Fire passage occupation identification method and system, electronic device, and storage medium

Abstract

This invention discloses a method, system, electronic device, and storage medium for identifying obstructed fire lanes. The method includes: a robot patrolling a fire lane area within a patrol environment; the robot is equipped with a camera; acquiring captured information within the fire lane area, including images and video streams; pushing the captured information to a backend, and using deep learning methods to analyze and judge the captured information to detect whether vehicles appear in the captured information; detecting vehicles appearing in the captured information, and processing the detection results using a tracking algorithm based on time sequence to determine the vehicle's bounding box and the type of vehicle being tracked; and judging the vehicle's status based on the vehicle's bounding box and vehicle type to determine whether the vehicle is illegally parked and obstructing the fire lane. This invention's method for identifying obstructed fire lanes reduces labor costs while improving efficiency, greatly enhancing the timeliness and safety of the overall solution.

Description

Technical Field

[0001] This invention relates to the field of fire lane obstruction identification technology, and more specifically, to a method, system, electronic device, and storage medium for fire lane obstruction identification. Background Technology

[0002] In industrial parks or large buildings, fire truck access routes may be blocked or obstructed. In the event of an emergency such as a fire, if these routes are occupied or blocked by other vehicles, the fire cannot be dealt with promptly, further aggravating the damage. Therefore, ensuring unobstructed access for fire trucks is an essential part of every industrial park management and fire patrol.

[0003] The current technical solutions mainly include the following:

[0004] 1. Based on fixed cameras, traditional image processing methods are used for analysis and judgment. Traditional image similarity algorithms are used to detect the similarity between images of normal channels and images of channels when they are congested, thereby determining whether the channel is congested, and then pushing alarm information to the server backend for processing.

[0005] 2. Assisted judgment based on special mechanical or electronic devices. By placing these devices near the fire truck access road, laser sensors and mechanical devices are used to detect whether there is vehicle congestion in the road, and if so, a corresponding alarm will be triggered.

[0006] 3. Judgment based on fixed cameras and RFID devices. RFID is used to read vehicle (electronic tag) information, and cameras are used to capture vehicle license plate information to determine if any vehicles are blocking fire lanes.

[0007] However, current solutions are costly in terms of equipment and deployment, requiring the installation of corresponding devices at all fire exits. Furthermore, the use of fixed cameras with image analysis is limited by camera location, making large-scale application very difficult. Summary of the Invention

[0008] One objective of this invention is to provide a new technological solution for identifying obstructed fire lanes, including a method, system, electronic device, and storage medium, which can at least solve problems such as high cost and limited camera placement in existing technologies.

[0009] A first aspect of the present invention provides a method for identifying obstructed fire lanes, comprising:

[0010] The robot patrols the fire truck access area in the environment to be patrolled; the robot is equipped with a camera.

[0011] Acquire photographic information within the fire truck access area, including images and video streams;

[0012] The captured information is pushed to the background, and deep learning methods are used to analyze and judge the captured information to detect whether a vehicle appears in the captured information;

[0013] The vehicles appearing in the captured information are detected, and the detection results are processed by a tracking algorithm according to the time sequence to determine the vehicle's bounding box and the type of vehicle being tracked.

[0014] Based on the vehicle's enclosure and vehicle category, determine the vehicle's status to ascertain whether the vehicle is illegally parked or obstructing traffic.

[0015] Optionally, the steps of patrolling the fire truck access area in the environment to be patrolled using a robot include:

[0016] Set up a robot patrol route and set up capture points along the patrol route;

[0017] According to the patrol route, the robot is controlled to patrol different fire lane areas.

[0018] Optionally, the step of obtaining the shooting information in the fire truck access area includes: taking a snapshot at the snapshot point to obtain the shooting information.

[0019] Optionally, the steps of detecting vehicles appearing in the captured information and processing the detection results using a tracking algorithm based on the time sequence to determine the vehicle's bounding box and the type of vehicle being tracked include:

[0020] Vehicle detection and tracking are performed on each frame of the captured information;

[0021] All tracking addresses are bounding box grouped;

[0022] Cluster the bounding boxes of each group to determine the bounding boxes of the vehicles and the vehicle categories being tracked.

[0023] Optionally, the step of clustering the bounding boxes of each group to determine the bounding boxes of vehicles and the tracked vehicle categories further includes:

[0024] If it is determined that all clustering results can be clustered into one class, the vehicle with the tracking address is determined to be a static vehicle.

[0025] If it is determined that the clustering results cannot be clustered into one class, the vehicle with the tracking address is identified as a dynamic vehicle.

[0026] Based on the identified static and dynamic vehicles, it is determined whether any illegally parked vehicles appear in the captured information.

[0027] Optionally, if it is determined that all vehicles at the tracked addresses are dynamic vehicles, it is determined that no illegally parked vehicles appear in the captured information.

[0028] Optionally, an alarm can be sent to the backend system when an illegally parked vehicle is detected.

[0029] A second aspect of the present invention provides a fire lane obstruction identification system, applied to the fire lane obstruction identification method described in the above embodiments, the fire lane obstruction identification system comprising:

[0030] A route setting module, which enables a robot to patrol the fire truck access area in the environment to be patrolled; the robot is equipped with a camera.

[0031] The acquisition module is used to acquire the shooting information within the fire truck access area, and the shooting information includes images and video streams;

[0032] The push module is used to push the shooting information to the background;

[0033] A deep learning module is used to analyze and judge the captured information to detect whether a vehicle appears in the captured information;

[0034] The detection and tracking module is used to detect vehicles appearing in the captured information and to process the detection results using a tracking algorithm based on the time sequence to determine the bounding box of the vehicle and the type of vehicle being tracked.

[0035] The judgment module is used to determine the status of a vehicle based on its surrounding box and vehicle category, so as to determine whether the vehicle is illegally parked and obstructing traffic.

[0036] A third aspect of the present invention provides an electronic device comprising: a processor and a memory, wherein computer program instructions are stored in the memory, wherein when the computer program instructions are executed by the processor, the processor causes the processor to perform the steps of the fire lane obstruction identification method described in the above embodiments.

[0037] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the fire lane obstruction identification method described in the above embodiments.

[0038] The fire lane obstruction identification method of this invention uses a robot patrol to inspect the fire truck access area, pushes the acquired image information to the backend, and processes it through deep learning methods and tracking algorithms to determine the vehicle's bounding box and the type of vehicle being tracked, and finally determines whether the vehicle is illegally parked and obstructing the access lane. This fire lane obstruction identification method reduces labor costs while improving efficiency, greatly enhancing the timeliness and safety of the overall solution.

[0039] Other features and advantages of the invention will become clear from the following detailed description of exemplary embodiments of the invention with reference to the accompanying drawings. Attached Figure Description

[0040] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments of the invention and, together with their description, serve to explain the principles of the invention.

[0041] Figure 1 This is a flowchart of a method for identifying obstructed fire lanes according to an embodiment of the present invention;

[0042] Figure 2 This is a logic judgment diagram based on deep learning of the fire lane obstruction identification method according to an embodiment of the present invention;

[0043] Figure 3 This is a route map for robot inspection according to an embodiment of the present invention;

[0044] Figure 4 A schematic diagram of the working principle of an electronic device according to an embodiment of the present invention.

[0045] Figure label:

[0046] Fire truck access area 10;

[0047] Robot 20;

[0048] 30 capture points;

[0049] Processor 201;

[0050] Memory 202; Operating System 2021; Application Program 2022;

[0051] Network interface 203;

[0052] Input device 204;

[0053] Hard drive 205;

[0054] Display device 206. Detailed Implementation

[0055] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention.

[0056] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use.

[0057] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0058] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0059] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0060] The method for identifying obstructed fire lanes according to an embodiment of the present invention will be described in detail below with reference to the accompanying drawings.

[0061] See Figure 1 The method for identifying obstructed fire lanes according to an embodiment of the present invention includes:

[0062] S1. The robot 20 patrols the fire truck access area 10 in the environment to be patrolled; the robot 20 is equipped with a camera.

[0063] S2. Acquire camera information within the fire truck access area 10, including images and video streams;

[0064] S3. Push the shooting information to the background and use deep learning methods to analyze and judge the shooting information to detect whether there is a vehicle in the shooting information;

[0065] S4. Detect vehicles appearing in the captured information, and process the detection results using a tracking algorithm based on the time sequence to determine the vehicle's bounding box and the type of vehicle being tracked.

[0066] S5. Based on the vehicle's enclosure and vehicle category, determine the vehicle's status to ascertain whether the vehicle is illegally parked or obstructing traffic.

[0067] In other words, in the fire lane obstruction identification method of this invention, such as Figure 1As shown, firstly, robot 20 can patrol the fire truck access area 10 within the environment to be patrolled. By patrolling back and forth, robot 20 collects relevant data, improving patrol efficiency and saving labor costs. Robot 20 is equipped with a camera to capture photos or videos, eliminating the need for additional equipment and reducing costs. Then, it acquires the captured information within the fire truck access area 10, including images and video streams. Next, the captured information is pushed to the backend, where deep learning methods are used to analyze and judge the information to detect whether vehicles are present in the captured data.

[0068] In this invention, the YOLOv5 method from deep learning is used to locate banners in images. To reduce detection time, the original YOLOv5 backbone network can be modified to a lightweight network like Mobilenetv3. Part of the model's training data comes from open-source networks, while the other part is collected and labeled by ourselves. To improve the model's generality and detection capability, the focus structure in YOLOv5 was removed, and the image input for both training and inference was changed to a 4:3 video resolution ratio for multiscale training. The detector can detect whether vehicles appear in an image and also determine the location of the vehicle's bounding box.

[0069] Then, vehicles appearing in the captured information can be detected, and the detection results can be processed by a tracking algorithm based on the time sequence to determine the vehicle's bounding box and the type of vehicle being tracked. Finally, based on the vehicle's bounding box and vehicle type, the vehicle's status can be determined to identify whether the vehicle is illegally parked or obstructing traffic.

[0070] This invention detects vehicles in each frame of the video and then processes the detection results using a tracking algorithm based on temporal sequence. The tracking method can be based on the idea of ​​Track by Detect, which determines whether two consecutive frames contain the same vehicle. The SORT tracking algorithm can also be used. After tracking, we can obtain the bounding box of each vehicle in each frame of the video and its tracking ID (address) (vehicle bounding boxes with the same ID represent the same vehicle).

[0071] Based on the tracking results, we can comprehensively determine the vehicle's status and thus determine whether any vehicles are illegally parked or obstructing traffic. By using the vehicle's bounding box and tracking ID, we can distinguish between static and dynamic vehicles in the video, thereby eliminating dynamic vehicles from the list of illegally parked vehicles.

[0072] This invention utilizes a robot (20 robots) to inspect and capture fixed-point images of fire lanes in parks and large building complexes. The captured videos are transmitted to a backend server for image algorithm analysis to determine if any illegal parking or blockages are present. The backend server's image algorithm employs a detection and tracking method within neural networks to analyze the images and identify vehicles violating regulations within the fire lane images. This overall solution not only reduces labor costs but also significantly improves efficiency and timeliness.

[0073] Therefore, the fire lane obstruction identification method according to an embodiment of the present invention uses a robot 20 to patrol and inspect the fire truck access area 10, pushes the acquired image information to the background, and processes it through deep learning methods and tracking algorithms to determine the vehicle's bounding box and the type of vehicle being tracked, and finally determines whether the vehicle is illegally parked and obstructing the access lane. This fire lane obstruction identification method reduces labor costs while improving efficiency, greatly enhancing the timeliness and safety of the overall solution.

[0074] According to one embodiment of the present invention, the steps of patrolling a fire truck access area 10 in a patrolled environment by means of a robot 20 include:

[0075] Set up a patrol route for robot 20, and set up capture points 30 along the patrol route;

[0076] Based on the patrol route, the control robot 20 patrols different fire lane areas.

[0077] The steps for obtaining camera information within the fire truck access area 10 include: taking a snapshot at camera point 30 to obtain the camera information.

[0078] In other words, see Figure 3 When the robot 20 patrols a fixed environment, it follows a circular path to repeatedly patrol the scene. The patrol route can be manually set (when setting the route, it must pass through the fire lane area). During the patrol, the robot 20 will perform snapshot operations at preset points. That is, the robot 20 will stop at fixed points and use the camera equipped on its head to capture video streams of the fire lane scene. The video is then transmitted to the backend algorithm server via a network module for analysis and early warning.

[0079] During the patrol of the fire truck access area 10 by the robot 20 in the environment to be patrolled, a patrol route for the robot 20 can be set, and snapshot points 30 can be set along the patrol route. According to the patrol route, the robot 20 is controlled to patrol different fire truck access areas. This invention is based on the inspection robot 20, with the camera placed on the head of the robot 20. It continuously inspects the points set in the fire truck access area and collects video in real time after each point is fixed, pushing the video to the backend algorithm server for image analysis. During the acquisition of image information within the fire truck access area 10, snapshot actions can be performed at the snapshot points 30 to obtain the captured information.

[0080] This invention primarily uses target detection and tracking to identify illegal parking in fire lanes. It employs a robot for patrol and fixed-point image capture to detect illegal parking in fire lane sections within a scenario. This invention breaks away from traditional methods that rely on security personnel or fixed equipment for detection, significantly reducing overall costs, greatly improving efficiency, and enhancing social safety and stability.

[0081] Meanwhile, the present invention uses a neural network method to replace the traditional image processing method, which not only makes it possible for the overall machine vision solution to replace human detection algorithms, but also ensures the robustness, detection rate and accuracy of the overall solution, making the overall solution more competitive.

[0082] In some specific embodiments of the present invention, the steps of detecting vehicles where the captured information appears, and processing the detection results according to the time sequence using a tracking algorithm to determine the vehicle's bounding box and the type of vehicle being tracked include:

[0083] Vehicle detection and tracking are performed on each frame of the captured data.

[0084] All tracking addresses are bounding box grouped;

[0085] Cluster the bounding boxes of each group to determine the bounding boxes of the vehicles and the vehicle categories being tracked.

[0086] The steps of clustering the bounding boxes of each group to determine the bounding boxes of vehicles and the tracked vehicle categories also include:

[0087] If it is determined that all clustering results can be clustered into one class, vehicles with tracking addresses are identified as static vehicles.

[0088] If it is determined that the clustering results cannot be clustered into one class, vehicles with tracking addresses are identified as dynamic vehicles.

[0089] Based on the identification of static and dynamic vehicles, determine whether any illegally parked vehicles appear in the captured information.

[0090] If it is determined that all vehicles at the tracked address are dynamic vehicles, then it is determined that no illegally parked vehicles appear in the captured information.

[0091] If an illegally parked vehicle is detected, an alarm will be sent to the back-end system.

[0092] In other words, the robot 20 patrols specific fire truck access areas 10 in parks, shopping malls, and buildings, collecting real-time video streams captured at fixed points and pushing them to the backend for analysis. When patrolling in a fixed environment, the robot 20 can patrol the scene repeatedly along a certain circular path. The patrol route can be manually set (when setting the route, it must pass through the fire truck access area). During the patrol, the robot 20 will capture images at preset points; that is, the robot 20 will stop at fixed points and use the camera equipped on its head to capture video streams of the fire truck access area. The video is then transmitted to the backend algorithm server via a network module for analysis and early warning.

[0093] The process involves acquiring real-time video images, processing them using deep learning techniques to obtain relevant information, and then determining whether any illegally parked vehicles appear in the image. The entire image analysis workflow can be summarized in three main steps: vehicle detection, vehicle tracking, and comprehensive logical judgment.

[0094] Specifically, this invention uses the YOLOv5 method from deep learning to locate banners in images. To reduce detection time, we modified the original YOLOv5 backbone network to a lightweight network like Mobilenetv3. Part of the training data for the model comes from open-source networks, while the other part is collected and labeled by ourselves. To improve the model's generalizability and detection capability, we removed the focus structure from YOLOv5 and changed the image input for both training and inference to a 4:3 video resolution ratio for multiscale training. Using the detector, we can detect whether vehicles appear in an image and determine the location of their bounding boxes.

[0095] Vehicles are detected in each frame of the video, and then the detection results are processed by a tracking algorithm based on the temporal sequence. This invention uses the Track by Detect approach, that is, it tracks vehicles by determining whether they are the same vehicle based on two consecutive frames. We use the SORT tracking algorithm for tracking. After tracking, the bounding boxes of vehicles in each frame of the video and their tracking IDs are obtained (vehicle bounding boxes with the same ID represent the same vehicle).

[0096] Based on the tracking results, the vehicle's status can be comprehensively determined, thus identifying whether any vehicles are illegally parked or obstructing traffic. By using the vehicle's bounding box and tracking ID, it's possible to distinguish between static and dynamic vehicles in the video, thereby eliminating dynamic vehicles from the illegal parking category. The specific judgment logic is as follows:

[0097] See Figure 2 Based on the tracking IDs of all vehicles in the video, the bounding boxes of vehicles with the same tracking ID are grouped into one category, which we denote as set A_tracking_id{box_k}, where box_k represents the position of the bounding box of the same tracking ID vehicle at different times within the set.

[0098] Perform k-means clustering on the center points of all bounding boxes in the A_tracking_id set, and set the cluster radius to 10 pixels.

[0099] If the clustering result in the set has only one cluster, it means that the vehicle with that tracking ID is a static vehicle; otherwise, it is a dynamic vehicle. In this way, we can obtain the dynamic and static status of each tracking ID.

[0100] We check each tracking ID. If a tracking ID of a static vehicle appears in the video, it indicates that the vehicle is illegally parked, and we push the vehicle's image and bounding box information to the business server to issue an alert. If all tracking IDs in the video are of dynamic vehicles, it means that no illegally parked vehicles are shown in the video, and no alert needs to be pushed.

[0101] In the fire lane obstruction identification method of this invention, a robot 20 is used for inspection combined with fixed-point image capture to detect illegal parking and obstruction of fire lane sections in the scene. This breaks away from the traditional approach of using security personnel or fixed equipment for detection, greatly reducing overall costs, significantly improving efficiency, and also enhancing social safety and stability.

[0102] Using neural networks to replace traditional image processing methods not only makes it possible for the overall machine vision solution to replace human detection algorithms, but also ensures the robustness, detection rate and accuracy of the overall solution, making it more competitive.

[0103] According to a second aspect of the present invention, a fire lane obstruction identification system is provided, which is applied to the fire lane obstruction identification method in the above embodiments. The fire lane obstruction identification system includes a route setting module, an acquisition module, a push module, a deep learning module, a detection and tracking module, and a judgment module.

[0104] Specifically, the route setting module uses robot 20 to patrol the fire truck access area 10 within the environment to be patrolled; robot 20 is equipped with a camera. The acquisition module acquires video information within the fire truck access area 10, including images and video streams. The push module pushes the video information to the backend. The deep learning module analyzes and judges the video information to detect whether any vehicles are present. The detection and tracking module detects vehicles appearing in the video information and processes the detection results using a tracking algorithm based on the time sequence to determine the vehicle's bounding box and the type of vehicle being tracked. The judgment module determines the vehicle's status based on the bounding box and vehicle type to determine whether the vehicle is illegally parked or obstructing traffic.

[0105] The fire lane obstruction identification system according to an embodiment of the present invention uses a robot 20 for inspection combined with fixed-point image capture to detect illegal parking and obstruction of fire lane sections in a scene. This breaks away from the traditional approach of using security personnel or fixed equipment for detection, significantly reducing overall costs, greatly improving efficiency, and also enhancing social safety and stability.

[0106] Using neural networks to replace traditional image processing methods not only makes it possible for the overall machine vision solution to replace human detection algorithms, but also ensures the robustness, detection rate and accuracy of the overall solution, making it more competitive.

[0107] According to a third aspect of the present invention, an electronic device is also provided, comprising: a processor 201 and a memory 202, wherein computer program instructions are stored in the memory 202, wherein when the computer program instructions are executed by the processor 201, the processor 201 causes the processor 201 to perform the steps of the fire lane obstruction identification method in the above embodiments.

[0108] Furthermore, such as Figure 4 As shown, the electronic device also includes a network interface 203, an input device 204, a hard disk 205, and a display device 206.

[0109] The various interfaces and devices described above can be interconnected via a bus architecture. The bus architecture can include any number of interconnecting buses and bridges. Specifically, various circuits of one or more central processing units 201 (CPUs), represented by processor 201, and one or more memories 202, represented by memory 202, are connected together. The bus architecture can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits. It is understood that the bus architecture is used to implement communication between these components. In addition to the data bus, the bus architecture also includes a power bus, a control bus, and a status signal bus, which are well known in the art and therefore will not be described in detail herein.

[0110] The network interface 203 can be connected to a network (such as the Internet, local area network, etc.), obtain relevant data from the network, and save it to the hard disk 205.

[0111] Input device 204 can receive various instructions input by the operator and send them to processor 201 for execution. Input device 204 may include a keyboard or clicking device (e.g., mouse, trackball, touchpad, or touch screen).

[0112] Display device 206 can display the results obtained by the processor 201 executing instructions.

[0113] The memory 202 is used to store the programs and data necessary for the operation of the operating system 2021, as well as intermediate results and other data during the calculation process of the processor 201.

[0114] It is understood that the memory 202 in the embodiments of the present invention may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM), which is used as an external cache. The memory 202 of the apparatus and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory 202.

[0115] In some implementations, memory 202 stores elements such as executable modules or data structures, or subsets thereof, or extended sets thereof: operating system 2021 and application 2022.

[0116] The operating system 2021 includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic business functions and handle hardware-based tasks. The application program 2022 includes various applications, such as a browser, used to implement various application functions. Programs implementing the methods of this invention can be included in the application program 2022.

[0117] When the processor 201 calls and executes the application program 2022 and data stored in the memory 202, specifically the program or instructions stored in the application program 2022, it executes the steps of the fire lane obstruction identification method according to the above embodiment.

[0118] The methods disclosed in the above embodiments of the present invention can be applied to processor 201, or implemented by processor 201. Processor 201 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in processor 201 or by instructions in the form of software. The processor 201 may be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), off-the-shelf programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, and can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor, or processor 201 may be any conventional processor 201, etc. The steps of the methods disclosed in the embodiments of the present invention can be directly embodied in the execution of hardware decoding processor, or executed by a combination of hardware and software modules in decoding processor. The software modules may be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 202. The processor 201 reads the information in memory 202 and, in conjunction with its hardware, completes the steps of the above method.

[0119] It is understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions of this application, or combinations thereof.

[0120] For software implementation, the techniques described herein can be implemented through modules (e.g., procedures, functions, etc.) that perform the functions described herein. The software code can be stored in memory 202 and executed by processor 201. Memory 202 can be implemented in processor 201 or external to processor 201.

[0121] Specifically, the processor 201 is also used to read the computer program and perform the following steps: predicting and outputting the answer to the user's question regarding the charging method for the charging pile.

[0122] According to a fourth aspect of the present invention, a computer-readable storage medium is also provided, which stores a computer program. When the computer program is run by a processor 201, the processor 201 performs the steps of the fire lane obstruction identification method described above.

[0123] In the several embodiments provided in this application, it should be understood that the disclosed methods and apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0124] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can be physically comprised separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0125] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute certain steps of the transmission and reception methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0126] While specific embodiments of the invention have been described in detail by way of examples, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of the invention. Those skilled in the art should understand that modifications can be made to the above embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

1. A method for identifying obstructed fire lanes, characterized in that, include: The robot patrols the fire truck access area in the environment to be patrolled; the robot is equipped with a camera; specifically: the robot patrol route is set and the capture points are set on the patrol route; according to the patrol route, the robot is controlled to patrol different fire truck access areas; Acquire photographic information within the fire truck access area, including images and video streams; The steps for obtaining the shooting information in the fire truck access area include: taking a snapshot at the snapshot point to obtain the shooting information; The captured information is pushed to the background, and a deep learning method is used to analyze and judge the captured information to detect whether a vehicle appears in the captured information; the deep learning method uses YOLOv5 to locate the banner in the image, modifies the original YOLOv5 backbone network to a Mobilenetv3 lightweight network, removes the focus structure in YOLOv5, and changes the image input for model training and inference to a video resolution ratio of 4:3 for multiscale training; The vehicles appearing in the captured information are detected, and the detection results are processed by a tracking algorithm according to the time sequence to determine the bounding boxes of the vehicles and the categories of the vehicles being tracked; the tracking algorithm is the SORT tracking algorithm; vehicle detection and tracking are performed on the information of each frame of the captured information to obtain the bounding boxes and tracking addresses of the vehicles in each frame of the video; the tracking addresses of all vehicles in the video are classified, and the positions of the bounding boxes of vehicles with the same tracking address at different times are grouped into one category, denoted as a set; k-means clustering is performed on the center points of the bounding boxes in the set, and the clustering radius is set to 10 pixels. Based on the vehicle's bounding box and vehicle category, the vehicle's status is determined to identify whether the vehicle is illegally parked or obstructing traffic. Specifically, this includes: if all clustering results are determined to be in one category, the vehicle with the tracking address is identified as a static vehicle; if the clustering results are determined not to be in one category, the vehicle with the tracking address is identified as a dynamic vehicle; and based on the identified static and dynamic vehicles, it is determined whether an illegally parked vehicle appears in the captured information.

2. The method for identifying obstructed fire lanes according to claim 1, characterized in that, If it is determined that all vehicles at the tracked addresses are dynamic vehicles, then it is determined that no illegally parked vehicles appear in the captured information.

3. The method for identifying obstructed fire lanes according to claim 2, characterized in that, If an illegally parked vehicle is detected, an alarm will be sent to the back-end system.

4. A fire lane obstruction identification system, applied to the fire lane obstruction identification method according to any one of claims 1-3, characterized in that, The fire lane obstruction identification system includes: A route setting module, which enables a robot to patrol the fire truck access area in the environment to be patrolled; the robot is equipped with a camera. The acquisition module is used to acquire the shooting information within the fire truck access area, and the shooting information includes images and video streams; The push module is used to push the shooting information to the background; A deep learning module is used to analyze and judge the captured information to detect whether a vehicle appears in the captured information; The detection and tracking module is used to detect vehicles appearing in the captured information and to process the detection results using a tracking algorithm based on the time sequence to determine the bounding box of the vehicle and the type of vehicle being tracked. The judgment module is used to determine the status of a vehicle based on its surrounding box and vehicle category, so as to determine whether the vehicle is illegally parked and obstructing traffic.

5. An electronic device, characterized in that, include: A processor and a memory, wherein computer program instructions are stored in the memory, wherein when the computer program instructions are executed by the processor, the processor causes the processor to perform the steps of the fire lane obstruction identification method according to any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the steps of the fire lane obstruction identification method according to any one of claims 1-3.

Images