Dynamic generation and early warning method and system of unmanned aerial vehicle electronic fence

Abstract

This invention provides a method and system for dynamic generation and early warning of UAV electronic fences, relating to the field of electronic fence technology. The method includes: obtaining image acquisition quality parameters of the UAV based on feature points of the target area and images acquired by the UAV in real time; performing dynamic evaluation of sampling accuracy based on the task priority level of the electronic fence generation command and the image acquisition quality parameters of the UAV, and performing dynamic adjustment of sampling configuration based on the task requirement parameters and sensor performance parameters of the electronic fence generation command according to the dynamic evaluation results; performing dynamic adjustment of the update speed based on the relationship between the UAV flight speed and the update speed of each boundary segment according to the task priority level of the electronic fence generation command; and performing UAV electronic fence generation and early warning based on the dynamic adjustment results of the sampling configuration and the dynamic adjustment results of the boundary update speed, thus achieving adaptive dynamic matching between the electronic fence boundary update speed and the real-time flight status of the UAV.

Description

Technical Field

[0001] This invention relates to the field of electronic fence technology, and in particular to a method and system for dynamic generation and early warning of electronic fences for drones. Background Technology

[0002] The booming development of the low-altitude economy has driven the continuous expansion of drone application scenarios. At the same time, the unauthorized and unregulated "black flight" of drones poses a significant threat to aviation safety, public safety, and national security. Drone electronic fence technology has become a core technical solution in the field of drone airspace management under this real need.

[0003] The generation of existing drone electronic fences is based on high-precision satellite positioning systems, assisted positioning technology, and geographic information databases. By inputting the precise coordinates and altitude thresholds of no-fly zones and restricted flight zones, virtual flight boundaries in the form of polygons, circles, etc. are constructed, and dynamic algorithms and real-time environmental information are combined to dynamically adjust the boundaries. Its early warning and intervention functions rely on the drone positioning module to collect flight data such as position and speed in real time, which is transmitted to the control center via a communication link and compared at high speed with preset boundary parameters. Once a drone is detected approaching or entering a restricted area, the system will trigger audible and visual warnings and text warnings according to the risk level gradient, or directly link the flight control system to implement control operations such as speed limiting, hovering, automatic return to home, and forced landing, thereby achieving precise and timely airspace control.

[0004] For example, the patent application publication number CN119789041A discloses an electronic fence monitoring method and system for drones, which includes: based on the real-time positioning data of the drone, calculating the drone's flight speed by analyzing the drone's positioning changes over a period of time, and analyzing the drone's flight speed change trend based on the drone's flight speed data over a period of time to obtain drone speed information.

[0005] For example, patent application CN120640237A discloses a method and system for constructing an aerial three-dimensional electronic fence for drones, which includes: dividing the area boundary of the electronic fence within a preset geographical range based on geographic information system and three-dimensional modeling technology, and calculating the range of the electronic fence; integrating the configured electronic fence into the drone system, establishing a real-time communication and data exchange channel with the drone, and receiving the location information and flight status data fed back by the drone during flight in real time; monitoring the flight status of the drone and the area boundary of the electronic fence in real time, and triggering an early warning mechanism when the drone is found to be crossing the boundary.

[0006] The above-mentioned technology has at least the following technical problems:

[0007] In existing technologies, the update speed of electronic fence boundaries is usually fixed. When drones fly at high speeds or make sudden maneuvers (such as emergency turns), the fixed update speed may cause the generation calculation of the fence boundary to lag significantly behind the real-time pose changes of the drone, resulting in the system failing to trigger the necessary warnings and causing the security protection to fail. Summary of the Invention

[0008] In view of this, embodiments of the present invention provide a method and system for dynamic generation and early warning of UAV electronic fences, which realizes adaptive dynamic matching between the update speed of electronic fence boundaries and the real-time flight status of UAVs, ensuring that the fence is always synchronized with the UAV's pose, thereby eliminating early warning loopholes caused by update lag, and achieving accurate and reliable real-time dynamic protection.

[0009] This invention provides a method for dynamic generation and early warning of UAV electronic fences, comprising: receiving an electronic fence generation command and obtaining task requirement parameters of the command, the task requirement parameters including task coverage requirements and task priority level; dynamically dividing a target area based on the task coverage requirements; obtaining image acquisition quality parameters of the UAV in real time based on feature points of the target area and images acquired by the UAV, the image acquisition quality parameters including image matching rate, target area feature point deviation, and feature point repetition rate; performing dynamic evaluation of sampling accuracy based on the task priority level of the electronic fence generation command and the image acquisition quality parameters of the UAV; and, based on the dynamic evaluation result of sampling accuracy, determining the target area based on the task requirement parameters of the electronic fence generation command and sensor data. The system dynamically adjusts the sensor performance parameters based on sampling configuration. Sensor performance parameters include sensor configuration and sensor latency, while sampling configuration includes sampling frequency and sampling accuracy. The system divides the electronic fence into boundary segments based on preset rules. It determines whether dynamic adjustment of the update speed is needed based on the task priority level of the electronic fence generation command and the relationship between the UAV's flight speed and the update speed of each boundary segment. If needed, the boundary update speed is dynamically adjusted based on the boundary update configuration of each boundary segment; otherwise, the current boundary update speed is maintained. The boundary update configuration includes boundary point update unit, boundary point density, and boundary priority level. Based on the results of the dynamic adjustment of sampling configuration and the dynamic adjustment of boundary update speed, the system generates and issues warnings for the UAV electronic fence.

[0010] This application also provides a UAV electronic fence dynamic generation and early warning system. This system is applied to the UAV electronic fence dynamic generation and early warning method and includes: an image quality assessment module, a sampling configuration control module, an update speed control module, and a fence generation early warning module. The image quality assessment module receives electronic fence generation instructions and obtains the task requirement parameters of the instructions, including task coverage requirements and task priority levels. Based on the task coverage requirements, it dynamically divides the target area and obtains the UAV's image acquisition quality parameters in real time based on the target area feature points and the UAV-acquired images. These parameters include image matching rate, target area feature point deviation, and feature point repetition rate. The sampling configuration control module performs dynamic evaluation of sampling accuracy based on the task priority level of the electronic fence generation instructions and the UAV's image acquisition quality parameters. Based on the dynamic evaluation results of sampling accuracy, the sampling configuration is dynamically adjusted according to the task requirement parameters and sensor performance parameters of the electronic fence generation command. The sensor performance parameters include sensor configuration and sensor latency, and the sampling configuration includes sampling frequency and sampling accuracy. The update speed adjustment module is used to divide the boundary segments of the electronic fence. Based on the task priority level of the electronic fence generation command, it determines whether dynamic adjustment of the update speed is needed based on the relationship between the UAV flight speed and the update speed of each boundary segment. If needed, the boundary update speed is dynamically adjusted based on the boundary update configuration of each boundary segment. If not needed, the current boundary update speed is maintained. The boundary update configuration includes boundary point update unit, boundary point density, and boundary priority level. The fence generation early warning module is used to generate and warn of UAV electronic fences based on the dynamic adjustment results of sampling configuration and dynamic adjustment results of boundary update speed.

[0011] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0012] 1. This invention establishes a complete closed-loop control system driven by task requirements, with multi-source perception feedback and dynamic resource allocation. This enables global, collaborative, and adaptive optimization of the UAV sensor sampling strategy and electronic fence boundary update mechanism. Consequently, the electronic fence is upgraded from a static geographical restriction tool to an intelligent flight safety management and decision support system that can autonomously adjust monitoring accuracy, update rate, and response strategy based on task urgency, environmental complexity, and real-time UAV dynamics. This maximizes the overall resource utilization efficiency and task adaptability of the system while ensuring that core security protection is free of vulnerabilities.

[0013] 2. This invention constructs a two-layer dynamic evaluation mechanism that combines a minimum sampling standard based on task priority with a comprehensive quantitative evaluation based on multi-dimensional image quality parameters. Based on the evaluation results, it triggers differentiated, step-by-step control strategies ranging from single-frequency fine-tuning to multi-parameter joint calibration. This achieves accurate, efficient, and adaptive configuration of UAV sensor acquisition resources, thereby ensuring the reliability of electronic fence-generated data and avoiding excessive resource consumption in unnecessary scenarios, achieving an optimal balance between safety and efficiency.

[0014] 3. This invention introduces a linkage adjustment mechanism between flight altitude and sampling density, which is constrained by the mission safety baseline and driven by real-time image quality feedback. This mechanism can automatically and collaboratively optimize the two key dimensions of spatial position and information acquisition intensity when the sampling accuracy is severely insufficient, and ensure that all adjustments are not lower than the mission's preset minimum performance standard. This enables the system to autonomously and dynamically improve its perception capabilities in complex or harsh operating environments, restore and maintain the sampling accuracy at a qualified level that meets the requirements for generating electronic fences, and effectively address the problem of perception quality degradation caused by environmental changes.

[0015] 4. This invention introduces a quantization mapping relationship based on the inherent delay of the sensor to dynamically adjust the sampling frequency to compensate for the time lag in data acquisition. It also uses the minimum requirements of the task and the maximum capabilities of the hardware as rigid constraints to limit the adjustment results. This achieves the optimal match between sampling frequency and data timeliness within the hardware performance boundary, thereby ensuring that the sensing data on which the electronic fence generation depends has both high real-time performance and high reliability, laying a solid data foundation for subsequent dynamic boundary generation and accurate early warning.

[0016] 5. This invention establishes a mechanism that uses task priority as a benchmark, matches the difference between the UAV's flight status and the boundary update speed in real time, and performs personalized acceleration control on different boundary segments based on their structural attributes and importance. This allows for the accurate identification of lagging boundary segments and targeted acceleration when the UAV is maneuvering at high speed, while always keeping the control within the system's capacity. This achieves real-time, accurate, and reliable synchronous updates of the electronic fence boundary and the UAV's dynamic flight trajectory, completely eliminating the risk of early warning failure caused by update delays.

[0017] 6. This invention integrates the dynamically optimized perception and boundary update capabilities from the preceding steps, and triggers a three-tiered progressive intelligent response strategy based on the real-time positional relationship between the UAV and the fence boundary. This strategy progresses from "routine monitoring" to "high-intensity evidence collection and emergency hovering" and then to "active heading correction." This constructs a closed-loop system that integrates high-precision dynamic fence generation, real-time status synchronization, and tiered active safety intervention. As a result, it achieves pre-warning, in-process evidence collection, and immediate intervention for potential UAV boundary crossings or intrusions. While ensuring absolute priority for flight safety, it maximizes the automation and intelligence of the entire monitoring process. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method for dynamic generation and early warning of UAV electronic fences provided in an embodiment of the present invention;

[0019] Figure 2 This is a flowchart illustrating the dynamic adjustment of sampling accuracy in the UAV electronic fence dynamic generation and early warning method provided in this embodiment of the invention.

[0020] Figure 3 This is a schematic diagram of the structure of the UAV electronic fence dynamic generation and early warning system provided in the embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0022] In this application, the terms "first," "second," "third," etc., are used to distinguish identical or similar items with substantially the same function and purpose. It should be understood that there is no logical or temporal dependency between "first," "second," and "nth," nor does it limit the quantity or execution order. It should also be understood that although the following description uses the terms "first," "second," etc., to describe various elements, these elements should not be limited by the terms. These terms are merely used to distinguish one element from another. For example, "first device," "second device," "third device," etc., are only used to distinguish devices. Similarly, "first sample data," "second sample data," and "third sample data," etc., are only used to distinguish sample data. Without departing from the scope of the various examples, a first device can be referred to as a second device, and similarly, a second device can be referred to as a first device. Both the first device and the second device are devices, and in some cases, they can be separate and distinct devices.

[0023] It should be understood that the terminology used in the description of the various examples herein is for the purpose of describing the particular examples only and is not intended to be limiting. As used in the description of the various examples and in the appended claims, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0024] It should also be understood that obtaining B from A does not mean determining B solely from A; B can also be determined from A and / or other information.

[0025] It should also be understood that the term “comprising” (also referred to as “includes”, “including”, “comprises” and / or “comprising”) as used in this specification specifies the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0026] It should also be understood that the term "if" can be interpreted as meaning "when" or "upon" or "in response to determination" or "in response to detection." Similarly, depending on the context, the phrases "if determination..." or "if detection [the stated condition or event]" can be interpreted as meaning "when determination..." or "in response to determination..." or "when detection [the stated condition or event]" or "in response to detection [the stated condition or event]."

[0027] This invention provides a method for dynamically generating and issuing early warnings for unmanned aerial vehicle (UAV) electronic fences. For example... Figure 1The flowchart shown illustrates a method for dynamic generation and early warning of drone electronic fences. This method's processing flow may include the following steps: receiving an electronic fence generation command and obtaining the task requirement parameters, including task coverage requirements and task priority levels; dynamically dividing the target area based on the task coverage requirements (task coverage requirements include the geographical range of the target area (latitude and longitude coordinates), coverage type (full area coverage / key area coverage), coverage duration (temporary fence / long-term fence), etc.); and obtaining the drone's image acquisition quality parameters in real time based on feature points of the target area and images acquired by the drone. This includes image matching rate, target area feature point deviation, and feature point repetition rate. Image matching rate represents the percentage of matches between preset feature points (such as landmarks and boundary inflection points) within the target area and corresponding points in the drone-collected images. A higher matching rate indicates more accurate capture of regional features. Target area feature point deviation represents the average deviation between the geographic coordinates (after pixel coordinate conversion) and the actual coordinates of each feature point with known precise coordinates in the target area. A smaller deviation indicates higher sampling accuracy. Feature point repetition rate is the average proportion of feature points in the target area that are repeatedly extracted in different acquisition perspectives and different frame images. For example, if a feature point is stably extracted 8 times in 10 consecutive acquired images, the repetition rate is 80%. A higher repetition rate indicates stronger stability in feature point acquisition and reduces sampling fluctuations caused by changes in perspective. Based on the task priority level of the electronic fence generation command and the image acquisition quality parameters of the UAV, dynamic evaluation of sampling accuracy is performed. Based on the dynamic evaluation results, dynamic adjustment of sampling configuration is performed based on the task requirement parameters of the electronic fence generation command and sensor performance parameters. Sensor performance parameters include sensor configuration and sensor latency, while sampling configuration includes sampling frequency and sampling accuracy. The electronic fence is divided into boundary segments based on preset rules. Based on the task priority level of the electronic fence generation command and the relationship between the UAV's flight speed and the update speed of each boundary segment, it is determined whether dynamic adjustment of the update speed is necessary. If necessary, the boundary update speed is dynamically adjusted based on the boundary update configuration of each boundary segment; otherwise, the current boundary update speed is maintained. The boundary update configuration includes the boundary point update unit, boundary point density, and boundary priority level. The boundary point update unit is the number of boundary points that need to be updated in each boundary segment during a single update cycle. The boundary point density is the number of boundary points contained within a unit length of each boundary segment. The priority coefficient is a quantitative value set based on the electronic fence task coverage requirements, representing the importance of each boundary segment update, with a value range of [0.1, 1.0], where higher priority results in larger values. Based on the dynamic adjustment results of sampling configuration and boundary update speed, the generation and early warning of UAV electronic fences are performed.

[0028] In this embodiment, the method achieves adaptability, reliability, and timely early warning in electronic fence generation through dynamic design and precise control throughout the entire process. Guided by task requirements, it first receives electronic fence generation instructions and extracts key parameters such as task coverage requirements and task priority levels. The task coverage requirements specify the latitude and longitude coordinate range of the target area, the coverage type (full area or key area), and the coverage duration (temporary or long-term fence). Based on this, dynamic and precise division of the target area can be achieved, ensuring that fence generation highly matches actual task requirements and avoiding resource waste caused by ineffective area coverage. Simultaneously, by acquiring drone images in real time and extracting image acquisition quality parameters such as image matching rate, target area feature point deviation, and feature point repetition rate, the image matching rate reflects the matching ratio between preset landmarks, boundary inflection points, and corresponding points in the acquired image. This directly characterizes the accuracy of regional feature capture; a higher matching rate indicates a higher degree of feature restoration in the sampled area. The deviation of feature points in the target area is quantified by comparing the average deviation between the geographic coordinates of known precise coordinate feature points after pixel coordinate transformation in the sampled image and their actual coordinates. The smaller the deviation, the more reliable the accuracy of the basic data for fence generation. The feature point repetition rate reflects the average proportion of each feature point that is repeatedly extracted in different acquisition views and different frame images. The higher the repetition rate, the stronger the stability of the sampled data, which can effectively avoid sampling fluctuations caused by changes in viewpoint and provide accurate data support for subsequent sampling control. Dynamic evaluation of sampling accuracy is carried out by combining task priority level and image acquisition quality parameters. Based on the evaluation results and the integration of sensor performance parameters such as sensor configuration and sensor delay, sampling configurations such as sampling frequency and sampling accuracy are dynamically controlled to achieve optimal allocation of sampling resources. When there are high-priority tasks or insufficient acquisition quality, sampling accuracy and frequency are increased to ensure data quality. When there are low-priority tasks or good acquisition quality, sampling resource consumption is reasonably reduced to balance accuracy and efficiency. The electronic fence boundary is segmented according to preset rules. Based on the task priority level and the matching relationship between the drone's flight speed and the update speed of each boundary segment, it is determined whether to perform dynamic adjustment of the boundary update speed. If adjustment is required, the boundary update speed is precisely controlled based on boundary update configurations such as boundary point update unit, boundary point density, and boundary priority level. This ensures that the boundary update is adapted to the drone's flight status and task priority. High-priority boundary segments can obtain a faster update speed to ensure the real-time performance of the fence boundary, while low-priority segments can maintain a reasonable update speed to balance resource consumption. If no adjustment is required, the current update speed is maintained, further improving the operational efficiency of the method.This invention achieves dynamic generation and early warning of UAV electronic fences based on the dual results of dynamic control of sampling configuration and dynamic control of boundary update speed. This full-process dynamic control mechanism enables the electronic fence to adapt to changes in task requirements, UAV flight status, and environmental sampling conditions in real time, effectively improving the flexibility and accuracy of electronic fence generation. It ensures that the fence boundary always remains consistent with the target area and task requirements, thereby guaranteeing the timeliness and reliability of the early warning mechanism. It can be widely adapted to UAV operation tasks with different coverage types, different time requirements, and different priorities, providing efficient and accurate technical support for UAV flight safety management.

[0029] like Figure 2 The flowchart shown illustrates the dynamic adjustment of sampling accuracy in a drone-based electronic fence dynamic generation and early warning method. The steps for dynamically evaluating sampling accuracy based on the task priority level of the electronic fence generation command and the drone's image acquisition quality parameters include: inputting the task priority level of the electronic fence generation command into a preset sampling accuracy threshold mapping table to obtain the corresponding minimum sampling accuracy threshold. The sampling accuracy threshold mapping table contains a quantified correspondence between different task priority levels and the minimum sampling accuracy threshold; determining whether the current sampling frequency of the drone sensor has reached the minimum sampling accuracy threshold; if so, marking the dynamic evaluation result of sampling accuracy as qualified; otherwise, performing comprehensive processing based on the image acquisition quality parameters to obtain an image quality comprehensive index; determining whether the image quality comprehensive index has reached a preset image quality threshold; if so, marking the dynamic evaluation result of sampling accuracy as needing adjustment; otherwise, marking the dynamic evaluation result of sampling accuracy as unqualified.

[0030] In this embodiment, the present invention achieves a deep adaptation between sampling accuracy assessment and task requirements through a logical design of task priority-oriented threshold matching and hierarchical precise determination. It also considers the scientific nature of the assessment and the forward-looking nature of the control. Using task priority as the core anchor, the present invention directly obtains the corresponding minimum sampling accuracy threshold by inputting it into a preset sampling accuracy threshold mapping table. This mapping table has pre-established a quantitative correspondence between different task priority levels and the minimum sampling accuracy threshold, enabling precise matching of differentiated accuracy baseline requirements based on task importance. High-priority tasks correspond to higher minimum sampling accuracy thresholds, ensuring the quality of sampling data for critical tasks; low-priority tasks correspond to reasonable accuracy thresholds, avoiding resource redundancy caused by excessive pursuit of accuracy. Based on this, a basic accuracy screening is first conducted to determine whether the current sampling frequency of the UAV sensor has reached the minimum sampling accuracy threshold obtained from the above matching. If the current sampling frequency meets the standard, it means that the existing sampling configuration can meet the basic accuracy requirements of the current task, and the dynamic evaluation result of the sampling accuracy is directly marked as qualified without additional adjustment, ensuring evaluation efficiency. If the current sampling frequency does not meet the standard, a secondary precision judgment stage is entered. By comprehensively weighting image acquisition quality parameters such as image matching rate, target area feature point deviation, and feature point repetition rate, a comprehensive image quality index that can fully characterize the quality of the sampled data is generated. This index is then compared with the preset image quality threshold. If the comprehensive image quality index meets the standard, it means that although the current sampling frequency has not reached the minimum threshold, the quality of the existing sampled data can still support the task to a certain extent. Therefore, the evaluation result is marked as needing adjustment, providing a basis for subsequent fine-tuning of the sampling configuration. If the comprehensive image quality index does not meet the standard, it means that the quality of the existing sampled data can no longer meet the basic requirements of the task, and the evaluation result is marked as unqualified, requiring the initiation of high-intensity sampling configuration adjustment. This hierarchical evaluation logic avoids misjudgments caused by using sampling frequency as the sole criterion (such as over-adjustment when the sampling frequency is below standard but the data quality is good). It also achieves comprehensive control over sampling accuracy through supplementary evaluation of image quality parameters, ensuring the accuracy and reliability of the evaluation results. At the same time, it provides clear directional guidance for subsequent dynamic adjustment of sampling configuration, enabling the adjustment measures to accurately match the current sampling accuracy with task requirements, taking into account both task execution accuracy and resource utilization efficiency.

[0031] Furthermore, the comprehensive image quality index is obtained as follows: Preset image acquisition quality weighting parameters and reference parameters are obtained. The image acquisition quality weighting parameters include the weighting of image matching rate, the weighting of target region feature point deviation, and the weighting of feature point repetition rate. The image acquisition quality reference parameters include the critical value of image matching rate, the critical value of target region feature point deviation, and the critical value of feature point repetition rate. The weighting of image matching rate is used to weight the analysis results of the ratio of image matching rate to the critical value of image matching rate, resulting in the image matching rate influence index. The weighting of target region feature point deviation is used to weight the analysis results of the ratio of target region feature point deviation critical value to the target region feature point deviation, resulting in the target region feature point deviation influence index. The weighting of feature point repetition rate is used to weight the analysis results of the ratio of feature point repetition rate to the critical value of feature point repetition rate, resulting in the feature point repetition rate influence index. The image matching rate influence index, the target region feature point deviation influence index, and the feature point repetition rate influence index are coupled to obtain the comprehensive image quality index, providing a quantitative basis for result determination in scenarios where the sampling frequency does not reach the minimum threshold in the dynamic evaluation of sampling accuracy.

[0032] Specifically, the formula for obtaining the overall image quality index is as follows:

[0033] ;

[0034] In the formula This represents the overall image quality index. , and These represent the weight proportions of image matching rate, target region feature point deviation, and feature point repetition rate, respectively. , and These represent the image matching rate, target region feature point deviation, and feature point repetition rate, respectively. , and These represent the critical value for image matching rate, the critical value for target region feature point deviation, and the critical value for feature point repetition rate, respectively.

[0035] In this embodiment, image matching rate, target area feature point deviation, and feature point repetition rate constitute a progressive, multi-dimensional image quality evaluation system. Image matching rate is fundamental; it assesses whether the sensor can successfully capture a sufficient number of preset features, solving the problem of feature coverage integrity. Target area feature point deviation is crucial; based on successful matching, it quantifies the spatial accuracy of feature points, directly determining the accuracy of the generated electronic fence's geographical boundary. Feature point repetition rate is a guarantee; it measures the stable occurrence probability of features in consecutive frames or different viewpoints, reflecting the robustness and reliability of image acquisition under time-series and viewpoint changes. The advantage of comprehensively evaluating these three factors is that it breaks the limitations of traditional reliance on a single indicator, enabling a comprehensive and in-depth diagnosis of the perception system's state: a high matching rate but high deviation may indicate systematic errors in sensor calibration or positioning; a low repetition rate may mean that changes in lighting, occlusion, or rapid movement have led to perception instability. This comprehensive evaluation provides a refined decision-making basis for subsequent precise adjustments (such as adjusting flight altitude or changing sampling strategies), thereby ensuring that the generation of the electronic fence is always based on high-quality, reliable perception data. This invention achieves precise characterization of image acquisition quality through quantitative design involving preset parameter anchoring, weighted sub-indicators, and multi-dimensional coupling. It provides a scientific and objective basis for evaluating sampling accuracy in scenarios where the sampling frequency does not reach the minimum threshold. By standardizing proportion analysis and assigning differentiated weights, it achieves comprehensive and accurate quantification of image acquisition quality, effectively avoiding the one-sidedness of single-indicator evaluation. At the same time, the generated comprehensive index provides a clear quantitative basis for judging the evaluation results in scenarios where the sampling frequency does not reach the minimum threshold corresponding to the task priority level. This makes the distinction between adjustment and non-compliance more objective and convincing, laying a solid foundation for the precise implementation of subsequent dynamic control of sampling configuration and further ensuring the adaptability of sampling accuracy to task requirements.

[0036] Furthermore, the steps for dynamically adjusting the sampling configuration include: if the dynamic evaluation result of the sampling accuracy is qualified, no additional processing is performed; if the dynamic evaluation result of the sampling accuracy is to be adjusted, then dynamic adjustment of the sampling frequency is performed based on the sensor configuration and sensor delay. Specifically, the task priority level is input into a preset sampling frequency threshold mapping table to obtain the corresponding minimum sampling frequency threshold. The sampling frequency threshold mapping table has a quantitative correspondence between different task priority levels and the minimum sampling frequency threshold. Sensor delay refers to the time deviation between the sensor completing image acquisition and the UAV acquiring the acquired image. The maximum sampling frequency threshold and the sensor focal length threshold are obtained based on the sensor configuration. When initializing or loading a specific UAV platform, the system will preset or automatically read its sensor hardware specification database. For the maximum sampling frequency threshold, the system obtains the value from the database based on the maximum physical frame rate specified by the sensor model and the maximum effective sampling rate that can be stably supported after comprehensively considering the data bus bandwidth and the processor's image processing capabilities, and sets it as an insurmountable hardware performance upper limit. For the sensor focal length threshold, the system calculates and sets a lower limit value of the focal length required to ensure basic monitoring functions based on the physically adjustable focal length range of the sensor's optical components (from wide-angle to telephoto) and the minimum field of view coverage or maximum permissible recognition error required by the mission safety principle. These two thresholds together constitute the inviolable hardware performance boundaries and safe operating baselines when adjusting the sampling frequency and focal length during subsequent dynamic control. The sensor delay is input to a preset sampling frequency adjustment mapping table to obtain the corresponding sampling frequency adjustment value. Based on the sampling frequency adjustment value, the current sampling frequency of the UAV sensor is processed to obtain the adjusted sampling frequency. The sampling frequency adjustment mapping table has a quantitative correspondence between different sensor delay ranges and sampling frequency adjustment values. If the adjusted sampling frequency is within the range of the minimum and maximum sampling frequency thresholds, the sensor sampling frequency is adjusted to the adjusted sampling frequency. If the adjusted sampling frequency is lower than the minimum sampling frequency threshold, the sensor sampling frequency is adjusted to the minimum sampling frequency threshold. If the adjusted sampling frequency is higher than the maximum sampling frequency threshold, the sensor sampling frequency is adjusted to the maximum sampling frequency threshold. If the dynamic evaluation result of the sampling accuracy is unqualified, the sampling frequency is dynamically adjusted based on the sensor configuration and sensor delay, and the sampling accuracy is dynamically adjusted based on the image acquisition quality parameters and task requirement parameters.

[0037] In this embodiment, the present invention employs a hierarchical response to evaluation results and a precise adaptation control logic for sensor parameters. This achieves dynamic matching between sampling configuration and task requirements and sensor performance, ensuring that sampling quality meets task requirements while avoiding redundant resource consumption. The present invention uses the dynamic evaluation result of sampling accuracy as the core basis to implement a differentiated control strategy: if the evaluation result is qualified, it indicates that the current sampling configuration has fully adapted to the accuracy requirements corresponding to the task priority level, and no additional control is needed. This minimizes resource consumption caused by ineffective control and improves overall operating efficiency. If the evaluation result is to be adjusted, it indicates that the current sampling frequency has not reached the minimum threshold, but the image quality is acceptable. Dynamic control needs to be focused on the sampling frequency. First, the core parameter definitions and boundary thresholds are clarified. Sensor delay refers to the time from when the sensor completes image acquisition to when the drone acquires the image. Image temporal deviation directly affects sampling timeliness. Based on sensor configuration, the highest sampling frequency threshold and the sensor focal length threshold can be determined (the focal length threshold is an auxiliary constraint for sampling frequency adjustment, ensuring the basic sharpness of the acquired image). Then, the sensor delay is input into a preset sampling frequency adjustment mapping table. The current sampling frequency is corrected using this adjustment value to obtain the adjusted sampling frequency. Finally, the final calibration is performed by comparing it with the lowest and highest sampling frequency thresholds: if the adjusted frequency is between the two thresholds, the image is directly acquired at this frequency to ensure a balance between sampling efficiency and accuracy; if it is below the lowest threshold, the lowest threshold is used as the final sampling frequency to ensure a minimum level of sampling accuracy; if it is above the highest threshold, the highest threshold is used as the final sampling frequency to avoid exceeding the sensor's performance limit, which could lead to acquisition failure or data distortion. If the evaluation result is unsatisfactory, it indicates that simply adjusting the sampling frequency cannot meet the task accuracy requirements. Therefore, based on the aforementioned dynamic adjustment of the sampling frequency, further dynamic adjustment of sampling accuracy is carried out by combining the current quality shortcomings (image acquisition quality parameters) with the anchored core accuracy requirements (task requirement parameters). Sampling accuracy is improved by optimizing sampling resolution and feature point extraction algorithm parameters, ensuring that the final sampling configuration fully covers the task accuracy requirements. This hierarchical adjustment logic achieves accurate response to different sampling accuracy statuses and ensures the feasibility and safety of adjustment through deep adaptation of sensor parameters. It also avoids resource waste or insufficient accuracy caused by a one-size-fits-all approach, ensuring that the sampling configuration always maintains optimal adaptation to task requirements and sensor status, providing high-quality data support for the accurate generation of subsequent electronic fences.

[0038] Furthermore, the step of dynamically adjusting the sampling accuracy based on image acquisition quality parameters and task requirement parameters includes: inputting the task priority level into a preset sampling density threshold mapping table to obtain the corresponding minimum sampling density threshold, wherein the sampling density threshold mapping table has a quantitative correspondence between different task priority levels and the minimum sampling density threshold; if the image matching rate is lower than the image matching rate threshold, the difference between the image matching rate threshold and the image matching rate is marked as the image matching rate deviation value, and the image matching rate deviation value is input into a preset sensor focal length adjustment mapping table to obtain the corresponding sensor focal length adjustment amount, wherein the sensor focal length adjustment mapping table has a quantitative correspondence between different image matching rate deviation ranges and the sensor focal length adjustment amount; performing summation processing on the current sensor focal length based on the sensor focal length adjustment amount to obtain the sensor focal length to be determined; if the image matching rate is not lower than the image matching rate threshold, the current sensor focal length is recorded as the sensor focal length to be determined; determining whether the sensor focal length to be determined is lower than the sensor focal length threshold, and if so, adjusting the sensor focal length to the sensor focal length threshold. The difference between the sensor focal length threshold and the focal length of the sensor to be determined is marked as the focal length deviation value. Based on the focal length deviation value, a dynamic adjustment warning for the UAV's flight altitude is issued. By increasing the UAV's flight altitude, the physical coverage of a single frame image from the sensor is expanded, thereby indirectly improving the image feature matching rate and pulling the sampling accuracy from an "unacceptable" state back to an acceptable range. This is a composite control method that coordinates optical sensing and flight control. Otherwise, the sensor focal length is adjusted to the focal length of the sensor to be determined. The difference between the image quality threshold and the image quality comprehensive index is marked as the image quality deviation value. The image quality deviation value is input into a preset sampling density adjustment mapping table to obtain the corresponding sampling density adjustment ratio. Based on the sampling density adjustment ratio, the sensor sampling density is dynamically adjusted to obtain the sampling density of the sensor to be determined. It is then determined whether the sampling density of the sensor to be determined has reached the minimum sampling density threshold. If so, the sensor sampling density is adjusted to the sampling density of the sensor to be determined; otherwise, the sensor sampling density is adjusted to the minimum sampling density threshold to ensure that the sampling accuracy meets the minimum sampling accuracy threshold requirement of the task.

[0039] In this embodiment, when a dynamic altitude adjustment warning for a UAV is issued, the focal length deviation value is input into a preset altitude adjustment mapping table. This table defines the engineering-calibrated altitude adjustment amounts corresponding to different UAV types with different deviation ranges (for example, it is recommended to increase the altitude by 2 meters for every 0.1mm focal length deviation). The generated warning is essentially a structured control command, mainly including: command type, target altitude value, constraint parameters, and a summary of adjustment basis. The constraint parameters include upper and lower limits. The upper limit constraint must simultaneously comply with the maximum safe flight altitude determined by the UAV's performance, and the lower limit constraint must meet the minimum flight altitude determined by the task coverage requirements. Based on the suggested target altitude, the system applies these constraints to finally calculate a legal and executable target flight altitude. The summary of adjustment basis is associated with the insufficient image matching rate event that triggered this warning and the calculation process, and is used for system logs and status tracking. This invention employs a hierarchical optimization logic of prioritizing focal length adjustment, supplementing sampling density, imposing threshold constraints, and linking early warnings to precisely target and solve the problem of substandard sampling accuracy. It achieves deep adaptation of sampling accuracy to task requirements and sensor performance. First, based on task priority, a minimum sensor sampling density value is determined, anchoring a rigid baseline requirement for sampling accuracy and setting an insurmountable accuracy threshold for subsequent adjustments. Then, sensor focal length adjustment is performed on a case-by-case basis for the core quality indicator, image matching rate. If the image matching rate is lower than a preset threshold, the difference between the threshold and the actual value is calculated as the image matching rate deviation. This deviation is input into a preset sensor focal length adjustment mapping table to obtain the corresponding sensor focal length adjustment amount and reduce the current focal length to obtain the sensor focal length to be determined. This improves the image detail capture capability and enhances the feature point matching effect. If the image matching rate is not lower than the threshold, the current focal length is directly recorded as the sensor focal length to be determined. Next, a threshold check is performed on the sensor focal length to be determined. If its value is lower than the threshold, the sensor focal length is adjusted accordingly. If the focal length threshold of the sensor is too low, it indicates that excessive reduction in focal length may lead to image blurring. In this case, the focal length is adjusted to the sensor focal length threshold, and the difference between this threshold and the focal length to be determined is marked as the focal length deviation value. At the same time, a dynamic adjustment warning for the UAV's flight altitude is issued based on this deviation value. The image acquisition clarity is further optimized through coordinated adjustment of flight altitude. If the focal length to be determined is not lower than the sensor focal length threshold, it is directly set as the final focal length. The difference between the image quality threshold and the image quality comprehensive index is calculated as the image quality deviation value. This deviation value is input into a preset sampling density adjustment mapping table to obtain the corresponding sampling density adjustment ratio. The sensor sampling density is dynamically adjusted according to this ratio to obtain the sensor sampling density to be determined. Finally, the sampling density to be determined is checked against the minimum value. If it reaches the minimum value of the sensor sampling density, it is assigned as the final sampling density of the UAV sensor. If it does not reach the minimum value, the minimum value of the sensor sampling density is used as the final sampling density, thereby ensuring that the sampling accuracy can meet the minimum threshold requirement of the task sampling accuracy.This invention precisely locates and solves the core problems of insufficient image matching rate and substandard overall quality by optimizing focal length and sampling density in a layered manner. At the same time, it avoids exceeding the sensor performance limit by using threshold constraints and forms a multi-dimensional collaborative control mechanism through flight altitude warning linkage. This effectively improves the accuracy and feasibility of sampling precision control, ensures that the quality of sampling data can support the generation of high-precision electronic fences, and provides a solid data guarantee for subsequent dynamic updates and warnings of fence boundaries.

[0040] Furthermore, the steps for determining whether to perform dynamic adjustment of update speed include: inputting the task priority level of the electronic fence generation command into a preset speed matching mapping table to obtain a speed deviation threshold. The speed matching mapping table contains a quantitative correspondence between different task priority levels and speed deviation thresholds; marking the difference between the boundary update speed of each boundary segment and the UAV flight speed as the speed deviation of each boundary segment; comparing the speed deviation of each boundary segment with the speed deviation threshold. If the speed deviation of any boundary segment is lower than the speed deviation threshold, then the boundary segment is marked as needing to perform dynamic adjustment of update speed; if the speed deviation of any boundary segment is not lower than the speed deviation threshold, then it is marked as not needing to perform dynamic adjustment of update speed.

[0041] In this embodiment, the present invention employs a precise judgment logic that compares task priority anchoring thresholds with segmented speed deviations. This enables the scientific identification of boundary update speed control requirements, ensuring that control measures are precisely matched with task importance and UAV flight status. Guided by task priority levels, the invention directly obtains the corresponding speed deviation thresholds by inputting them into a preset speed matching mapping table. This mapping table pre-establishes a quantitative correspondence between different task priority levels and speed deviation thresholds. Higher priority tasks correspond to smaller speed deviation thresholds to strictly constrain the matching degree between boundary update speed and UAV flight speed, ensuring the real-time performance and accuracy of the fence boundary. Lower priority tasks correspond to larger speed deviation thresholds, reducing unnecessary control frequency and saving resources. Furthermore, by calculating the difference between the boundary update speed of each boundary segment and the current flight speed of the UAV, the specific speed deviation for each segment is obtained, quantitatively representing the degree of matching between the boundary update speed and flight speed for each segment. Subsequently, the speed deviation of each segment is compared one by one with the speed deviation threshold obtained above to form a comprehensive judgment result: if the speed deviation of any boundary segment is lower than the speed deviation threshold, it means that the boundary update speed of that segment can no longer keep up with the drone's flight speed. If no adjustment is made, the fence boundary will lag behind the drone's flight trajectory, thus affecting the effectiveness of fence control. Therefore, this segment is marked as needing dynamic adjustment of the update speed. If the speed deviation of all boundary segments is not lower than the speed deviation threshold, it means that the current boundary update speed of each segment can adapt to the drone's flight speed. No additional adjustment is needed, and the current update state can be maintained to ensure operational efficiency. This judgment logic achieves personalized adaptation of adjustment needs identification through differentiated threshold settings based on task priorities, avoiding the lack of adjustment of key segments or the over-adjustment of non-key segments caused by a one-size-fits-all judgment. At the same time, based on the accurate comparison of segment speed deviations, the boundary segments that need adjustment can be accurately located, providing a clear target for subsequent targeted boundary update speed adjustment. This ensures that the adjustment measures focus on core needs and effectively ensures that the electronic fence boundary always keeps in sync with the drone's flight status and task requirements, improving the real-time performance and reliability of fence generation.

[0042] Furthermore, the steps for dynamically adjusting the boundary update speed include: marking the difference between the UAV flight speed and the speed deviation threshold as the basic update speed; dynamically processing the boundary update configuration based on the boundary segment to obtain the dynamic adjustment coefficient of the boundary segment; multiplying the basic update speed using the dynamic adjustment coefficient of the boundary segment to obtain the preliminary adjustment speed; obtaining the maximum allowable update speed based on UAV performance. The system queries the preset UAV platform performance database or obtains the airborne system status parameters in real time, comprehensively considering core performance constraints such as the maximum instruction processing frequency of the flight control unit, the stable transmission bandwidth of the data link, and the real-time path planning and boundary rendering capabilities of the airborne computing module. Through a preset performance-speed mapping model, the system calculates the highest sustainable speed that can stably and reliably process electronic fence boundary update data under the current system load. This speed is the maximum allowable update speed, serving as a rigid safety upper limit to ensure the real-time performance and stability of the system. If the preliminary adjustment speed does not exceed the maximum allowable update speed, the preliminary adjustment speed is used as the adjusted boundary update speed for the boundary segment; if the preliminary adjustment speed exceeds the maximum allowable update speed, the maximum allowable update speed is used as the adjusted boundary update speed for the boundary segment.

[0043] In this embodiment, the present invention employs a multi-level control framework of basic speed anchoring, segmented dynamic adaptation, and performance threshold constraints. This framework achieves precise matching between the boundary update speed and the UAV flight status, task requirements, and equipment performance. It ensures the real-time nature of the fence boundary while mitigating the risk of exceeding the equipment's capacity. First, based on the speed deviation threshold between the UAV flight speed and the corresponding task priority level, the difference between the two is calculated to obtain the basic update speed. This speed serves as the baseline for boundary updates, ensuring that the update speed meets the core requirement of keeping up with the UAV's flight rhythm, thus fundamentally preventing the fence boundary from becoming disconnected from the actual flight trajectory due to update lag. Subsequently, a comprehensive quantitative processing is performed based on the boundary update configuration of the controllable boundary segments to generate a dynamic adjustment coefficient specific to each segment. Higher boundary priority levels and greater boundary point density result in larger adjustment coefficient values, meaning that high-priority, high-density boundary segments receive stronger speed boost weights, thus tilting update resources towards key segments. The boundary point update unit, by influencing the workload of a single update cycle, assists in calibrating the rationality of the adjustment coefficient. The dynamic adjustment coefficient is multiplied by the base update speed to obtain the initial adjustment speed, thus completing the personalized optimization of the base speed and enabling the update speed to accurately adapt to the differentiated needs of each segment. Finally, the maximum allowable update speed determined by the drone's performance parameters is introduced as a hard constraint to verify the initial adjustment speed: if the initial adjustment speed does not exceed the maximum allowable update speed, it means that the speed is within the drone's equipment capacity range, and it is directly used as the adjusted boundary update speed for that segment; if the initial adjustment speed exceeds the maximum allowable update speed, the maximum allowable update speed is used as the final update speed to avoid drone computing power overload, data transmission congestion, and other failures caused by excessively increasing the update speed. This control logic not only ensures the real-time baseline of the update through base speed anchoring, but also achieves segmented differentiated control through dynamic adjustment coefficients, and ensures the safety and feasibility of control through performance threshold constraints. It effectively improves the adaptability and reliability of the boundary update speed, ensuring that the electronic fence boundary can be dynamically adjusted in real time and accurately according to the drone's flight status, providing key support for the effectiveness of the subsequent early warning mechanism.

[0044] Furthermore, the dynamic adjustment coefficient is obtained as follows: The boundary priority level is input into a preset priority coefficient mapping table to obtain the corresponding boundary priority coefficient. The priority coefficient mapping table contains a quantitative correspondence between different boundary priority levels and boundary priority coefficients. Preset update weight percentage parameters and update weight reference parameters are obtained. The update weight percentage parameters include update unit weight percentage and boundary point density weight percentage, used to define the influence weights of boundary point update units (number of boundary points updated in a single round) and boundary point density (number of boundary points per unit length) on the adjustment coefficient. The update weight reference parameters include update unit critical values ​​and boundary point density critical values, serving as benchmarks for measuring whether the two indicators meet the basic update requirements, providing a unified reference for standardized percentage analysis. The update unit weight percentage is used to weight the percentage analysis results of boundary point update units and update unit critical values ​​to obtain the update unit influence index. The boundary point density weight percentage is used to weight the percentage analysis results of boundary point density and boundary point density critical values ​​to obtain the boundary point density influence index. The boundary priority coefficient is used to weight the coupling processing results of the boundary point density influence index and the update unit influence index to obtain the dynamic adjustment coefficient for that boundary segment.

[0045] Specifically, the formula for obtaining the dynamic adjustment coefficient is as follows:

[0046] ;

[0047] In the formula, This represents the dynamic adjustment coefficient. , and These represent the boundary priority coefficient, the update unit weight ratio, and the boundary point density weight ratio, respectively. and These represent the boundary point update unit and boundary point density, respectively. and These represent the critical value for updating the unit and the critical value for boundary point density, respectively.

[0048] In this embodiment, the coupling process is called the operation and coupling process. This invention achieves precise adaptation between the adjustment coefficient and the characteristics of boundary segments through priority anchoring, multi-dimensional weighting, and quantitative coupling design. This provides a scientific and objective quantitative basis for differentiated control of boundary update speed. Using boundary priority levels as the core guide, the coefficients are input into a preset priority coefficient mapping table to obtain the corresponding boundary priority coefficients. These coefficients accurately quantify the update importance of each boundary segment, providing a core basis for subsequent weight allocation. This invention eliminates the differences in the dimensions of different indicators through standardized proportion analysis, accurately matches the impact value of each indicator through differentiated weighting, and strengthens the control priority of key segments through the integration of boundary priority coefficients. This allows the dynamic adjustment coefficients to comprehensively and accurately represent the update needs and characteristics of each boundary segment. The generated quantitative coefficients provide core support for the personalized optimization of subsequent basic update speeds, ensuring that high-priority, high-density, and high-update-unit-demand boundary segments receive stronger speed adjustment weights, achieving deep adaptation between boundary update speed and segment characteristics, and further improving the accuracy and rationality of boundary update speed control.

[0049] Furthermore, based on the dynamic adjustment results of sampling configuration and boundary update speed, the steps for generating and issuing early warnings for drone electronic fences include: controlling the drone to acquire images and extract features from the target area according to the adjusted sampling frequency and sampling accuracy to generate boundary data for the electronic fence; dynamically generating and updating the electronic fence boundary in real time based on the boundary data and the adjusted boundary update speed for each boundary segment; and during electronic fence monitoring, monitoring the positional relationship between the drone and the electronic fence boundary in real time and performing corresponding operations according to different positional states, specifically including: triggering an electronic fence anomaly signal when the drone enters the electronic fence boundary area. The sampling frequency of the UAV sensor is directly adjusted to the highest sampling frequency threshold, and a hovering command containing the target hovering coordinates, hovering altitude, and stable hovering attitude requirements is sent to the UAV flight control system. When the UAV reaches the electronic fence boundary, an electronic fence warning signal is issued, the sampling frequency of the UAV sensor is adjusted to the highest sampling frequency threshold, and a direction vector perpendicular to the tangent of the electronic fence boundary and pointing outward from the area is calculated based on the current position of the UAV. A heading adjustment command containing the direction vector and a preset safety distance is generated and sent to the UAV flight control system. When the UAV does not reach the electronic fence boundary, the current sampling frequency and flight status of the UAV sensor are maintained.

[0050] In this embodiment, the present invention constructs a closed-loop control logic of data acquisition, boundary update, status monitoring, and differentiated response. By deeply integrating the sampling configuration and boundary update speed control results, it achieves accurate generation, real-time updating, and security early warning of electronic fences. First, based on the sampling frequency and sampling accuracy after dynamic adjustment of the sampling configuration, the UAV is controlled to carry out image acquisition and feature extraction of the target area. This ensures that the acquired image data and feature information can accurately represent the boundary features of the target area, providing a high-quality and highly adaptable basic data source for the generation of electronic fence boundary data, thus ensuring the accuracy of electronic fence generation from the root. Subsequently, based on the generated electronic fence boundary data, combined with the boundary update speed of each boundary segment after dynamic adjustment, the electronic fence boundary is dynamically generated and updated in real time. This allows the fence boundary to closely follow the UAV flight status and changes in target area features, effectively avoiding the fence control failure problem caused by boundary update lag, and ensuring the real-time performance and effectiveness of the fence boundary. During the electronic fence monitoring phase, a differentiated response mechanism is established by monitoring the positional relationship between the drone and the fence boundary in real time: When a drone enters the electronic fence boundary area, an electronic fence anomaly signal is immediately triggered, and the drone's sensor sampling frequency is directly increased to the highest threshold to enhance data acquisition accuracy. Simultaneously, a hovering command containing hovering coordinates, altitude, and attitude requirements is sent to the flight control system to quickly curb further unauthorized intrusion and reduce safety risks. When the drone reaches the electronic fence boundary, an early warning signal is issued, and the sampling frequency is adjusted to the highest threshold. Simultaneously, by calculating a direction vector perpendicular to the fence boundary tangent and pointing outwards, a heading adjustment command containing this direction vector and a preset safe distance is generated to accurately guide the drone away from the fence boundary, achieving proactive risk avoidance. When the drone does not reach the electronic fence boundary, the current sampling frequency and flight status are maintained to avoid resource waste caused by excessive control. This invention deeply links sampling configuration, boundary updates, and flight control through closed-loop management logic. This ensures the accuracy and real-time nature of electronic fence generation and achieves graded risk handling through a differentiated early warning response mechanism, effectively improving the safety and reliability of drone flight control and providing comprehensive fence protection support for drone missions with different priorities and coverage requirements.

[0051] like Figure 3The diagram shows the structure of a drone-based electronic fence dynamic generation and early warning system. This system includes: an image quality assessment module, a sampling configuration control module, an update speed control module, and a fence generation early warning module. The image quality assessment module receives the electronic fence generation command and obtains the task requirement parameters, including task coverage requirements and task priority levels. Based on the task coverage requirements, it dynamically divides the target area and obtains the drone's image acquisition quality parameters in real time based on the target area feature points and the drone's acquired images. These parameters include image matching rate, target area feature point deviation, and feature point repetition rate. The sampling configuration control module performs dynamic evaluation of sampling accuracy based on the task priority level of the electronic fence generation command and the drone's image acquisition quality parameters. As a result, the sampling configuration is dynamically adjusted based on the task requirement parameters and sensor performance parameters of the electronic fence generation command. The sensor performance parameters include sensor configuration and sensor latency, and the sampling configuration includes sampling frequency and sampling accuracy. The update speed adjustment module is used to divide the electronic fence into boundary segments. Based on the task priority level of the electronic fence generation command, it determines whether dynamic adjustment of the update speed is needed based on the relationship between the UAV flight speed and the update speed of each boundary segment. If needed, the boundary update speed is dynamically adjusted based on the boundary update configuration of each boundary segment. If not needed, the current boundary update speed is maintained. The boundary update configuration includes boundary point update unit, boundary point density, and boundary priority level. The fence generation early warning module is used to generate and warn of UAV electronic fences based on the results of the sampling configuration dynamic adjustment and the results of the boundary update speed dynamic adjustment.

[0052] The above description is only an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for dynamically generating and issuing early warnings for unmanned aerial vehicle (UAV) electronic fences, characterized in that, The method includes: The system receives an electronic fence generation instruction and obtains the task requirement parameters of the electronic fence generation instruction. The task requirement parameters include task coverage requirements and task priority level. The system dynamically divides the target area based on the task coverage requirements and obtains the image acquisition quality parameters of the UAV in real time based on the feature points of the target area and the images acquired by the UAV. The image acquisition quality parameters include image matching rate, target area feature point deviation, and feature point repetition rate. Based on the task priority level of the electronic fence generation command and the image acquisition quality parameters of the UAV, the sampling accuracy is dynamically evaluated. Based on the dynamic evaluation result of the sampling accuracy, the sampling configuration is dynamically adjusted based on the task requirement parameters of the electronic fence generation command and the sensor performance parameters. The sensor performance parameters include sensor configuration and sensor latency, and the sampling configuration includes sampling frequency and sampling accuracy. The boundary segments of the electronic fence are divided. Based on the task priority level of the electronic fence generation command, it is determined whether dynamic adjustment of the update speed is required based on the relationship between the drone flight speed and the update speed of each boundary segment. If required, the boundary update speed is dynamically adjusted based on the boundary update configuration of each boundary segment. If not required, the current boundary update speed is maintained. The boundary update configuration includes the boundary point update unit, boundary point density, and boundary priority level. Based on the dynamic adjustment results of sampling configuration and boundary update speed, the generation and early warning of UAV electronic fences are performed.

2. The method for dynamic generation and early warning of UAV electronic fences as described in claim 1, characterized in that, The steps for dynamically evaluating sampling accuracy based on the task priority level of the electronic fence generation command and the image acquisition quality parameters of the UAV include: The task priority level of the electronic fence generation instruction is input into a preset sampling accuracy threshold mapping table to obtain the corresponding minimum sampling accuracy threshold. The sampling accuracy threshold mapping table has a quantitative correspondence between different task priority levels and the minimum sampling accuracy threshold. Determine whether the current sampling frequency of the drone sensor has reached the minimum threshold for sampling accuracy; if so, mark the dynamic evaluation result of sampling accuracy as qualified. Otherwise, based on the image acquisition quality parameters, a comprehensive image quality index is obtained. It is then determined whether the comprehensive image quality index reaches the preset image quality threshold. If it does, the dynamic evaluation result of sampling accuracy is marked as needing adjustment; otherwise, the dynamic evaluation result of sampling accuracy is marked as unqualified.

3. The method for dynamic generation and early warning of UAV electronic fences as described in claim 2, characterized in that, The comprehensive image quality index is obtained as follows: Obtain preset image acquisition quality weight ratio parameters and image acquisition quality reference parameters. The image acquisition quality weight ratio parameters include image matching rate weight ratio, target region feature point deviation weight ratio, and feature point repetition rate weight ratio. The image acquisition quality reference parameters include image matching rate threshold, target region feature point deviation threshold, and feature point repetition rate threshold. The image matching rate influence index is obtained by weighting the results of the analysis of the proportion of image matching rate and image matching rate threshold using the weight ratio of image matching rate. The target region feature point deviation influence index is obtained by weighting the target region feature point deviation critical value and the target region feature point deviation ratio analysis results using the target region feature point deviation weight ratio. The feature point repetition rate influence index is obtained by weighting the feature point repetition rate and the feature point repetition rate threshold by using the feature point repetition rate weight ratio. By coupling the image matching rate influence index, the target region feature point deviation influence index, and the feature point repetition rate influence index, a comprehensive image quality index is obtained, which provides a quantitative basis for the result judgment in the scenario where the sampling frequency does not reach the minimum threshold in the dynamic evaluation of sampling accuracy.

4. The method for dynamic generation and early warning of UAV electronic fences as described in claim 2, characterized in that, The steps for dynamically adjusting the sampling configuration include: If the dynamic evaluation result of sampling accuracy is qualified, no additional processing is required; If the dynamic evaluation result of the sampling accuracy is that it needs adjustment, then the sampling frequency will be dynamically adjusted based on the sensor configuration and sensor delay, specifically as follows: The task priority level is input into a preset sampling frequency threshold mapping table to obtain the corresponding minimum sampling frequency threshold. The sampling frequency threshold mapping table has a quantitative correspondence between different task priority levels and the minimum sampling frequency threshold. The sensor delay refers to the time difference between when the sensor completes image acquisition and when the drone acquires the acquired image. The highest sampling frequency threshold and the sensor focal length threshold are obtained based on the sensor configuration. The sensor delay is input to a preset sampling frequency adjustment mapping table to obtain the corresponding sampling frequency adjustment value. Based on the sampling frequency adjustment value, the current sampling frequency of the UAV sensor is processed to obtain the adjusted sampling frequency. The sampling frequency adjustment mapping table has a quantitative correspondence between different sensor delay ranges and sampling frequency adjustment values. If the adjusted sampling frequency is within the range of the lowest and highest sampling frequency thresholds, then adjust the sensor sampling frequency to the adjusted sampling frequency. If the adjusted sampling frequency is lower than the minimum sampling frequency threshold, then adjust the sensor sampling frequency to the minimum sampling frequency threshold. If the adjusted sampling frequency is higher than the highest sampling frequency threshold, then adjust the sensor sampling frequency to the highest sampling frequency threshold. If the dynamic evaluation result of sampling accuracy is unqualified, the sampling frequency will be dynamically adjusted based on the sensor configuration and sensor delay, and the sampling accuracy will be dynamically adjusted based on the image acquisition quality parameters and task requirement parameters.

5. The method for dynamic generation and early warning of UAV electronic fences as described in claim 4, characterized in that, The steps for dynamically adjusting the sampling accuracy based on image acquisition quality parameters and task requirement parameters include: The task priority level is input into a preset sampling density threshold mapping table to obtain the corresponding minimum sampling density threshold. The sampling density threshold mapping table contains a quantitative correspondence between different task priority levels and the minimum sampling density threshold. If the image matching rate is lower than the image matching rate threshold, the difference between the image matching rate threshold and the image matching rate is marked as the image matching rate deviation value. The image matching rate deviation value is input into the preset sensor focal length adjustment mapping table to obtain the corresponding sensor focal length adjustment amount. The sensor focal length adjustment mapping table has a quantitative correspondence between different image matching rate deviation ranges and sensor focal length adjustment amounts. The current focal length of the sensor is reduced based on the sensor focal length adjustment amount to obtain the focal length of the sensor to be determined. If the image matching rate is not lower than the image matching rate threshold, then the current focal length of the sensor is recorded as the focal length of the sensor to be determined. Determine whether the focal length of the sensor to be determined is lower than the sensor focal length threshold. If so, adjust the sensor focal length to the sensor focal length threshold and mark the difference between the sensor focal length threshold and the focal length of the sensor to be determined as the focal length deviation value. Based on the focal length deviation value, issue a dynamic control warning for the UAV's flight altitude. Otherwise, adjust the sensor focal length to the focal length of the sensor to be determined. The difference between the image quality threshold and the comprehensive image quality index is marked as the image quality deviation value. The image quality deviation value is input into the preset sampling density adjustment mapping table to obtain the corresponding sampling density adjustment ratio. The sensor sampling density is dynamically adjusted based on the sampling density adjustment ratio to obtain the sampling density of the sensor to be judged. Determine whether the sampling density of the sensor to be judged has reached the minimum sampling density threshold. If so, adjust the sensor sampling density to the sampling density of the sensor to be judged. Otherwise, adjust the sensor sampling density to the minimum sampling density threshold.

6. The method for dynamic generation and early warning of UAV electronic fences as described in claim 1, characterized in that, The step of determining whether to perform dynamic adjustment of the update speed includes: The task priority level of the electronic fence generation command is input into a preset speed matching mapping table to obtain the speed deviation threshold. The speed matching mapping table has a quantitative correspondence between different task priority levels and speed deviation thresholds. The difference between the boundary update speed of each boundary segment and the flight speed of the UAV is marked as the speed deviation of each boundary segment. The speed deviation of each boundary segment is compared with the speed deviation threshold. If the speed deviation of any boundary segment is lower than the speed deviation threshold, the boundary segment is marked as needing to perform dynamic speed control. If the speed deviation of any boundary segment is not lower than the speed deviation threshold, it is marked as not requiring dynamic speed control updates.

7. The method for dynamic generation and early warning of UAV electronic fences as described in claim 1, characterized in that, The steps for dynamically adjusting the boundary update speed include: The difference between the drone's flight speed and the speed deviation threshold is used as the basis for updating the speed; Dynamic processing is performed based on the boundary update configuration of this boundary segment to obtain the dynamic adjustment coefficient of this boundary segment. The basic update rate is processed using the dynamic adjustment coefficient of this boundary segment to obtain the preliminary adjustment rate; The maximum allowable update speed is determined based on the performance of the drone; If the initial adjustment speed does not exceed the maximum allowable update speed, then the initial adjustment speed will be used as the adjusted boundary update speed for that boundary segment. If the initial adjustment speed exceeds the maximum allowable update speed, then the maximum allowable update speed will be used as the adjusted boundary update speed for that boundary segment.

8. The method for dynamic generation and early warning of UAV electronic fences as described in claim 7, characterized in that, The dynamic adjustment coefficient is obtained as follows: The boundary priority level is input into a preset priority coefficient mapping table to obtain the corresponding boundary priority coefficient. The priority coefficient mapping table contains a quantitative correspondence between different boundary priority levels and boundary priority coefficients. Obtain preset update weight percentage parameters and update weight reference parameters. The update weight percentage parameters include update unit weight percentage and boundary point density weight percentage. The update weight reference parameters include update unit critical value and boundary point density critical value. The update unit influence index is obtained by weighting the analysis results of the ratio of update unit to update unit threshold at boundary points using the update unit weight ratio. The boundary point density influence index is obtained by weighting the boundary point density and the boundary point density critical value by using the boundary point density weight ratio. By using the boundary priority coefficient to weight the coupling processing results of the boundary point density influence index and the update unit influence index, the dynamic adjustment coefficient of the boundary segment is obtained.

9. The method for dynamic generation and early warning of UAV electronic fences as described in claim 1, characterized in that, The steps for generating and issuing early warnings for drone electronic fences based on the dynamic adjustment results of sampling configuration and boundary update speed include: Based on the adjusted sampling frequency and sampling accuracy, the drone is controlled to acquire images and extract features from the target area, generating boundary data for the electronic fence. Based on the boundary data of the electronic fence, and the adjusted boundary update speed of each boundary segment, the electronic fence boundary is dynamically generated and updated in real time. During electronic fence monitoring, the positional relationship between the drone and the electronic fence boundary is monitored in real time, and corresponding operations are performed based on different positional states, specifically including: When a drone enters the boundary of the electronic fence, an abnormal signal is triggered, and the sampling frequency of the drone's sensors is directly adjusted to the highest sampling frequency threshold. When the drone reaches the boundary of the electronic fence, an electronic fence warning signal is issued, and the sampling frequency of the drone's sensors is adjusted to the highest sampling frequency threshold. When the drone does not reach the boundary of the electronic fence, the current sampling frequency and flight status of the drone's sensors are maintained.

10. A drone electronic fence dynamic generation and early warning system, applied in the drone electronic fence dynamic generation and early warning method as described in any one of claims 1-9, characterized in that, It includes an image quality assessment module, a sampling configuration control module, an update speed control module, and a fence generation and early warning module. The image quality assessment module is used to receive the electronic fence generation instruction and obtain the task requirement parameters of the electronic fence generation instruction. The task requirement parameters include task coverage requirements and task priority level. The target area is dynamically divided based on the task coverage requirements. The image acquisition quality parameters of the UAV are obtained in real time based on the feature points of the target area and the images acquired by the UAV. The image acquisition quality parameters include image matching rate, target area feature point deviation and feature point repetition rate. The sampling configuration control module is used to perform dynamic evaluation of sampling accuracy based on the task priority level of the electronic fence generation command and the image acquisition quality parameters of the UAV, and to perform dynamic control of sampling configuration based on the task requirement parameters of the electronic fence generation command and the sensor performance parameters according to the dynamic evaluation results of sampling accuracy. The sensor performance parameters include sensor configuration and sensor delay, and the sampling configuration includes sampling frequency and sampling accuracy. The update speed control module is used to divide the boundary segments of the electronic fence. Based on the task priority level of the electronic fence generation command, it determines whether dynamic update speed control is needed based on the relationship between the drone flight speed and the update speed of each boundary segment. If needed, the boundary update speed is dynamically controlled based on the boundary update configuration of each boundary segment. If not needed, the current boundary update speed is maintained. The boundary update configuration includes the boundary point update unit, boundary point density, and boundary priority level. The fence generation and early warning module is used to generate and issue early warnings for drone electronic fences based on the dynamic adjustment results of sampling configuration and the dynamic adjustment results of boundary update speed.

Images