Gridding unmanned aerial vehicle detection system, unmanned aerial vehicle detection single station and detection method

Abstract

The invention discloses a gridding unmanned aerial vehicle detection system, an unmanned aerial vehicle detection single station and a detection method, and relates to the technical field of unmanned aerial vehicle monitoring and early warning. Based on the problem; the method comprises the following steps: performing scanning data acquisition on a test unmanned aerial vehicle through each unmanned aerial vehicle detection single station deployed in a monitoring area, and obtaining an unmanned aerial vehicle feature data sequence of each unmanned aerial vehicle detection single station; based on the test unmanned aerial vehicle feature data of the unmanned aerial vehicle detection single station, constructing a detection feature of the unmanned aerial vehicle detection single station, marking the detection feature as an abnormal unmanned aerial vehicle detection single station, and isolating the abnormal unmanned aerial vehicle detection single station; according to the number and distribution of the isolated unmanned aerial vehicle detection single stations, obtaining a terrain shielding loss rate of the detection network corresponding to the unmanned aerial vehicle trajectory; according to the terrain shielding loss rate of the detection network of different unmanned aerial vehicle trajectories, obtaining detection risks of different unmanned aerial vehicle trajectories, and storing the detection risks in a database; and obtaining the predicted detection risk of the target unmanned aerial vehicle according to the similarity between the target unmanned aerial vehicle trajectory and the unmanned aerial vehicle trajectory in the database.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) monitoring and early warning, and specifically to a grid-based UAV detection system, a UAV detection single station, and a detection method. Background Art

[0002] With the rapid development of UAV technology, its wide application in civilian, commercial, and military fields has brought severe security supervision challenges. Traditional grid-based UAV detection methods mainly rely on single-point monitoring or a single sensor, and have the following limitations: Traditional methods obtain data through a single detection station, which is vulnerable to terrain occlusion, electromagnetic interference, or sensor blind spots, resulting in missed reports or false alarms. Most systems use fixed thresholds to determine anomalies and cannot adapt to the dynamic changes in UAV flight states or environments. The impact of complex terrain on the performance of the detection network is not considered, and it is difficult to quantify the monitoring blind spots caused by terrain occlusion, resulting in insufficient accuracy of risk prediction. Multi-source sensor data is not fully integrated, resulting in low information utilization rate and difficulty in dealing with miniaturized and stealth UAVs. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a grid-based UAV detection method, including the following steps: Step 1, scan and collect data on the test UAV through each UAV detection single station deployed in the monitoring area to obtain the UAV feature data sequence of each UAV detection single station; Step 2, construct the detection features of the detection UAV detection single station based on the test UAV feature data of the UAV detection single station, calculate the relative change values of the detection features of each UAV detection single station respectively. If the relative change value is less than the preset dynamic threshold, it is determined that the detection consistency of the UAV detection single station is consistent, and proceed to Step 3; otherwise, mark it as an abnormal UAV detection single station and isolate it; Step 3, align the test UAV trajectory data of each detection-consistent UAV detection single station according to the spatial coordinates of the UAV detection single station, generate the consistent test UAV trajectories corresponding to each UAV detection single station respectively, calculate the global spatial consistency coefficient. If the global spatial consistency coefficient is greater than the preset threshold, proceed to Step 5; otherwise, proceed to Step 4; Step 4, obtain the deviation between the test UAV trajectory of each UAV detection single station and the test UAV standard trajectory, and locate and isolate the abnormal detection UAV detection single station according to the deviation; Step 5, obtain the terrain occlusion loss rate of the detection network of the corresponding UAV trajectory according to the number and distribution of the isolated UAV detection single stations; obtain the detection risks of different UAV trajectories according to the terrain occlusion loss rates of the detection networks of different UAV trajectories, and store them in the database; Step 6: Obtain the predicted detection risk of the target UAV based on the similarity between the target UAV trajectory and the UAV trajectories in the database.

[0004] Furthermore, data collection is performed on the test UAV by the UAV detection single stations deployed in the monitoring area to obtain the UAV feature data sequences of each UAV detection single station, including: The UAV feature data includes the azimuth angle, elevation angle, and relative distance of the test UAV; according to the set acquisition duration and acquisition interval, the UAV feature data sequence is obtained; the relative distance is the distance between the test UAV and the UAV detection single station.

[0005] Furthermore, the detection features of the UAV detection single stations are constructed based on the signal data sequences of the UAV detection single stations, and the relative change values of the signal features of each UAV detection single station are calculated respectively, including: The detection feature of the UAV detection single station is: the relative change amount of the selected UAV feature data within the set acquisition interval; according to the difference between the relative change amounts of each UAV detection single station, the relative change value of the signal feature of the UAV detection single station is obtained.

[0006] Furthermore, the test UAV trajectory data of the UAV detection single stations with consistent detection consistencies are aligned according to the spatial coordinates of the UAV detection single stations, and the consistent test UAV trajectories corresponding to the UAV detection single stations are generated respectively, including: The test UAV trajectory data of the UAV detection single station is the test UAV trajectory data based on the spatial position of the UAV detection single station; the test UAV trajectory data of each UAV detection single station is converted to the same trajectory space to obtain the consistent test UAV trajectory corresponding to the UAV detection single station.

[0007] Furthermore, calculating the global spatial consistency coefficient includes: The global spatial consistency coefficient is: the proportion of the coincidence rate between the consistent test UAV trajectories of each corresponding UAV detection single station and the test UAV standard trajectory that is greater than the coincidence rate threshold.

[0008] Furthermore, obtaining the deviation between the test UAV trajectory of each UAV detection single station and the test UAV standard trajectory includes: The non - coincidence rate between the consistent test UAV trajectory of the corresponding UAV detection single station and the test UAV standard trajectory is the deviation between the test UAV trajectory of the UAV detection single station and the test UAV standard trajectory.

[0009] Furthermore, obtaining the terrain occlusion loss rate of the detection network corresponding to the UAV trajectory according to the number and distribution of the isolated UAV detection single stations includes: The terrain occlusion loss rate of the detection network corresponding to the UAV trajectory is obtained according to the ratio of the sum of the products of each isolated UAV detection single station and the corresponding loss weight to the sum of the products of the UAV detection single stations deployed within the monitoring area and the corresponding loss weight.

[0010] Further, obtaining the detection risk of the target UAV according to the similarity between the target UAV trajectory and the UAV trajectories in the database includes: Obtain the target UAV trajectory, match the UAV trajectory with the highest similarity in the database according to the target UAV trajectory, and obtain the terrain occlusion loss rate of the corresponding detection network; the terrain occlusion loss rate of the corresponding detection network is the predicted terrain occlusion loss rate of the target UAV trajectory, and the detection risk of the target UAV is obtained according to the predicted terrain occlusion loss rate of the detection network.

[0011] Grid the UAV detection single stations, and apply the described grid UAV detection method, including a UAV monitoring module, a power supply module, a positioning module, a communication module, and a data processing module; The UAV monitoring module, the power supply module, the positioning module, and the communication module are respectively connected to the data processing module.

[0012] A grid UAV detection system applies the UAV detection single stations, including multiple grid UAV detection single stations and a UAV detection control server; the multiple UAV detection single stations are respectively communicatively connected to the UAV detection control server; The UAV detection single stations are used to collect UAV data; The UAV detection control server is used to perform UAV monitoring and early warning according to the UAV data collected by the UAV detection single stations; The UAV detection control server includes a UAV detection single station networking control module and a database module; The UAV detection single station networking control module is used to control the UAV detection single stations to form a network; The database module is used to store the terrain occlusion loss rates of the detection networks corresponding to different UAV trajectories and the corresponding detection risks.

[0013] The beneficial effects of the present invention are as follows: By detecting the detection consistency of each UAV detection single station, the present invention can timely discover and isolate abnormal detection single stations, reducing detection errors caused by individual single station failures or environmental factors. At the same time, when calculating the global spatial consistency coefficient, the coincidence rate between the consistency test UAV trajectories of each single station and the standard trajectory is considered, further improving the accuracy of detection.

[0014] By calculating the relative change values of the detection characteristics of each single UAV detection station and setting dynamic thresholds for consistency judgment, the detection stability of each single station can be effectively guaranteed. Even if some single stations fail or are interfered, they can be detected and isolated in time to ensure the reliability of the entire detection system.

[0015] The present invention introduces the concept of terrain occlusion loss rate. According to the number and distribution of isolated UAV detection single stations, the terrain occlusion loss rate of the detection network for different UAV trajectories is calculated, and then the detection risk is evaluated. This risk assessment method is more comprehensive and accurate, and can provide a more effective decision-making basis for the supervision and disposal of UAVs.

[0016] By storing the detection risks of different UAV trajectories and quickly obtaining the predicted detection risk of the target UAV according to the similarity between the target UAV trajectory and the trajectories in the database, potential UAV threats can be discovered in advance, warnings can be issued in time, and time can be gained for subsequent disposal measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic flow chart of the grid UAV detection method; Figure 2 is a schematic principle diagram of the grid UAV detection single station; Figure 3 is a schematic principle diagram of the grid UAV detection system. DETAILED DESCRIPTION OF THE INVENTION

[0018] The technical solutions of the present invention will be further described in detail below with reference to the drawings, but the protection scope of the present invention is not limited to the following.

[0019] The features and performance of the present invention will be further described in detail below with reference to the embodiments.

[0020] As Figure 1 shown, the grid UAV detection method includes the following steps: Step 1, each UAV detection single station deployed in the monitoring area scans and collects data of the test UAV to obtain the UAV feature data sequence of each UAV detection single station; Step 2, based on the test UAV feature data of the UAV detection single station, the detection features of the UAV detection single station are constructed, and the relative change values of the detection features of each UAV detection single station are calculated respectively. If the relative change value is less than the preset dynamic threshold, it is determined that the detection consistency of the UAV detection single station is consistent, and step 3 is entered; otherwise, it is marked as an abnormal UAV detection single station and isolated; Step 3: Align the test UAV trajectory data of each UAV detection single station with consistent detection consistency according to the spatial coordinates of the UAV detection single station, respectively generate the consistent test UAV trajectories corresponding to the UAV detection single stations, calculate the global spatial consistency coefficient. If the global spatial consistency coefficient is greater than the preset threshold, go to Step 5; otherwise, go to Step 4; Step 4: Obtain the deviation between the test UAV trajectory and the test UAV standard trajectory of each UAV detection single station, and locate and isolate the abnormal UAV detection single station according to the deviation; Step 5: Obtain the terrain occlusion loss rate of the detection network of the corresponding UAV trajectory according to the number and distribution of the isolated UAV detection single stations; obtain the detection risks of different UAV trajectories according to the terrain occlusion loss rates of the detection networks of different UAV trajectories, and store them in the database; Step 6: Obtain the predicted detection risk of the target UAV according to the similarity between the target UAV trajectory and the UAV trajectories in the database.

[0021] Specifically, deploy multiple UAV detection single stations reasonably in the monitoring area. These single stations, like the nodes of the monitoring network, undertake the important task of data collection. Each UAV detection single station scans and collects data on the test UAV. The collected UAV feature data mainly includes the azimuth angle, elevation angle, and relative distance of the test UAV. The azimuth angle is used to determine the position of the UAV in the horizontal direction, the elevation angle reflects the angle of the UAV in the vertical direction, and the relative distance reflects the spatial interval between the UAV and the detection single station. According to the set collection duration and collection interval, the system can obtain a series of ordered UAV feature data, and these data are arranged in chronological order to form a UAV feature data sequence, providing the original data basis for subsequent analysis. For example, in a large park monitoring scenario, data is collected every 10 seconds for 5 consecutive minutes, thus obtaining 30 sets of UAV feature data sequences. Based on the test UAV feature data of the UAV detection single station obtained in the first step, the detection features of each detection single station are constructed. The detection feature here is defined as the relative change of the selected UAV feature data within the set acquisition interval. Taking the azimuth angle as an example, if the azimuth angle changes from 30 degrees to 35 degrees in two adjacent acquisition intervals, then the relative change of the azimuth angle is 5 degrees. By calculating the difference between the relative changes of each UAV detection single station, the relative change value of the UAV detection single station signal feature is obtained. Subsequently, this relative change value is compared with a preset dynamic threshold. If the relative change value is less than the preset dynamic threshold, it indicates that the detection data of this UAV detection single station has good consistency, the data is stable and reliable, and it can be determined that the detection consistency is consistent, and proceed to the next step; otherwise, mark this UAV detection single station as abnormal and isolate it to prevent abnormal data from interfering with subsequent analysis. This step is like checking the quality of the data to ensure that the data entering the subsequent processing link is accurate and effective. Align the test UAV trajectory data of each UAV detection single station with consistent detection consistency according to the spatial coordinates of the UAV detection single station. Since the spatial positions of each detection single station are different, there are differences in the trajectory data collected in space. It is necessary to convert these test UAV trajectory data based on different single station spatial positions to the same trajectory space to generate the consistent test UAV trajectory corresponding to the UAV detection single station. After completing the trajectory data alignment, calculate the global spatial consistency coefficient. The global spatial consistency coefficient is defined as the proportion of the coincidence rate of the consistent test UAV trajectories corresponding to each UAV detection single station and the test UAV standard trajectory that is greater than the coincidence rate threshold. If the global spatial consistency coefficient is greater than the preset threshold, it indicates that the consistency of the overall trajectory data is good; otherwise, it means that there may still be problems with some detection single station data. When the global spatial consistency coefficient does not reach the preset threshold, obtain the deviation between the test UAV trajectory of each UAV detection single station and the test UAV standard trajectory. Here, the non - coincidence rate between the consistent test UAV trajectory corresponding to the UAV detection single station and the test UAV standard trajectory is the deviation between the test UAV trajectory of the UAV detection single station and the test UAV standard trajectory. By analyzing these deviations, the abnormal detection UAV detection single stations with problems can be accurately located and isolated to further optimize the data quality. Calculate the terrain occlusion loss rate of the detection network corresponding to the UAV trajectory based on the number and distribution of isolated UAV detection single stations. The specific calculation method is as follows: obtain the terrain occlusion loss rate of the detection network corresponding to the UAV trajectory according to the ratio of the sum of the products of each isolated UAV detection single station and the corresponding loss weight to the sum of the products of the UAV detection single stations deployed within the monitoring area and the corresponding loss weight. Different terrain and environmental factors will have different degrees of influence on UAV detection, and these influences can be quantified by setting loss weights. For example, in mountainous areas and other regions with complex terrain, the loss weight can be set relatively high. According to the terrain occlusion loss rates of the detection networks corresponding to different UAV trajectories, the detection risks of different UAV trajectories can be evaluated, and these risk data can be stored in the database to provide data support for subsequent predictions. Obtain the target UAV trajectory, and match the UAV trajectory with the highest similarity in the database through a specific algorithm to obtain the terrain occlusion loss rate of the corresponding detection network. This terrain occlusion loss rate of the corresponding detection network is the terrain occlusion loss rate of the predicted detection network of the target UAV trajectory. Based on the terrain occlusion loss rate of the predicted detection network, the detection risk of the target UAV is finally obtained. In this way, the detection risk of the target UAV can be quickly and accurately predicted based on historical data, providing a decision-making basis for UAV monitoring and management. As Figure 2 Shown in the figure, the grid UAV detection single station is the hardware basis for implementing the above detection method, including a UAV monitoring module, a power supply module, a positioning module, a communication module, and a data processing module. The UAV monitoring module is responsible for scanning and data collection of UAVs to obtain key UAV feature data; the power supply module provides stable power supply for the entire detection single station to ensure the normal operation of each module; the positioning module can accurately determine the spatial coordinates of the detection single station to provide position information for operations such as trajectory data alignment; the communication module realizes data transmission between the detection single station and other devices or servers; the data processing module is the core of the detection single station. It is respectively connected to the UAV monitoring module, the power supply module, the positioning module, and the communication module, and is responsible for processing and analyzing the collected data, and executing various algorithms and logical judgments in the detection method. As Figure 3As shown in the figure, the grid unmanned aerial vehicle (UAV) detection system consists of multiple grid UAV detection single stations and a UAV detection control server. Multiple UAV detection single stations are respectively communicatively connected to the UAV detection control server. The UAV detection single station, as a front-end device, transmits the collected UAV data to the UAV detection control server in real time. The UAV detection control server is used to conduct UAV monitoring and early warning based on the UAV data collected by the UAV detection single station, including a UAV detection single station networking control module and a database module. The UAV detection single station networking control module is responsible for controlling the UAV detection single stations to form a network, optimizing the network structure, and ensuring the efficiency and stability of data transmission; the database module is used to store important data such as the terrain occlusion loss rate of the detection network for different UAV trajectories and the corresponding detection risks, providing data support for detection risk assessment and prediction, so that the system can continuously optimize the monitoring and early warning capabilities based on historical data.

[0022] Embodiment: UAV Detection in Urban Commercial Centers In a certain commercial center, including surrounding office buildings, shopping malls, squares and other areas, with a total area of about 5 square kilometers, a grid UAV detection system is deployed. The area has dense buildings, complex terrain, and a large flow of people. The monitoring of illegal UAV flights aims to ensure public safety and the normal progress of commercial activities. 20 UAV detection single stations are deployed in this area to build a monitoring network. Data Collection Set the collection duration from 9:00 to 22:00 on weekdays, and the collection interval is 3 seconds. Each detection single station collects the azimuth angle, elevation angle, and relative distance data of the UAV in real time. For example, detection single station E collects the UAV azimuth angle of 135°, elevation angle of 20°, and relative distance of 1500 meters at a certain moment, and continuously records to form a data sequence. Detection Feature Analysis and Anomaly Detection Construct the detection features of each detection single station and calculate the relative change value. For example, in the adjacent collection intervals of detection single station F, the elevation angle changes from 18° to 25°, and the relative change amount is 7°. Calculate the difference in relative change amounts of all single stations, set the preset dynamic threshold to 8°, and mark and isolate the abnormal detection single stations that exceed the threshold. Trajectory Data Alignment and Global Consistency Evaluation Perform spatial coordinate transformation on the trajectory data of the detection single stations with good consistency to generate a consistency test UAV trajectory. Set the coincidence rate threshold to 75%, calculate the global spatial consistency coefficient, and determine whether to enter the next step. Localization and Isolation of Abnormal Detection Single Stations If the global spatial consistency coefficient does not meet the standard, locate the abnormal detection single station by calculating the trajectory non - coincidence rate. For example, if the non - coincidence rate of detection single station G is relatively high, isolate it to purify the data. Terrain occlusion loss rate calculation and detection risk assessment Considering the dense buildings in this area, the loss weights of detection single stations at different positions are different. The loss weight of the detection single station near the high - rise building is set to 2, and the others are set to 1. Suppose 4 detection single stations are isolated, among which 2 have a loss weight of 2 and 2 have a loss weight of 1. Then the terrain occlusion loss rate = (2×2 + 2×1)÷(18×1 + 2×2)×100% ≈ 18.2%. Evaluate the detection risk and store the data. Target UAV detection risk prediction When the target UAV appears, after obtaining its trajectory, match the similar trajectory in the database. According to the terrain occlusion loss rate of the matched historical trajectory, determine the predicted detection risk of the target UAV, so that the relevant departments can respond in time to maintain the safety order of the urban commercial center.

Claims

1. Grid-based UAV detection method, characterized in that It includes the following steps: Step 1: Use each drone detection single station deployed in the monitoring area to scan and collect data from the test drone, and obtain the drone feature data sequence of each drone detection single station; Step 2: Based on the test drone feature data of the drone detection single station, construct the detection features of the detection drone detection single station, and calculate the relative change values of the detection features of each drone detection single station respectively. If the relative change value is less than the preset dynamic threshold, it is determined that the detection consistency of the drone detection single station is consistent, and proceed to Step 3; otherwise, mark it as an abnormal drone detection single station and isolate it; Step 3: Align the test drone trajectory data of each detection-consistent drone detection single station according to the spatial coordinates of the drone detection single station, and generate the consistent test drone trajectories corresponding to each drone detection single station respectively. Calculate the global spatial consistency coefficient. If the global spatial consistency coefficient is greater than the preset threshold, proceed to Step 5; otherwise, proceed to Step 4; Step 4: Obtain the deviation between the test drone trajectory of each drone detection single station and the test drone standard trajectory, and locate and isolate the abnormal detection drone detection single station according to the deviation; Step 5: According to the quantity and distribution of the isolated drone detection single stations, obtain the terrain occlusion loss rate of the detection network corresponding to the drone trajectory; according to the terrain occlusion loss rate of the detection network of different drone trajectories, obtain the detection risks of different drone trajectories, and store them in the database; Step 6: According to the similarity between the target drone trajectory and the drone trajectories in the database, obtain the predicted detection risk of the target drone.

2. The grid-based UAV detection method according to claim 1, wherein, The data collection of the test drone by the drone detection single station deployed in the monitoring area to obtain the drone feature data sequence of each drone detection single station includes: The drone feature data includes the azimuth angle, elevation angle, and relative distance of the test drone; according to the set collection duration and collection interval, obtain the drone feature data sequence; the relative distance is the distance between the test drone and the drone detection single station.

3. The grid-based UAV detection method according to claim 2, wherein, The construction of the detection features of the detection drone detection single station based on the signal data sequence of the drone detection single station and the calculation of the relative change values of the signal features of each drone detection single station respectively include: The detection feature of the drone detection single station is: the relative change amount of the selected drone feature data within the set collection interval; according to the difference between the relative change amounts of each drone detection single station, obtain the relative change value of the signal feature of the drone detection single station.

4. The grid-based UAV detection method according to claim 1, wherein, The alignment of the test drone trajectory data of each detection-consistent drone detection single station according to the spatial coordinates of the drone detection single station to generate the consistent test drone trajectories corresponding to each drone detection single station respectively includes: The test drone trajectory data of the drone detection single station is the test drone trajectory data based on the spatial position of the drone detection single station; convert the test drone trajectory data of each drone detection single station to the same trajectory space to obtain the consistent test drone trajectories corresponding to each drone detection single station.

5. The grid-based UAV detection method according to claim 4, characterized in that, The calculation of the global spatial consistency coefficient includes: The global spatial consistency coefficient is: the ratio of the proportion of the coincidence rate of the consistency test UAV trajectory and the test UAV standard trajectory of each corresponding UAV detection single station being greater than the coincidence rate threshold.

6. The grid-based UAV detection method according to claim 5, characterized in that, Obtaining the deviation between the test UAV trajectory and the test UAV standard trajectory of each UAV detection single station includes: The non - coincidence rate between the consistency test UAV trajectory and the test UAV standard trajectory of the corresponding UAV detection single station is the deviation between the test UAV trajectory and the test UAV standard trajectory of the UAV detection single station.

7. The grid unmanned aerial vehicle detection method according to claim 1, wherein, Obtaining the terrain occlusion loss rate of the detection network corresponding to the UAV trajectory according to the number and distribution of the isolated UAV detection single stations includes: Obtaining the terrain occlusion loss rate of the detection network corresponding to the UAV trajectory according to the ratio of the sum of the products of each isolated UAV detection single station and the corresponding loss weight to the sum of the products of the UAV detection single stations deployed in the monitoring area and the corresponding loss weight.

8. The grid-based UAV detection method according to claim 1, characterized in that, Obtaining the detection risk of the target UAV according to the similarity between the target UAV trajectory and the UAV trajectories in the database includes: Obtaining the target UAV trajectory, matching the UAV trajectory with the highest similarity in the database according to the target UAV trajectory, obtaining the terrain occlusion loss rate of the corresponding detection network; the terrain occlusion loss rate of the corresponding detection network is the predicted terrain occlusion loss rate of the target UAV trajectory, and obtaining the detection risk of the target UAV according to the predicted terrain occlusion loss rate of the detection network.

9. Grid-based unmanned aerial vehicle detection single station, characterized in that, Applying the grid - based UAV detection method according to any one of claims 1 - 8, including a UAV monitoring module, a power supply module, a positioning module, a communication module, and a data processing module; The UAV monitoring module, the power supply module, the positioning module, and the communication module are respectively connected to the data processing module.

10. Grid-based UAV detection system, characterized in that, Applying the grid - based UAV detection single station according to claim 9, including a plurality of UAV detection single stations and a UAV detection control server; the plurality of UAV detection single stations are respectively communicatively connected to the UAV detection control server; The UAV detection single station is used to collect UAV data; The UAV detection control server is used to perform UAV monitoring and early warning according to the UAV data collected by the UAV detection single station; The UAV detection control server includes a UAV detection single station networking control module and a database module; The UAV detection single station networking control module is used to control the UAV detection single stations to form a network; The database module is used to store the terrain occlusion loss rate of the detection network corresponding to different UAV trajectories and the corresponding detection risks.

Images