the fire location information, and the fire location information is reported to the target object.

[0091] Optionally, in the embodiment of the present application, a possible embodiment is provided to determine the fire location information, and the specific steps are as follows: Step 220: Obtain a relationship table, and query the relationship table to obtain the target distance corresponding to the current shooting magnification.

[0092] The relationship table is used to record the candidate distances corresponding to each candidate shooting magnification, where the candidate distance is the straight-line distance between the center point of the shooting picture and the installation position of the target ball camera at the corresponding candidate shooting magnification. The target distance is the straight-line distance between the center point of the shooting picture and the installation position of the target ball camera at the current shooting magnification.

[0093] In the embodiment of the present application, the straight-line distance between the center point position of the shooting picture and the installation position of the target ball camera under different candidate shooting magnifications is pre-calculated and stored in a relationship table. Specifically, for each candidate shooting magnification, the following operations are performed respectively: based on the installation height of the target ball camera (i.e., the distance between the target ball camera and the ground), and the distance between the horizontal position of the target ball camera installation and the horizontal position of the center point under the candidate shooting magnification, the straight-line distance between the center point position of the shooting picture and the installation position of the target ball camera under the candidate shooting magnification is calculated.

[0094] The calculation formula is: m 2 =h m 2 +s m 2 , m is the candidate shooting multiple, h m is the installation height of the target ball machine, s m The distance between the horizontal position of the target ball camera installation and the horizontal position of the center point in multiples of m.

[0095] For example, see Figure 5 As shown, it is a schematic diagram of distance calculation in an embodiment of the present application, where the shooting magnification is 1, based on the installation height h1 of the target ball camera, and the distance s1 between the horizontal position of the target ball camera installation and the horizontal position of the center point at the magnification of 1, the straight-line distance d1 between the center point position of the shooting picture and the installation position of the target ball camera when the candidate shooting magnification is 1 is obtained, that is, the corresponding candidate distance when the candidate shooting magnification is 1.

[0096] Step 220: Based on the current rotation angle, determine the direction deviation between the current shooting image and the true north direction.

[0097] Step 220: Based on the target distance, direction deviation and the installation position of the target ball camera, determine the center point position of the current shooting picture, and use the center point position of the current shooting picture as the fire position information.

[0098] In an embodiment of the present application, the initial center point position is determined based on the target distance and the installation position of the target ball camera, and then the initial center point position is corrected by the direction deviation to obtain the center point position of the current shooting picture, and the center point position of the current shooting picture is used as the fire location information.

[0099] In this way, the fire location information can be accurately and quickly determined, providing effective location guidance for subsequent fire fighting.

[0100] In an embodiment of the present application, after obtaining the fire location information, the fire location information is reported to the target object through the operation and maintenance platform. Specifically, based on the fire location information, video frames and the installation position of the target camera, early warning information is generated, and the preset early warning method is used to report the early warning information to the target object.

[0101] Among them, the target objects include: forest rangers, emergency management personnel and fire departments, etc., and the methods of reporting early warning information include: text messages, phone calls and emails, etc.

[0102] In this way, rich warning information can be quickly reported to the target object, achieving real-time warning and solving the problem of further spread of fire due to untimely human warning. At the same time, subsequent relevant emergency departments can further plan the optimal fire fighting route based on the fire point information in the obtained warning information and combined with road information.

[0103] Based on the same inventive concept, the present application embodiment provides a forest fire warning device, please refer to Figure 6 , the device comprises: The determination module 601 is used to determine the installation position of the ball camera set in the target forest, wherein the monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera; The detection module 602 is used to receive the video frames collected by the target ball camera and perform fire detection on the video frames using an image detection model, wherein the target ball camera is any ball camera in the ball camera set; The early warning module 603 is used to determine the fire location information based on the installation position of the target ball camera and the pan / tilt information carried by the video frame when a fire is detected in the video frame, and report the fire location information to the target object.

[0104] In a possible embodiment, when determining the installation location of the ball camera set in the target forest, the determination module 601 is further used to: Obtain the location information of each low point of the target forest; Based on the location information of each low point, each low point is clustered to obtain multiple cluster centers; Based on multiple cluster centers and the visible distance of the ball cameras, the installation location of the ball camera collection in the target forest is determined.

[0105] In a possible embodiment, when determining the installation location of the ball camera set in the target forest based on the multiple cluster centers and the ball camera visible distances, the determination module 601 is further used to: For multiple cluster centers, perform the following operations respectively: If the distance from a cluster center to a target low point is less than the visible distance of the ball camera, then a cluster center is used as an installation location of the ball camera set in the target forest, where a target low point is any low point contained in the cluster corresponding to a cluster center; If the distance from a cluster center to a target low point is greater than or equal to the visible distance of the ball camera, then a cluster center is used as an installation location information of the ball camera set in the target forest, and the point whose distance from a target low point is the visible distance of the ball camera is used as an installation location of the ball camera set in the target forest.

[0106] In a possible embodiment, when the image detection model is used to perform fire detection on the video frame, the detection module 602 is further used to: Obtain a forest fire data set, and train an image detection model to be trained based on the forest fire data set to obtain a trained image detection model, wherein the forest fire data set includes: fire image data of a target forest; Preprocess the video frames; The preprocessed video frames are input into the trained image detection model to obtain the fire detection results.

[0107] In a possible embodiment, the gimbal information includes: a current shooting magnification and a current rotation angle; When determining the fire location information based on the installation location of the target ball camera and the pan / tilt information carried by the video frame, the early warning module 603 is further used to: Obtain a relationship table, and query the relationship table to obtain the target distance corresponding to the current shooting magnification, wherein the relationship table is used to: record the candidate distances corresponding to each candidate shooting magnification, and the candidate distances are: the straight-line distance between the center point position of the shooting picture and the installation position of the target ball camera under the corresponding candidate shooting magnification; Based on the current rotation angle, determine the direction deviation between the current shooting image and the true north direction; Based on the target distance, direction deviation and the installation position of the target ball camera, the center point position of the current shooting picture is determined, and the center point position of the current shooting picture is used as the fire position information.

[0108] In a possible embodiment, when reporting the fire location information to the target object, the early warning module 603 is further used to: Based on the fire location information, video frames and the installation location of the target dome camera, early warning information is generated and reported to the target object using a preset early warning method.

[0109] In a possible embodiment, the ball camera set has a scheduled patrol task, and the monitoring ranges of different ball cameras in the ball camera set cover the same area in different time periods.

[0110] Based on the above embodiments, see Figure 7 Shown is a schematic diagram of the structure of an electronic device in an embodiment of the present application.

[0111] An embodiment of the present application provides an electronic device, which may include a processor 710 (Center Processing Unit, CPU), a memory 720, an input device 730 and an output device 740, etc. The input device 730 may include a keyboard, a mouse, a touch screen, etc., and the output device 740 may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), a cathode ray tube (Cathode Ray Tube, CRT), etc.

[0112] The memory 720 may include a read-only memory (ROM) and a random access memory (RAM), and provides the processor 710 with program instructions and data stored in the memory 720. In the embodiment of the present application, the memory 720 may be used to store the program of any forest fire warning method in the embodiment of the present application.

[0113] The processor 710 calls the program instructions stored in the memory 720, and the processor 710 is used to execute any forest fire warning method in the embodiments of the present application according to the obtained program instructions.

[0114] Based on the above embodiments, in an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the forest fire warning method in any of the above method embodiments is implemented.

[0115] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0116] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0117] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0119] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A forest fire early warning method, characterized in that: include: Determine an installation location of a ball camera set in a target forest, wherein a monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera; Receiving video frames captured by a target dome camera, and performing fire detection on the video frames using an image detection model, wherein the target dome camera is any dome camera in the dome camera set; When a fire is detected in the video frame, the fire location information is determined based on the installation position of the target dome camera and the pan / tilt information carried by the video frame, and the fire location information is reported to the target object.

2. The method according to claim 1, characterized in that Determining the installation location of the ball camera set in the target forest includes: Obtaining location information of each low point of the target forest; Based on the position information of each low point, clustering each low point to obtain a plurality of cluster centers; Based on the multiple cluster centers and the visible distance of the ball cameras, the installation location of the ball camera set in the target forest is determined.

3. The method according to claim 1, characterized in that The step of determining the installation location of the ball camera set in the target forest based on the multiple cluster centers and the ball camera visual distances includes: For the multiple cluster centers, the following operations are performed respectively: If the distance from a cluster center to a target low point is less than the visual distance of the ball camera, the cluster center is used as an installation position of the ball camera set in the target forest, wherein the target low point is any low point included in the cluster corresponding to the cluster center; If the distance from a cluster center to a target low point is greater than or equal to the visible distance of the ball camera, the cluster center will be used as an installation location information of the ball camera collection in the target forest, and the point whose distance from the target low point is the visible distance of the ball camera will be used as an installation location of the ball camera collection in the target forest.

4. The method according to claim 1, characterized in that The adopting an image detection model to perform fire detection on the video frame includes: Acquire a forest fire data set, and train an image detection model to be trained based on the forest fire data set to obtain a trained image detection model, wherein the forest fire data set includes: fire image data of the target forest; Preprocessing the video frame; The preprocessed video frames are input into a trained image detection model to obtain fire detection results.

5. The method according to claim 1, characterized in that The pan / tilt information includes: current shooting magnification and current rotation angle; Then, the fire location information is determined based on the installation position of the target ball camera and the pan / tilt information carried by the video frame, including: Obtain a relationship table, and query the relationship table to obtain the target distance corresponding to the current shooting magnification, wherein the relationship table is used to: record the candidate distances corresponding to each candidate shooting magnification, and the candidate distances are: the straight-line distance between the center point position of the shooting picture and the installation position of the target ball camera under the corresponding candidate shooting magnification; Based on the current rotation angle, determining a direction deviation between the current shooting picture and the true north direction; Based on the target distance, the direction deviation and the installation position of the target ball camera, the center point position of the current shooting picture is determined, and the center point position of the current shooting picture is used as the fire position information.

6. The method according to claim 1, characterized in that The reporting the fire location information to the target object includes: Based on the fire location information, the video frame and the installation location of the target dome camera, early warning information is generated, and the early warning information is reported to the target object using a preset early warning method.

7. The method according to claim 1, characterized in that The ball camera set has a scheduled patrol mission, and the monitoring ranges of different ball cameras in the ball camera set cover the same area in different time periods.

8. A forest fire early warning device, characterized in that: include: A determination module, used to determine the installation location of the ball camera set in the target forest, wherein the monitoring range of the ball camera set covers the target forest, and the ball camera set includes at least one ball camera; A detection module, used for receiving video frames collected by a target ball camera, and performing fire detection on the video frames using an image detection model, wherein the target ball camera is any ball camera in the ball camera set; The early warning module is used to determine the fire location information based on the installation position of the target ball camera and the pan / tilt information carried by the video frame when a fire is detected in the video frame, and report the fire location information to the target object.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images