CN118915776A - Autonomous patrol and decision method of intelligent patrol robot in park - Google Patents

Abstract

The embodiment of the application discloses an autonomous patrol and decision-making method of an intelligent patrol robot in a park. According to the technical scheme provided by the embodiment of the application, the inspection route is constructed by monitoring the target position and the inspection position in real time, and the abnormal position detection behavior abnormality detection is carried out on the target position of the inspection route. With this can combine to patrol and examine the target and patrol and examine the position and carry out the garden and patrol and examine the action and the position of target and carry out anomaly detection and suggestion, promote the comprehensiveness that the garden was patrolled and examined, ensure the manual operation standard in garden, optimize the garden operation effect.

Description

An autonomous inspection and decision-making method for a park intelligent inspection robot

Technical Field

The embodiments of the present application relate to the technical field of non-electric variable control systems, and in particular to an autonomous inspection and decision-making method for a park intelligent inspection robot.

Background Art

At present, in the application scenarios of non-electric variable control system technology, inspection robots are set up in the park to conduct park inspections, and safety hazards are discovered in time to ensure the safety of production and operation in the park. During the park inspection process, the inspection robot generally conducts abnormal inspections along a fixed route, and collects relevant parameters (such as sensor data) at each inspection location to make abnormal judgments, and then gives abnormal prompts when abnormalities occur.

However, the relevant inspection method can only detect fixed parameters along a fixed route, lacks effective detection and prompts for related abnormal behaviors of park personnel, and its inspection function is relatively simple, making it difficult to achieve ideal inspection results.

Summary of the invention

The embodiment of the present application provides an autonomous inspection and decision-making method for a park intelligent inspection robot, which can conduct park inspections in combination with inspection targets and inspection locations, and perform abnormal detection and prompts on the behavior and location of the inspection targets, thereby improving the comprehensiveness of park inspections, ensuring standardized operations of park personnel, and optimizing park operating results.

In a first aspect, an embodiment of the present application provides an autonomous inspection and decision-making method for a park intelligent inspection robot, comprising:

Call the real-time monitoring video of the park camera, perform target detection based on the real-time monitoring video, and determine the target position according to the target detection result and the position of the corresponding park camera;

Importing the set inspection position, planning the inspection route based on the target position and the inspection position, and moving along the inspection route to perform route inspection;

When moving to the inspection position, collecting setting parameter information of the inspection position, and performing parameter abnormality detection and prompting based on the setting parameter information;

When moving to the target position, a video image of the target position is collected, the inspection target is determined based on the video image, and the behavior information, facial information and position information of the inspection target are identified. It is judged whether the inspection target has behavioral abnormalities based on the behavior information and the facial information, and whether the inspection target has positional abnormalities based on the position information and the facial information. Abnormal prompts are given based on the detected behavioral abnormalities and/or positional abnormalities, and the park camera is called to collect tracking video of the inspection target in real time, and the tracking video is reported to the system background.

Further, the planning of the inspection route based on the target location and the inspection location includes:

The electronic map of the park is called, and an inspection route is planned on the electronic map of the park based on a set path planning algorithm, wherein the inspection route includes the target location and the inspection location.

Further, determining the target position according to the target detection result and the position of the corresponding park camera includes:

When it is determined according to the target detection result that the monitored target enters the designated inspection area, or when it is determined that the number of monitored targets reaches the set number threshold, the target position is determined based on the position of the monitored target and the corresponding park camera.

Further, after planning the inspection route based on the target location and the inspection location, it also includes:

Acquire newly added target location information and/or target location change information, and update the inspection route based on the newly added target location information and/or the target location change information.

Further, judging whether the inspection target has abnormal behavior based on the behavior information and the face information includes:

Based on the behavior information, the specified detection behavior is compared. If the behavior information matches the specified detection behavior, based on the facial information, the whitelist facial data pre-bound to the specified detection behavior is compared. If the facial information does not match the whitelist facial data, it is determined that the inspection target has abnormal behavior.

Further, judging whether the inspection target has a position abnormality based on the position information and the face information includes:

When it is determined that the location information is within the specified abnormality detection area, the facial information is compared with the authorized list data of the specified abnormality detection area, and when the facial information does not match the authorized list data, it is determined that the inspection target has a location abnormality.

Further, after determining that the inspection target has a position abnormality, the method further includes:

Based on the data of the authorized list, the park camera is called to determine the authorized person closest to the inspection target, and an unauthorized entry abnormality alarm is sent to the work terminal bound to the authorized person.

In a second aspect, an embodiment of the present application provides an autonomous inspection and decision-making device for a park intelligent inspection robot, comprising:

A target determination module is used to call the real-time monitoring video of the park camera, perform target detection based on the real-time monitoring video, and determine the target position according to the target detection result and the position of the corresponding park camera;

A route planning module, used to import a set inspection position, plan an inspection route based on the target position and the inspection position, and move along the inspection route to perform route inspection;

The inspection module is used to collect the set parameter information of the inspection position when moving to the inspection position, and perform parameter anomaly detection and prompt based on the set parameter information; when moving to the target position, collect the video image of the target position, determine the inspection target based on the video image, and identify the behavior information, face information and position information of the inspection target, judge whether the inspection target has behavioral abnormalities based on the behavior information and the face information, judge whether the inspection target has positional abnormalities based on the position information and the face information, and give abnormal prompts based on the detected behavior abnormalities and/or positional abnormalities, call the park camera to collect the tracking video of the inspection target in real time, and report the tracking video to the system background.

In a third aspect, an embodiment of the present application provides an electronic device, including:

memory and one or more processors;

The memory is used to store one or more programs;

When the one or more programs are executed by the one or more processors, the one or more processors implement the autonomous inspection and decision-making method of the park intelligent inspection robot as described in the first aspect.

In a fourth aspect, an embodiment of the present application provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute the autonomous inspection and decision-making method of the campus intelligent inspection robot as described in the first aspect.

The embodiment of the present application calls the real-time monitoring video of the campus camera, performs target detection based on the real-time monitoring video, and determines the target position according to the target detection result and the position of the corresponding campus camera; imports the set inspection position, plans the inspection route based on the target position and the inspection position, and moves along the inspection route to perform route inspection; when moving to the inspection position, collects the set parameter information of the inspection position, and performs parameter anomaly detection and prompts based on the set parameter information; when moving to the target position, collects the video image of the target position, determines the inspection target based on the video image, and identifies the behavior information, face information and position information of the inspection target, determines whether the inspection target has behavioral abnormalities based on the behavior information and face information, determines whether the inspection target has positional abnormalities based on the position information and face information, and performs abnormality prompts based on the detected behavior abnormalities and/or positional abnormalities, calls the campus camera to collect the tracking video of the inspection target in real time, and reports the tracking video to the system background. By adopting the above-mentioned technical means, it is possible to conduct park inspections in combination with inspection targets and inspection locations, and to detect and prompt abnormalities in the behavior and location of the inspection targets. This can improve the comprehensiveness of park inspections, ensure the standardized operations of park personnel, and optimize the park's operating results.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of an autonomous inspection and decision-making method of a park intelligent inspection robot provided in Example 1 of the present application;

FIG2 is a schematic diagram of the interaction of the inspection robot in Example 1 of the present application;

FIG3 is a flowchart of abnormal behavior detection in Example 1 of the present application;

FIG4 is a flowchart of position anomaly detection in Embodiment 1 of the present application;

5 is a schematic diagram of the structure of an autonomous inspection and decision-making device of a park intelligent inspection robot provided in Example 1 of the present application;

FIG6 is a schematic diagram of the structure of an electronic device provided in Embodiment 1 of the present application.

DETAILED DESCRIPTION

In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. It is understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for the convenience of description, only part of the present application is shown in the accompanying drawings, but not all of the content. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow chart describes each operation (or step) as a sequential process, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

Embodiment 1:

Figure 1 shows a flow chart of an autonomous inspection and decision-making method for a campus intelligent inspection robot provided in Example 1 of the present application. The autonomous inspection and decision-making method for the campus intelligent inspection robot provided in this embodiment can be executed by the campus intelligent inspection robot, which can be implemented by software and/or hardware. The campus intelligent inspection robot can be composed of two or more physical entities, or it can be composed of one physical entity.

The following description is made by taking the park intelligent inspection robot as the subject of the autonomous inspection and decision-making method of the park intelligent inspection robot as an example. Referring to FIG1 , the autonomous inspection and decision-making method of the park intelligent inspection robot specifically includes:

S110, calling the real-time monitoring video of the park camera, performing target detection based on the real-time monitoring video, and determining the target position according to the target detection result and the position of the corresponding park camera.

The autonomous inspection and decision-making method of the park intelligent inspection robot of this application aims to construct an inspection route by real-time monitoring of the target position and inspection position, and perform position anomaly detection and behavior anomaly detection on the target position of the inspection route. In this way, the park inspection can be carried out in combination with the inspection target and inspection position, and the behavior and position of the inspection target can be detected and prompted abnormally, so as to improve the comprehensiveness of the park inspection, ensure the operation standardization of the park personnel, and optimize the park operation effect.

As shown in Figure 2, the park intelligent inspection robot 11 calls the video stream transmitted in real time by the cameras 12 distributed in the park through the network interface. The video stream is used for target detection using deep learning or traditional computer vision algorithms to identify potential inspection targets such as people and vehicles in the park. Based on the detected target and camera location information, the park map server 13 is called to obtain the park map, and the specific location of the target is determined using the map or spatial positioning technology.

Specifically, in the process of calling the real-time monitoring videos of the campus cameras and performing target detection based on these videos and then determining the target location, it is necessary to combine video processing, computer vision technology and spatial positioning technology.

First, the inspection robot needs to establish a network connection with the camera system in the park, such as through the park local area network (LAN) or wireless network (such as Wi-Fi). The inspection robot sends a request to the camera through the preset IP address, port number and authentication information (such as user name and password) to obtain the real-time video stream. The video stream is transmitted using protocols such as RTSP (Real Time Streaming Protocol), RTMP (Real Time Messaging Protocol) or HTTP.

The acquired video stream needs to be decoded first, and converted from a compressed format (such as H.264, H.265, etc.) into a processable frame image. Use a computer vision library (such as OpenCV) to process the decoded frame image, including resizing, color space conversion (such as conversion from BGR to grayscale or HSV), etc., to facilitate subsequent target detection.

Then, deep learning models (such as YOLO, SSD, Faster R-CNN, etc.) or traditional computer vision methods (such as background difference, frame difference, optical flow, etc.) are applied to detect targets in the processed images. These algorithms can identify target objects in the image, such as people and vehicles, and give the target's bounding box and possible category label.

It is understandable that each camera has a fixed installation location in the park, and this location information can be obtained through the database of the park map, GIS system or camera management system. The location information includes longitude and latitude, altitude, and coordinates relative to a reference point in the park. According to the bounding box in the target detection result and the camera's own viewing angle, focal length and other parameters, the relative position of the target in the camera's field of view can be calculated. Then, combined with the camera's location information and the camera's orientation, pitch and other posture information, the target position is converted from the two-dimensional coordinates in the camera's field of view to three-dimensional coordinates or two-dimensional plane coordinates in the park using spatial positioning algorithms (such as geometric projection, three-dimensional reconstruction, etc.). The calculated target position is then mapped to the park map to facilitate subsequent inspection route planning and exception handling.

Optionally, determining the target position according to the target detection result and the position of the corresponding park camera includes:

When it is determined according to the target detection result that the monitored target enters the designated inspection area, or when it is determined that the number of monitored targets reaches the set number threshold, the target position is determined based on the position of the monitored target and the corresponding park camera.

When determining the target location based on the target detection results and the location of the corresponding park camera, consider

There are many detected targets in the park. For some areas that do not need to be monitored, or when the number of targets is small, they can be appropriately filtered without abnormal inspections. Therefore, this application determines the target location by monitoring whether the target enters the designated inspection area or whether the number of monitored targets reaches the set number threshold.

First, a series of inspection areas are delineated on the park map. These areas can be key equipment areas, safety exits, restricted areas and other areas that need to be monitored. Each inspection area has clear boundaries and signs. When the inspection robot or back-end server receives the real-time monitoring video from the camera and processes it through the target detection algorithm, a series of target detection results will be obtained, including the target category, bounding box and other information. The bounding box of each detected target is compared with the delineated inspection area to determine whether the target is completely or partially within the inspection area. This usually involves geometric calculations, such as calculating the number of intersections between the bounding box and the inspection area boundary, the positional relationship between the center point of the bounding box and the inspection area, etc. If the target is judged to have entered the inspection area, the specific position of the target in the inspection area (such as the coordinates relative to the center of the inspection area) can be further calculated based on the camera position information and the target's bounding box information.

In addition, according to inspection requirements and safety regulations, set a threshold for the number of monitoring targets for different inspection areas or the entire park. This threshold can be a fixed value or dynamically adjusted. During the target detection process, the number of monitoring targets in each inspection area or the entire park is counted in real time. The number of monitoring targets obtained by counting is compared with the set number threshold. If the number of monitoring targets reaches or exceeds the threshold, further location determination of these targets is required. This can be achieved by traversing all detected targets and calculating their locations separately. Since multiple targets may be involved at this time, it is necessary to determine the specific location of each target in the park separately, and target tracking or trajectory analysis is required.

In practical applications, both the target entering the inspection area and the target number reaching the threshold can be considered at the same time to form a more comprehensive inspection strategy. For example, when the number of monitored targets in a certain inspection area suddenly increases and reaches the threshold, the system can automatically trigger a key inspection of the area and track and record the location and behavior of the targets entering the area in real time.

Through the above method, the target position of the corresponding target can be accurately determined for use in subsequent intelligent inspections by inspection robots.

S120, importing the set inspection position, planning the inspection route based on the target position and the inspection position, and moving along the inspection route to perform route inspection.

Furthermore, before the park intelligent inspection robot conducts autonomous inspection, it imports the set inspection position, plans the inspection route in combination with the above target position and the inspection position, and conducts inspection along the planned route to ensure the comprehensiveness of the inspection.

First, according to the actual situation of the park and the inspection needs, a series of inspection locations are preset in the system. These locations can be key equipment, safety exits, monitoring blind spots, crowded areas and other areas that need to be focused on. The detailed information of each inspection location (such as location coordinates, inspection content, inspection cycle, etc.) is entered into the inspection system. This information can be stored and managed through map annotation, database records, etc.

Before the inspection begins, the target location information is updated based on the target detection results of the real-time monitoring video.

Then, the optimal or feasible inspection route is planned by using a path planning algorithm (such as A* algorithm, Dijkstra algorithm, genetic algorithm, etc.) in combination with the inspection location and the target location. The path planning algorithm can consider a variety of factors, such as the distance between inspection locations, road conditions, obstacle distribution, inspection time, etc. After planning the preliminary inspection route, the route can be optimized. For example, the inspection order can be adjusted, similar inspection locations can be merged to reduce repeated paths, and high-risk areas can be avoided.

The inspection robot uses navigation systems (such as GPS, SLAM, inertial navigation, etc.) and positioning technologies (such as laser radar, cameras, etc.) to obtain its own location information in the park in real time and move according to the planned inspection route. When the inspection robot arrives at the preset inspection location, it will use the sensors it carries (such as temperature sensors, pressure sensors, cameras, etc.) to collect parameter information at that location and perform abnormality detection. If an abnormality is found, an abnormal prompt will be given immediately, and further emergency measures may be triggered. During the inspection process, if an emergency occurs (such as road congestion, target location changes, etc.), the inspection robot needs to be able to adjust the inspection route and inspection strategy in real time to ensure the smooth completion of the inspection task. Through the above steps, the park's intelligent inspection robot can efficiently complete the inspection task, promptly detect and handle abnormal situations, and provide strong guarantees for the safe operation of the park.

Optionally, an inspection route is planned based on the target location and the inspection location, including:

The electronic map of the park is called, and an inspection route is planned on the electronic map of the park based on a set path planning algorithm. The inspection route includes a target location and an inspection location.

When planning the inspection route based on the target location and the inspection location, the route is planned by calling the electronic map of the park and using the set path planning algorithm.

First, you need to obtain an electronic map of the park. The electronic map of the park contains information such as the detailed layout of the park, road network, and location of key facilities. The electronic map can come from the park's GIS system, professional map service providers, or information provided by the park management department. According to the requirements of the inspection task, clearly identify the target locations that need to be paid attention to and the preset inspection locations (such as key equipment, emergency exits, etc.), and mark these location information on the electronic map.

Then, according to the actual situation of the park and the inspection requirements, select the appropriate path planning algorithm. Common algorithms include A* algorithm, Dijkstra algorithm, genetic algorithm, etc. These algorithms can calculate the optimal or feasible path based on the given starting point, end point and map information. Set the corresponding parameters according to the requirements of the selected algorithm. For example, in the A* algorithm, it is necessary to set the heuristic function (used to evaluate the estimated cost from the current node to the target node); in the genetic algorithm, it is necessary to set parameters such as population size, crossover probability, and mutation probability. Take the target location and inspection location as the input of the algorithm, call the park electronic map as the source of map information, and run the path planning algorithm. Based on the map information and the set parameters, the algorithm will calculate one or more optimal or feasible paths from the starting point (which can be the inspection starting point or a certain inspection location) to the end point (the target location or the next inspection location). These paths will be saved as inspection routes. Through the above steps, a reasonable and efficient inspection route can be planned on the park electronic map based on the target location and inspection location to ensure the smooth completion of the inspection task.

Optionally, after planning the inspection route based on the target location and the inspection location, the method further includes:

Acquire newly added target location information and/or target location change information, and update the inspection route based on the newly added target location information and/or target location change information.

After planning the inspection route based on the target location and inspection location, in order to cope with the dynamic changes that may occur in the park, such as the addition of new inspection targets or changes in target locations, the inspection system needs to be flexible and real-time so that the inspection route can be updated in a timely manner.

Among them, the inspection system continuously monitors the dynamic changes of the park and sets up corresponding mechanisms to receive notifications of new target locations or changes in target locations. For example, it is implemented through interfaces with park cameras, monitoring systems or other related systems. After receiving the new or changed information, the system verifies it to ensure the accuracy and validity of the information. For example, check whether the coordinates of the target location are within the park range, whether the target location conflicts with the existing inspection location, etc. After verification, based on the new or changed target location information, the system re-runs the path planning algorithm to calculate a new optimal or feasible inspection route. For example, adjust the inspection order, add or delete inspection locations, optimize the path length, etc. When replanning the inspection route, you can also consider optimization factors, such as reducing repeated paths, avoiding high-risk areas, etc.

Through the above steps, the inspection system can flexibly respond to dynamic changes in the park, update the inspection routes in a timely manner, and ensure the comprehensiveness and accuracy of the inspection tasks.

S130. When moving to the inspection position, collect the set parameter information of the inspection position, and perform parameter anomaly detection and prompt based on the set parameter information; when moving to the target position, collect the video image of the target position, determine the inspection target based on the video image, and identify the behavior information, face information and position information of the inspection target, judge whether the inspection target has behavioral abnormalities based on the behavior information and face information, judge whether the inspection target has positional abnormalities based on the position information and face information, and give abnormal prompts based on the detected behavior abnormalities and/or positional abnormalities, call the park camera to collect the tracking video of the inspection target in real time, and report the tracking video to the system background.

During the inspection process, when the inspection robot moves to the inspection location and target location, it will perform a series of sophisticated operations to ensure the comprehensiveness and accuracy of the inspection.

Among them, GPS, SLAM or other positioning technologies are used to confirm that the inspection robot has arrived at the preset inspection location. According to the specific requirements of the inspection location, the set parameter information is collected. Including physical quantities such as temperature, humidity, pressure, current, voltage, etc., and can also include logical quantities such as device status and running time. The collected parameter information is compared with the preset normal range or threshold to determine whether there is any parameter abnormality.

If an abnormal parameter is detected, an immediate abnormal prompt will be given, which can be in the form of sound, light, screen display, etc., to ensure that the inspectors or monitoring system can notice it in time.

The positioning technology is also used to confirm that the inspection robot has reached the target location, and the camera on the inspection robot is called to collect video images of the target location. Based on the video image, the inspection target is determined using image recognition or machine learning algorithms. The inspection target can be a person or an object, depending on the requirements of the inspection task. The inspection target is further analyzed to identify its behavior information (such as walking, staying, running, etc.), facial information (if applicable), and location information (specific location in the video image).

Based on behavior information and face information, determine whether the inspection target has abnormal behavior, such as violations of regulations, abnormal actions, etc. Based on location information and face information (combined with historical records or preset areas), determine whether the inspection target has abnormal location, such as unauthorized entry into sensitive areas, etc. If abnormal behavior or location is detected, an abnormal prompt will be issued immediately, and further emergency response measures may be triggered, such as alarms, notifications to relevant personnel, etc. In addition, by calling the park camera to collect tracking videos of the inspection target in real time, ensure that its dynamics can be continuously monitored. Upload the tracking video to the system background for subsequent analysis, archiving or use as evidence.

According to the abnormal prompts and tracking videos, corresponding abnormal handling measures can be taken, such as dispatching personnel to handle on-site, adjusting inspection strategies, etc. In addition, data analysis can be performed on the collected parameter information, video images, behavior information, face information, and location information to discover potential problems and optimize the inspection process.

Finally, an inspection report is generated based on the inspection results, including detailed information such as inspection time, inspection location, inspection target, abnormal conditions, etc., for management reference and decision-making.

Through the above steps, the inspection robot can complete the inspection tasks comprehensively and accurately, detect and handle abnormal situations in time, and ensure the safety and normal operation of the park.

In this way, by combining real-time monitoring video and preset inspection positions, multi-dimensional inspections can be carried out in the park, including equipment status, personnel behavior and other aspects, which can improve the comprehensiveness and accuracy of inspections; by introducing target detection and behavior recognition technology, abnormal behaviors such as illegal operations and illegal entry into restricted areas in the park can be discovered and handled in a timely manner, thereby improving the safety management level of the park; by autonomously planning inspection routes and executing them, the dependence on manual labor is reduced, the inspection efficiency is improved, and the labor cost is reduced; by

When an abnormality is detected, it can quickly issue an abnormal prompt and report the tracking video, providing park managers with timely and accurate decision-making basis and enhancing emergency response capabilities; through continuous and comprehensive inspections, potential safety hazards and violations can be discovered and dealt with in a timely manner, ensuring the safety and stability of the park's production and operation, thereby optimizing the park's overall operating results.

Optionally, referring to FIG. 3 , judging whether the inspection target has abnormal behavior based on the behavior information and the face information includes:

S1301, comparing the specified detection behavior based on the behavior information, and when the behavior information matches the specified detection behavior, comparing the whitelist face data pre-bound to the specified detection behavior based on the face information;

S1302: When the facial information does not match the facial data in the whitelist, it is determined that the inspection target has abnormal behavior.

In the process of judging whether the inspection target has abnormal behavior based on behavior information and face information, S first collects the behavior information of the inspection target through devices such as cameras. This behavior information includes the target's movements, postures, speed, etc. By pre-defining a series of behaviors that need to be detected in the system or database, these behaviors are considered abnormal or require special attention. For example, wandering in a prohibited area, running fast, abnormal movements, etc. The collected behavior information is compared with the detection behavior specified in the system. For example, algorithms such as image recognition, machine learning, or deep learning are used to analyze and identify behavioral characteristics.

If the behavior information matches the specified detection behavior, it means that the inspection target may be performing an abnormal or attention-grabbing behavior. At this point, it is necessary to further verify whether the behavior is performed by a specific person to avoid false alarms. After confirming that the behavior is abnormal, the system needs to obtain facial information related to the current behavior. For example, the facial image of the inspection target captured by the camera. The system also needs to prepare a whitelist face data, which contains the facial features of people who are allowed to perform specific behaviors or enter specific areas. Compare the acquired facial information with the facial data in the whitelist. If the facial information successfully matches the facial data in the whitelist, it means that the person performing the abnormal behavior is authorized or known, and no abnormal prompt is required. If the facial information does not match the facial data in the whitelist, it means that the person performing the abnormal behavior is unknown or unauthorized. In this case, the system should determine that the inspection target has abnormal behavior.

Once it is confirmed that the inspection target has abnormal behavior, the system should immediately give an abnormal prompt, such as sounding an alarm, displaying abnormal information on the monitoring screen, sending notifications to relevant personnel, etc. At the same time, the system can also automatically take further response measures according to preset rules or processes, such as starting tracking cameras, recording abnormal videos, notifying security personnel, etc.

Through the above process, the system can accurately determine whether the inspection target has abnormal behavior based on behavioral information and facial information, and take corresponding measures in a timely manner to ensure the safety and normal operation of the park.

Optionally, referring to FIG. 4 , judging whether the inspection target has a position abnormality based on the position information and the face information includes:

S1303, when it is determined that the location information is within the designated abnormal detection area, the facial information is compared with the data of the authorized list of persons in the designated abnormal detection area;

S1304: When the facial information does not match the data of the person on the authorized list, it is determined that the inspection target has a positional abnormality.

In the process of judging whether the inspection target has positional abnormality based on location information and facial information, the location information of the inspection target is first obtained through camera positioning or setting positioning technology. A series of areas that require special attention are pre-defined in the system or database. These areas are considered to be abnormal detection areas, such as areas containing sensitive equipment, important assets or restricted access. The location information of the inspection target is compared with the specified abnormal detection areas defined in the system to determine whether the inspection target is currently located in these areas. If the inspection target is located in the specified abnormal detection area, the system needs to obtain the target's facial information, such as the facial image of the inspection target captured by the camera.

The system also needs to prepare a list of authorized personnel data, which contains the facial features or identity information of the personnel allowed to enter the designated anomaly detection area. The acquired facial information is compared with the personnel data in the authorized list, such as using facial recognition technology to analyze and identify facial features and match them with records in the database.

If the facial information matches the personnel data in the authorization list, it means that the inspection target is authorized to enter the abnormal detection area, and there is no need to further determine the position abnormality. If the facial information does not match the personnel data in the authorization list, it means that the inspection target has entered the designated abnormal detection area without authorization. In this case, the system determines that the inspection target has a position abnormality. Once it is confirmed that the inspection target has a position abnormality, the system immediately issues an abnormal prompt, such as sounding an alarm, displaying abnormal information on the monitoring screen, and sending notifications to relevant personnel. At the same time, the system can also automatically take further response measures according to preset rules or processes, such as starting tracking cameras, recording abnormal videos, and notifying security personnel to go to the scene.

Through the above process, the system can accurately determine whether the inspection target has any position abnormalities based on location information and facial information, and take corresponding measures in a timely manner to maintain the safety and order of the park.

After determining that the inspection target has an abnormal position, the following steps are also performed:

Based on the data of the authorized personnel, the campus camera is called to determine the authorized personnel closest to the inspection target, and an unauthorized entry abnormality alarm is sent to the work terminal bound to the authorized personnel.

After determining that the inspection target has an abnormal position, the system can take a series of follow-up measures to respond and handle the situation in a timely manner.

Among them, the system determines the level of abnormal alarm according to the abnormal degree and potential risk of the inspection target's location. This helps authorized personnel understand the urgency of the situation and respond accordingly. Based on the personnel data on the authorized list and the camera network of the park, call the camera or camera combination closest to the inspection target. These cameras should be able to provide clear video images for subsequent analysis and identification of the nearest authorized personnel. Through the analysis of the camera video images and the personnel location information in the authorized list, the system can calculate the authorized personnel closest to the inspection target.

Once the authorized personnel closest to the inspection target are identified, the system immediately sends an unauthorized access exception alarm to the bound work terminal (such as mobile phones, tablet computers, intercoms, etc.). The alarm information contains necessary information such as the current location of the inspection target, abnormal details, and the authorized personnel's response suggestions. After receiving the alarm, the authorized personnel can respond quickly according to the instructions of the alarm information, such as going to the site to confirm the situation, communicating with the inspection target, and taking necessary control measures.

Through the above process, the system can quickly respond to abnormal position of the inspection target and send alarm information to the nearest authorized personnel to achieve fast and effective processing, improve the safety management level of the park, and reduce potential safety risks.

In the above, by calling the real-time monitoring video of the park camera, target detection is performed based on the real-time monitoring video, and the target position is determined according to the target detection result and the position of the corresponding park camera; the set inspection position is imported, the inspection route is planned based on the target position and the inspection position, and the route inspection is performed along the inspection route; when moving to the inspection position, the set parameter information of the inspection position is collected, and parameter anomaly detection and prompts are performed based on the set parameter information; when moving to the target position, the video image of the target position is collected, the inspection target is determined based on the video image, and the behavior information, face information and position information of the inspection target are identified, and it is determined whether the inspection target has behavioral abnormalities based on the behavior information and face information, and whether the inspection target has positional abnormalities based on the position information and face information, and abnormal prompts are given based on the detected behavior abnormalities and/or position abnormalities, and the park camera is called to collect the tracking video of the inspection target in real time, and the tracking video is reported to the system background. By adopting the above-mentioned technical means, it is possible to conduct park inspections in combination with inspection targets and inspection locations, and to detect and prompt abnormalities in the behavior and location of the inspection targets. This can improve the comprehensiveness of park inspections, ensure the standardized operations of park personnel, and optimize the park's operating results.

Embodiment 2:

Based on the above embodiments, FIG5 is a schematic diagram of the structure of an autonomous inspection and decision-making device of a park intelligent inspection robot provided in Embodiment 2 of the present application. Referring to FIG5, the autonomous inspection and decision-making device of the park intelligent inspection robot provided in this embodiment specifically includes:

The target determination module 21 is used to call the real-time monitoring video of the park camera, perform target detection based on the real-time monitoring video, and determine the target position according to the target detection result and the position of the corresponding park camera;

The route planning module 22 is used to import the set inspection position, plan the inspection route based on the target position and the inspection position, and move along the inspection route to perform route inspection;

The inspection module 23 is used to collect the set parameter information of the inspection position when moving to the inspection position, and perform parameter anomaly detection and prompt based on the set parameter information; when moving to the target position, collect the video image of the target position, determine the inspection target based on the video image, and identify the behavior information, face information and position information of the inspection target, judge whether the inspection target has behavioral abnormalities based on the behavior information and face information, judge whether the inspection target has positional abnormalities based on the position information and face information, and give abnormal prompts based on the detected behavior abnormalities and/or positional abnormalities, call the park camera to collect the tracking video of the inspection target in real time, and report the tracking video to the system background.

Specifically, the target position is determined according to the target detection result and the position of the corresponding park camera, including:

When it is determined according to the target detection result that the monitored target enters the designated inspection area, or when it is determined that the number of monitored targets reaches the set number threshold, the target position is determined based on the position of the monitored target and the corresponding park camera.

Specifically, the inspection route is planned based on the target location and the inspection location, including:

The electronic map of the park is called, and an inspection route is planned on the electronic map of the park based on a set path planning algorithm. The inspection route includes a target location and an inspection location.

Specifically, after planning the inspection route based on the target location and the inspection location, it also includes:

Acquire newly added target location information and/or target location change information, and update the inspection route based on the newly added target location information and/or target location change information.

Specifically, based on the behavior information and face information, it is judged whether the inspection target has abnormal behavior, including:

The specified detection behavior is compared based on the behavior information. If the behavior information matches the specified detection behavior, the whitelist face data pre-bound to the specified detection behavior is compared based on the face information. If the face information does not match the whitelist face data, it is judged that the inspection target has abnormal behavior.

Based on the location information and face information, determine whether the inspection target has any abnormal location, including:

When it is determined that the location information is within the specified abnormal detection area, the facial information is compared with the data of the authorized list of people in the specified abnormal detection area. If the facial information does not match the data of the authorized list of people, it is determined that there is a location abnormality in the inspection target.

Specifically, after determining that the inspection target has an abnormal position, the following steps are also included:

Based on the data of the authorized personnel, the campus camera is called to determine the authorized personnel closest to the inspection target, and an unauthorized entry abnormality alarm is sent to the work terminal bound to the authorized personnel.

In the above, by calling the real-time monitoring video of the park camera, target detection is performed based on the real-time monitoring video, and the target position is determined according to the target detection result and the position of the corresponding park camera; the set inspection position is imported, the inspection route is planned based on the target position and the inspection position, and the route inspection is performed along the inspection route; when moving to the inspection position, the set parameter information of the inspection position is collected, and parameter anomaly detection and prompts are performed based on the set parameter information; when moving to the target position, the video image of the target position is collected, the inspection target is determined based on the video image, and the behavior information, face information and position information of the inspection target are identified, and it is determined whether the inspection target has behavioral abnormalities based on the behavior information and face information, and whether the inspection target has positional abnormalities based on the position information and face information, and abnormal prompts are given based on the detected behavior abnormalities and/or position abnormalities, and the park camera is called to collect the tracking video of the inspection target in real time, and the tracking video is reported to the system background. By adopting the above-mentioned technical means, it is possible to conduct park inspections in combination with inspection targets and inspection locations, and to detect and prompt abnormalities in the behavior and location of the inspection targets. This can improve the comprehensiveness of park inspections, ensure the standardized operations of park personnel, and optimize the park's operating results.

The autonomous inspection and decision-making device of the park intelligent inspection robot provided in Example 2 of the present application can be used to execute the autonomous inspection and decision-making method of the park intelligent inspection robot provided in Example 1 above, and has corresponding functions and beneficial effects.

Embodiment three:

Embodiment 3 of the present application provides an electronic device, referring to FIG6 , the electronic device includes: a processor 31, a memory 32, a communication module 33, an input device 34, and an output device 35. The number of processors in the electronic device may be one or more, and the number of memories in the electronic device may be one or more. The processor, memory, communication module, input device, and output device of the electronic device may be connected via a bus or other means.

As a computer-readable storage medium, the memory can be used to store software programs, computer executable programs and modules, such as the program instructions/modules corresponding to the autonomous inspection and decision-making method of the park intelligent inspection robot described in any embodiment of the present application (for example, the various modules in the autonomous inspection and decision-making device of the park intelligent inspection robot). The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the device, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The communication module is used for data transmission.

The processor executes various functional applications and data processing of the device by running software programs, instructions and modules stored in the memory, thereby realizing the autonomous inspection and decision-making method of the above-mentioned park intelligent inspection robot.

The input device can be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device can include display devices such as display screens.

The electronic device provided above can be used to execute the autonomous inspection and decision-making method of the park intelligent inspection robot provided in the above embodiment 1, and has corresponding functions and beneficial effects.

Embodiment 4:

The embodiment of the present application also provides a storage medium containing computer executable instructions, which are used to execute an autonomous inspection and decision-making method for a park intelligent inspection robot when executed by a computer processor. The autonomous inspection and decision-making method for the park intelligent inspection robot includes: calling the real-time monitoring video of the park camera, performing target detection based on the real-time monitoring video, and determining the target position according to the target detection result and the position of the corresponding park camera; importing the set inspection position, planning the inspection route based on the target position and the inspection position, and moving along the inspection route to perform route inspection; when moving to the inspection position, collecting the set parameter information of the inspection position, and performing parameter abnormality detection and prompting based on the set parameter information; when moving to the target position, collecting the video image of the target position, determining the inspection target based on the video image, and identifying the behavior information, face information and position information of the inspection target, judging whether the inspection target has abnormal behavior based on the behavior information and face information, judging whether the inspection target has abnormal position based on the position information and face information, and giving abnormal prompts based on the detected behavior abnormality and/or position abnormality, calling the park camera to collect the tracking video of the inspection target in real time, and reporting the tracking video to the system background.

Storage medium - any of various types of memory devices or storage devices. The term "storage medium" is intended to include: installation media, such as CD-ROM, floppy disk or tape device; computer system memory or random access memory, such as DRAM, DDR RAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory, such as flash memory, magnetic media (such as hard disk or optical storage); registers or other similar types of memory elements, etc. Storage media may also include other types of memory or combinations thereof. In addition, the storage medium may be located in the first computer system in which the program is executed, or may be located in a different second computer system, which is connected to the first computer system via a network (such as the Internet). The second computer system can provide program instructions to the first computer for execution. The term "storage medium" may include two or more storage media residing in different locations (for example, in different computer systems connected by a network). The storage medium may store program instructions (for example, embodied as a computer program) that can be executed by one or more processors.

Of course, the storage medium containing computer executable instructions provided in the embodiment of the present application is not limited to the autonomous inspection and decision-making method of the park intelligent inspection robot as described above, and can also execute related operations in the autonomous inspection and decision-making method of the park intelligent inspection robot provided in any embodiment of the present application.

The autonomous inspection and decision-making device, storage medium and electronic device of the park intelligent inspection robot provided in the above embodiments can execute the autonomous inspection and decision-making method of the park intelligent inspection robot provided in any embodiment of the present application. For technical details not described in detail in the above embodiments, please refer to the autonomous inspection and decision-making method of the park intelligent inspection robot provided in any embodiment of the present application.

The above are only preferred embodiments of the present application and the technical principles used. The present application is not limited to the specific embodiments described herein, and various obvious changes, readjustments and substitutions that can be made by those skilled in the art will not deviate from the scope of protection of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the claims.

Claims (10)

1. An autonomous patrol and decision method of an intelligent patrol robot in a park is characterized by comprising the following steps:

Invoking a real-time monitoring video of the park camera, performing target detection based on the real-time monitoring video, and determining a target position according to a target detection result and the position of the corresponding park camera;

Importing a set routing inspection position, planning a routing inspection route based on the target position and the routing inspection position, and moving along the routing inspection route to perform route routing inspection;

When the mobile terminal moves to the inspection position, acquiring set parameter information of the inspection position, and detecting and prompting parameter abnormality based on the set parameter information;

when the target position is moved, a video image of the target position is acquired, a patrol target is determined based on the video image, behavior information, face information and position information of the patrol target are identified, whether the patrol target is abnormal in behavior is judged based on the behavior information and the face information, whether the patrol target is abnormal in position is judged based on the position information and the face information, abnormal prompt is carried out based on the detected abnormal behavior and/or abnormal position, a park camera is called to acquire tracking video of the patrol target in real time, and the tracking video is reported to a system background.

2. The autonomous patrol and decision-making method for a campus intelligent patrol robot according to claim 1, wherein said planning a patrol route based on said target location and said patrol location comprises:

And calling a park electronic map, and planning a routing inspection route on the park electronic map based on a set path planning algorithm, wherein the routing inspection route comprises the target position and the routing inspection position.

3. The autonomous patrol and decision-making method of the intelligent patrol robot for a campus as recited in claim 1, wherein the determining the target position according to the target detection result and the position of the corresponding camera for the campus comprises:

and determining the position of the monitoring target based on the positions of the monitoring target and the cameras of the corresponding park under the condition that the monitoring target enters the defined inspection area according to the target detection result or the number of the monitoring target reaches the set number threshold.

4. The autonomous patrol and decision-making method for a campus intelligent patrol robot according to claim 1, further comprising, after said planning a patrol route based on said target location and said patrol location:

And acquiring new target position information and/or target position change information, and updating the routing inspection route based on the new target position information and/or the target position change information.

5. The autonomous patrol and decision-making method of a campus intelligent patrol robot according to claim 1, wherein said determining whether the patrol target is behaviorally abnormal based on the behavioural information and the face information comprises:

And comparing the specified detection behaviors based on the behavior information, comparing the pre-bound white-name single-person face data of the specified detection behaviors based on the face information under the condition that the behavior information is matched with the specified detection behaviors, and judging that the inspection target is abnormal under the condition that the face information is not matched with the white-name single-person face data.

6. The autonomous patrol and decision-making method of a campus intelligent patrol robot according to claim 1, wherein said determining whether the patrol target is abnormal based on the position information and the face information comprises:

and under the condition that the position information is determined to be in the appointed abnormal detection area, comparing the face information with the authorized list personnel data of the appointed abnormal detection area, and under the condition that the face information is not matched with the authorized list personnel data, judging that the position of the inspection target is abnormal.

7. The autonomous patrol and decision-making method for a intelligent patrol robot for a campus of claim 6, further comprising, after said determining that the patrol target is abnormal in position:

And calling a park camera based on the authorized list personnel data, determining authorized personnel closest to the patrol target, and sending an unauthorized entering abnormal alarm to a working terminal bound by the authorized personnel.

8. Autonomous patrol and decision-making device of intelligent patrol robot in park, which is characterized by comprising:

The target determining module is used for calling the real-time monitoring video of the park camera, detecting the target based on the real-time monitoring video and determining the target position according to the target detection result and the position of the corresponding park camera;

the route planning module is used for importing a set routing inspection position, planning a routing inspection route based on the target position and the routing inspection position, and moving along the routing inspection route to carry out route routing inspection;

The inspection module is used for collecting set parameter information of the inspection position when moving to the inspection position, and carrying out parameter abnormality detection and prompt based on the set parameter information; when the target position is moved, a video image of the target position is acquired, a patrol target is determined based on the video image, behavior information, face information and position information of the patrol target are identified, whether the patrol target is abnormal in behavior is judged based on the behavior information and the face information, whether the patrol target is abnormal in position is judged based on the position information and the face information, abnormal prompt is carried out based on the detected abnormal behavior and/or abnormal position, a park camera is called to acquire tracking video of the patrol target in real time, and the tracking video is reported to a system background.

9. An electronic device, comprising:

A memory and one or more processors;

The memory is used for storing one or more programs;

When the one or more programs are executed by the one or more processors, the one or more processors implement the autonomous patrol and decision-making method of the intelligent patrol robot of a campus as recited in any one of claims 1-7.

10. A storage medium containing computer executable instructions which, when executed by a computer processor, are for performing the autonomous inspection and decision making method of the intelligent inspection robot of a campus as claimed in any one of claims 1 to 7.