Low-altitude unmanned aerial vehicle anti-reverse method based on radar positioning and image recognition technology

Abstract

The invention provides a radar positioning and image recognition technology-based low-altitude unmanned aerial vehicle anti-reverse method, and relates to the technical field of unmanned aerial vehicle anti-reverse. The method comprises the following steps: establishing a coordinate system, and setting an unmanned aerial vehicle no-fly area and a risk area; obtaining first coordinate data of the unmanned aerial vehicle in the coordinate system, and determining whether there is a risk of losing a target by the radar; and if the risk that the radar loses the target exists, determining the first camera and the orientation thereof according to the first coordinate data and the camera coordinate data. Determining a risk coefficient of the unmanned aerial vehicle; and when the risk coefficient is higher than a preset risk threshold value, determining an alarm level, recording the image and determining a countering measure. According to the method and the device, when the radar has the risk of losing the target, the appropriate camera can be selected in time to shoot the unmanned aerial vehicle so as to reduce the risk of losing the target, and after the unmanned aerial vehicle enters the risk area, the risk of the unmanned aerial vehicle can be judged in real time, the interference to the passing unmanned aerial vehicle without risk is reduced, and the unmanned aerial vehicle with risk is effectively countered.

Description

Technical Field

[0001] The present invention relates to the technical field of anti-drone technology, and in particular to a method for preventing and countering low-altitude drones based on radar positioning and image recognition technology. Background Art

[0002] In the related art, a radar is usually used to detect the position of a drone, and anti-drone measures such as signal interference are taken to force the drone to stay away from the no-fly zone. However, when the flight altitude of the drone is relatively low, there is a risk that the radar may lose the target. Moreover, not all drones approaching the no-fly zone pose a threat, and it is difficult for the related art to determine the risk of approaching drones. Summary of the Invention

[0003] The present invention provides a method for preventing and countering low-altitude drones based on radar positioning and image recognition technology, which can solve the technical problems in the related art that the radar has a risk of losing the target and it is difficult to determine the risk of approaching drones.

[0004] According to a first aspect of the present invention, there is provided a method for preventing and countering low-altitude drones based on radar positioning and image recognition technology, including:

[0005] Taking the radar as the coordinate origin, establishing a coordinate system, and setting a no-fly zone for drones and a risk area in the coordinate system, wherein the no-fly zone for drones is located within the risk area, and the second range of the risk area is larger than the first range of the no-fly zone for drones;

[0006] When the radar detects that a drone enters the risk area, obtaining first coordinate data of the drone in the coordinate system;

[0007] Determining whether there is a risk that the radar loses the target according to the first coordinate data;

[0008] If there is a risk that the radar loses the target, determining the first camera to be turned on and the orientation of the first camera according to the first coordinate data and the camera coordinate data of multiple cameras, wherein the multiple cameras are arranged at multiple positions outside the no-fly zone for drones and within the risk area;

[0009] Determining the risk coefficient of the drone according to at least one of the first coordinate data and the video frames captured by the first camera, the duration of the drone entering the risk area, and the first range of the no-fly zone for drones;

[0010] Determining the alarm level when the risk coefficient is higher than a preset risk threshold;

[0011] Selecting a second camera to capture a recorded image of the drone according to the first coordinate data;

[0012] Determine countermeasures according to the alarm level.

[0013] According to the present invention, determining whether there is a risk of radar target loss based on the first coordinate data includes:

[0014] Based on the first coordinate data, determine the planar coordinates (x t , y t ) of the UAV at the current moment and the planar coordinates (x t-1 , y t-1 ) at the previous moment, where the current moment is the t-th moment after the UAV enters the risk area;

[0015] If then obtain the altitude coordinate z of the UAV at the current moment t and the altitude coordinate z at the previous moment t-1 ;

[0016] According to the formula

[0017]

[0018] Determine the judgment condition C1, judgment condition C2, and judgment condition C3, where h ll is the lower limit of the detection altitude of the radar, δ1 is a coefficient greater than 1, Δt is the time difference between the current moment and the previous moment, and n d is the number of moments corresponding to the preset camera calibration duration;

[0019] When the judgment condition C1, judgment condition C2, and judgment condition C3 are all satisfied, it is determined that there is a risk of radar target loss.

[0020] According to the present invention, if there is a risk of radar target loss, then based on the first coordinate data and the camera coordinate data of multiple cameras, determine the first camera to be turned on and the orientation of the first camera, including:

[0021] Input the first coordinate data of multiple moments after the UAV enters the risk area into the trained position prediction model to obtain the predicted coordinate data for the first preset number of future moments;

[0022] Determine the first connection line between the planar coordinates corresponding to the camera coordinate data and the planar coordinates corresponding to the first coordinate data, and determine the second connection line between the planar coordinates corresponding to the camera coordinate data and the planar coordinates corresponding to the predicted coordinate data;

[0023] Determine the third range of the obstacle area within the risk area, where the third range includes the first range;

[0024] Among multiple cameras, determine the cameras to be determined for which the first connection line and the second connection line have no intersection with the third range;

[0025] According to the camera coordinate data of the cameras to be determined, the first coordinate data, the predicted coordinate data, and the third range, determine the preference coefficient of the cameras to be determined;

[0026] According to the preference coefficients of each camera to be determined, determine the first camera;

[0027] According to the camera coordinate data of the first camera and the first coordinate data, determine the orientation of the first camera.

[0028] According to the present invention, determining the preference coefficient of the cameras to be determined according to the camera coordinate data of the cameras to be determined, the first coordinate data, the predicted coordinate data, and the third range includes:

[0029] According to the formula

[0030]

[0031] Determine the preference coefficient C of the i-th camera to be determined S,i , where L 1,i is the distance between the first coordinate data and the camera coordinate data of the i-th camera to be determined, l 3,i is the minimum planar distance between the planar coordinate corresponding to the first coordinate data and the third range, L 1,i,p,j is the distance between the predicted coordinate data at the future j-th moment and the camera coordinate data of the i-th camera to be determined, l 3,i,p,j is the minimum planar distance between the planar coordinate corresponding to the predicted coordinate data at the future j-th moment and the third range, γ is a coefficient less than 1 and greater than 0, N1 is the first preset quantity, j ≤ N1, and both j and N1 are positive integers.

[0032] According to the present invention, determining the risk coefficient of the unmanned aerial vehicle according to at least one of the first coordinate data and the video frames captured by the first camera, the duration of the unmanned aerial vehicle entering the risk area, and the first range of the no-fly zone of the unmanned aerial vehicle includes:

[0033] If the radar does not lose the target, use the first coordinate data as the positioning data of the unmanned aerial vehicle;

[0034] If the radar loses the target, determine the positioning data of the unmanned aerial vehicle according to the image coordinates of the unmanned aerial vehicle in the video frames captured by multiple first cameras and the calibration parameters of the multiple first cameras;

[0035] Determine the risk coefficient of the drone based on the positioning data of the drone at multiple moments after entering the risk area, the duration of the drone entering the risk area, and the first range of the no-fly zone of the drone.

[0036] According to the present invention, determining the risk coefficient of the drone based on the positioning data of the drone at multiple moments after entering the risk area, the duration of the drone entering the risk area, and the first range of the no-fly zone of the drone includes:

[0037] Determine the centroid plane coordinates of the first range;

[0038] Obtain the third connection line between the centroid plane coordinates and the plane coordinates corresponding to the positioning data at each moment;

[0039] Determine the first intersection point of the third connection line and the projection edge of the first range;

[0040] Obtain the first plane distance between the first intersection point and the plane coordinates corresponding to the positioning data at each moment;

[0041] Fit the first plane distance corresponding to each moment with each moment to obtain a plane distance function;

[0042] Obtain the first distance between the first intersection point corresponding to each moment and the first intersection point corresponding to the moment when the radar detects the drone entering the risk area on the projection edge of the first range;

[0043] Determine the first ratio of the first distance to the total length of the projection edge of the first range;

[0044] Fit the first ratio corresponding to each moment with each moment to obtain an observation ratio function;

[0045] Determine the risk coefficient of the drone according to the observation ratio function, the plane distance function, and the duration of the drone entering the risk area.

[0046] According to the present invention, determining the risk coefficient of the drone according to the observation ratio function, the plane distance function, and the duration of the drone entering the risk area includes:

[0047] According to the formula

[0048]

[0049] Determine the risk coefficient R of the drone, where f o (T) is the observation ratio function, f l (T) is the plane distance function, T is the time variable, T t is the duration of the drone entering the risk area, r is the radius of the risk area, is the average planar movement speed of the drone at multiple moments after the drone enters the risk area, 0 ≤ T ≤ T t , where w1 and w2 are preset weights.

[0050] According to the present invention, the method further includes:

[0051] When at least one first camera loses sight of the drone or at the end of the first preset number of future moments, re-determine the first camera.

[0052] According to the present invention, selecting a second camera to capture a recorded image of the drone based on the first coordinate data includes:

[0053] Immediately re-determine the first camera and determine the re-determined first camera as the second camera to capture the recorded image.

[0054] By adopting the above technical solutions, the present invention can achieve the following technical effects:

[0055] According to the present invention, when the unmanned aerial vehicle (UAV) enters a risk area and approaches the no-fly zone of the UAV, the coordinate data of the UAV can be detected by radar. When there is a risk of the radar losing the target, a suitable camera can be selected in a timely manner to photograph the UAV, thereby reducing the risk of losing the target. After the UAV enters the risk area, the risk of the UAV can be judged in real time, reducing the interference to the passing UAVs without risk. When judging whether there is a risk of the radar losing the target, it can be judged whether the UAV entering the risk area is leaving the risk area. If the UAV has not left the risk area, it is judged whether the UAV is at a relatively low altitude and the altitude is still decreasing through judgment conditions. Moreover, the time required for the UAV to reach the lower limit of the detection altitude of the radar is short, so as to judge whether there is a risk of the radar losing the target, and a calibration time is reserved for the camera to track and photograph the UAV, so that the UAV can be accurately tracked, the tracking accuracy can be improved, and the probability of losing the target can be reduced. When selecting the first camera, the ratio of the minimum planar distance between the planar coordinates corresponding to the first coordinate data and the third range and the distance between the first coordinate data and the camera coordinate data of the i-th undetermined camera, and the ratio of the minimum planar distance between the planar coordinates corresponding to the predicted coordinate data at the future j-th moment and the third range and the distance between the predicted coordinate data at the future j-th moment and the camera coordinate data of the i-th undetermined camera can be used to determine the possibility that the undetermined camera at each moment can continuously, clearly and stably track and photograph the UAV. When solving the possibility that the undetermined camera at the future moment can continuously, clearly and stably track and photograph the UAV, considering the accuracy of the position prediction model, a discount rate is set for the future possibility, so as to improve the accuracy of the preference coefficient, provide an accurate objective standard for screening the first camera, improve the accuracy and stability of tracking and photographing the UAV, and reduce the probability of losing the target. When determining the risk coefficient of the UAV, the risks of the UAV's observation clarity of the no-fly zone of the UAV and the risks of the UAV's observation comprehensiveness of the no-fly zone of the UAV can be processed by integration to obtain the risk of the UAV observing the no-fly zone of the UAV. The possibility that the UAV deliberately stays to observe the no-fly zone of the UAV is represented by the ratio of the duration of the UAV entering the risk area to the longest time required for the UAV to cross the risk area at its average speed, so as to obtain the risk coefficient of the UAV, which can accurately represent the risk that the UAV deliberately stays in the risk area to observe the no-fly zone of the UAV and observes more information, and can accurately represent the risk of information leakage of the prohibited shooting information in the no-fly zone of the UAV. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 Exemplarily shows a schematic flow chart of a low-altitude UAV anti-reversal method based on radar positioning and image recognition technology according to an embodiment of the present invention;

[0057] Figure 2Exemplarily shown is a schematic diagram of a no-fly zone and a risk zone of a drone according to an embodiment of the present invention;

[0058] Figure 3 Exemplarily shown is a schematic diagram of first coordinate data and predicted coordinate data according to an embodiment of the present invention;

[0059] Figure 4 Exemplarily shown is a schematic diagram of a third connection line according to an embodiment of the present invention. Detailed implementation manners

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0061] The technical solutions of the present invention will be described in detail below with specific embodiments. These specific embodiments may be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0062] Figure 1 Exemplarily shown is a schematic flowchart of a low-altitude drone anti-reflection method based on radar positioning and image recognition technologies according to an embodiment of the present invention. The method includes:

[0063] Step S1: Taking the radar as the coordinate origin, establishing a coordinate system, and setting a no-fly zone and a risk zone for the drone in the coordinate system. Among them, the no-fly zone for the drone is located within the risk zone, and the second range of the risk zone is larger than the first range of the no-fly zone for the drone;

[0064] Step S2: When the radar detects that the drone enters the risk zone, obtaining the first coordinate data of the drone in the coordinate system;

[0065] Step S3: Determining whether there is a risk of the radar losing the target according to the first coordinate data;

[0066] Step S4: If there is a risk of the radar losing the target, determining the first camera to be turned on and the orientation of the first camera according to the first coordinate data and the camera coordinate data of multiple cameras, where the multiple cameras are arranged at multiple positions outside the no-fly zone for the drone and within the risk zone;

[0067] Step S5: Determine the risk coefficient of the UAV based on at least one of the first coordinate data and the video frames captured by the first camera, the duration of the UAV entering the risk area, and the first range of the UAV no-fly zone.

[0068] Step S6: Determine the alarm level when the risk coefficient is higher than the preset risk threshold.

[0069] Step S7: Select a second camera to capture a recorded image of the UAV according to the first coordinate data.

[0070] Step S8: Determine countermeasures according to the alarm level.

[0071] According to the low-altitude UAV anti-countermeasure method based on radar positioning and image recognition technology of the embodiments of the present invention, after the UAV enters the risk area and approaches the UAV no-fly area, the coordinate data of the UAV can be detected by radar, and when there is a risk of losing the target by the radar, a suitable camera can be selected in time to capture the UAV, thereby reducing the risk of losing the target. After the UAV enters the risk area, the risk of the UAV can be judged in real time, and the interference to the passing UAVs without risk can be reduced.

[0072] According to an embodiment of the present invention, in step S1, the radar can be located within the risk area or within the UAV no-fly area. The position where the radar is located can be used as the coordinate origin to establish a coordinate system. For example, the ground plane is set as the xoy plane, and the vertical direction is determined as the z-axis direction.

[0073] Figure 2 Exemplarily shows a schematic diagram of the UAV no-fly area and the risk area according to an embodiment of the present invention. Figure 2 It is a plan view of the UAV no-fly area and the risk area. The UAV no-fly area is located within the risk area. The second range of the risk area (the second range is a circular range centered on the position where the radar is located in the top view, that is, the range shown by the circular dotted line) is larger than the first range of the UAV no-fly area (the first range is a rectangular dotted line range in the top view). The heights of the first range and the second range can be set to infinity. In other words, as long as the plane coordinates of the UAV (that is, the x-axis coordinate and the y-axis coordinate) are within the circular range, it can be considered that the UAV enters the risk area. As long as the plane coordinates of the UAV are within the rectangular range, it can be considered that the UAV enters the UAV no-fly area. A radar is set within the UAV no-fly area (and the radar is located at the coordinate origin, that is, the coordinates of the radar are (0, 0, 0)). Outside the UAV no-fly area and at multiple positions within the risk area, multiple cameras are set. The multiple cameras are arranged around the UAV no-fly area. For example, the multiple cameras are evenly arranged around the UAV no-fly area.

[0074] According to an embodiment of the present invention, in step S2, the radar can scan the surroundings every preset time period (for example, 2 seconds). If it detects that the drone enters the second range of the risk area, it can determine the first coordinate data of the drone. For example, the distance between the drone and the radar can be determined by the time difference between the transmitted signal and the echo. By rotating or electronically scanning the radar antenna, the direction with the strongest echo can be determined, and the azimuth angle of the drone can be determined using this direction. The height of the drone can be determined by the elevation angle of the echo, so as to determine the relative position relationship between the drone and the radar. Furthermore, based on the coordinates of the radar in the coordinate system, the first coordinate data of the drone in the coordinate system can be determined.

[0075] According to an embodiment of the present invention, in step S3, due to factors such as a certain elevation angle of the radar, the detection difficulty of aircraft with a relatively low flight altitude increases, and the target may be lost. For example, if the flight altitude of the drone is lower than 50 meters, the risk that the radar cannot detect the drone (i.e., loses the target) increases. Therefore, it can be determined whether there is a risk of the radar losing the target based on whether the flight altitude of the drone is relatively low or whether the drone has a trend of decreasing flight altitude.

[0076] According to an embodiment of the present invention, in step S3, determining whether there is a risk of the radar losing the target based on the first coordinate data includes:

[0077] Step S31: According to the first coordinate data, determine the planar coordinates (x t , y t ) of the drone at the current moment and the planar coordinates (x t-1 , y t-1 ) of the previous moment, where the current moment is the t-th moment after the drone enters the risk area;

[0078] Step S32: If then obtain the height coordinate z t of the drone at the current moment and the height coordinate z t-1 of the previous moment;

[0079] Step S33: Determine the judgment conditions C1, C2, and C3 according to formula (1),

[0080]

[0081] where h ll is the lower limit of the detection height of the radar, δ1 is a coefficient greater than 1, Δt is the time difference between the current moment and the previous moment, and n d is the number of moments corresponding to the preset camera calibration duration;

[0082] Step S34: When judgment conditions C1, C2, and C3 are all satisfied, it is determined that there is a risk of the radar losing the target.

[0083] According to an embodiment of the present invention, in step S31, based on the above method for obtaining the first coordinate data, the first coordinate data of the UAV can be obtained at each moment, and based on the first coordinate data, the planar coordinates of the UAV at the current moment and the planar coordinates at the previous moment are determined. That is, the coordinate data of the first two dimensions in the first coordinate data (three-dimensional coordinate data) (i.e., the x-axis coordinate and the y-axis coordinate) are determined as the planar coordinates, and the planar coordinates of the UAV at the current moment and the planar coordinates at the previous moment are obtained. The interval between the two moments is the above-mentioned preset duration (for example, 2 seconds).

[0084] According to an embodiment of the present invention, in step S32, it can be judged whether the UAV is leaving the risk area. If That is, the distance between the UAV and the coordinate origin is shrinking. In other words, the UAV is approaching the central area of the risk area (i.e., the no-fly area of the UAV), then it can continue to judge whether there is a risk of the radar losing the target. Otherwise, if It indicates that the UAV is leaving the risk area, and there is no need to further judge the risk of the radar losing the target.

[0085] According to an embodiment of the present invention, in step S33, formula (1) describes three judgment conditions for judging whether there is a risk of the radar losing the target. In judgment condition C1, the lower limit of the detection height of the radar can be set to 50 meters. When the flight height of the UAV is lower than 50 meters, the probability that the radar cannot detect the UAV (i.e., loses the target) increases. That is, there is a risk of losing the target. The judgment height can be set to δ1h ll , where δ1 can be set to 1.5 or 2. If the height coordinate of the UAV is less than or equal to δ1h ll , it is judged that the UAV is approaching the lower limit of the radar's detection height.

[0086] According to an embodiment of the present invention, in judgment condition C2, The height change speed between the current moment and the previous moment, It indicates that the direction of the height change speed is downward, that is, the height of the UAV is decreasing.

[0087] According to an embodiment of the present invention, in judgment condition C3, the height change z of the UAV t -z t-1 The required duration is Δt, that is, a preset duration. At this descent speed, the duration required for the UAV to descend from the current height to the lower limit of the radar's detection height is When the UAV approaches the lower limit of the detection altitude of the radar, the camera can be used to aim at the UAV for shooting. Thus, when the radar has difficulty detecting the UAV, the UAV can be continuously tracked through the camera. However, the process of aiming the camera at the UAV and shooting requires a certain amount of time. For example, actions such as adjusting the orientation angle of the camera and adjusting the focal length require a certain amount of time, which is collectively referred to as the calibration time, n d Δt is the average calibration time of the camera and can be used as the preset calibration time of the camera (for example, n d = 1). If the time required for the UAV to descend from the current altitude to the lower limit of the detection altitude of the radar is less than the preset calibration time of the camera, the camera will also have difficulty shooting the UAV, which may cause the loss of the target. Therefore, in That is, When, a judgment can be made in advance, so as to give the camera calibration time, which is convenient for the camera to track and shoot the UAV at a lower altitude.

[0088] According to an embodiment of the present invention, in step S34, if the judgment conditions C1, C2, and C3 are simultaneously satisfied, it means that the UAV is at a low altitude and the altitude is still decreasing, and the time required for the UAV to reach the lower limit of the detection altitude of the radar is short, approaching the preset calibration time of the camera. In this case, it can be judged that there is a risk of the radar losing the target, and the camera calibration time is given, so that the UAV can be timely aimed at for tracking and shooting, so that when the radar has difficulty detecting the UAV, the UAV can be continuously tracked through the camera.

[0089] In this way, it can be judged whether the UAV entering the risk area is leaving the risk area. If the UAV does not leave the risk area, it is judged whether the UAV is at a low altitude and the altitude is still decreasing through the judgment conditions, and the time required for the UAV to reach the lower limit of the detection altitude of the radar is short, so as to judge whether there is a risk of the radar losing the target, and reserve the calibration time for the camera to track and shoot the UAV, so that the UAV can be accurately tracked, the tracking accuracy can be improved, and the probability of losing the target can be reduced.

[0090] According to an embodiment of the present invention, in step S4, as shown above, if there is a risk of the radar losing the target, in order to accurately track the UAV, the UAV can be tracked and shot through the camera, so as to reduce the risk of completely losing the target. A suitable first camera can be selected from multiple cameras to shoot the UAV. The first camera can be a camera that is relatively close to the UAV and there are no obstacles in the space between the first camera and the UAV, so that the picture of the UAV can be successfully shot.

[0091] According to an embodiment of the present invention, in step S4, if there is a risk of the radar losing the target, then according to the first coordinate data and the camera coordinate data of multiple cameras, determine the first camera to be turned on and the orientation of the first camera, including:

[0092] Step S41, input the first coordinate data at multiple moments after the UAV enters the risk area into the trained position prediction model to obtain the predicted coordinate data at the first preset number of future moments;

[0093] Step S42, determine the first connection line between the plane coordinates corresponding to the camera coordinate data and the plane coordinates corresponding to the first coordinate data, and determine the second connection line between the plane coordinates corresponding to the camera coordinate data and the plane coordinates corresponding to the predicted coordinate data;

[0094] Step S43, determine the third range of the obstacle area within the risk area, where the third range includes the first range;

[0095] Step S44, among multiple cameras, determine the pending cameras whose first connection line and second connection line have no intersection with the third range;

[0096] Step S45, according to the camera coordinate data of the pending camera, the first coordinate data, the predicted coordinate data, and the third range, determine the preference coefficient of the pending camera;

[0097] Step S46, according to the preference coefficients of each pending camera, determine the second preset number of first cameras;

[0098] Step S47, according to the camera coordinate data of the first camera and the first coordinate data, determine the orientation of the first camera.

[0099] Figure 3 Exemplarily, a schematic diagram of the first coordinate data and the predicted coordinate data according to an embodiment of the present invention is shown.

[0100] According to an embodiment of the present invention, in step S41, the position prediction model can be a BP neural network model, an LSTM model, etc. The present invention does not limit the type of the position prediction model. The position prediction model can be trained through the historical coordinate data of the drone collected in the past time period. For example, input the positions at the 1st to the nth historical moments into the position prediction model, and the position prediction model can output the predicted positions at the (n + 1)th to the (n + m)th moments. Then, the historical position data at the (n + 1)th to the (n + m)th historical moments can be compared with the predicted positions, and the error of the predicted positions can be determined. Furthermore, the loss function of the position prediction model can be determined, and the position prediction model can be trained by backpropagating the loss function. After multiple trainings and verifying the accuracy of the position prediction model, the trained position prediction model can be obtained, where n and m are both positive integers. The trained position prediction model can be used to process the first coordinate data (such as Figure 3 shown by the solid small rectangle in Figure 3 ) at multiple moments after the drone enters the risk area to obtain the predicted coordinate data at the first preset number of future moments (such as

[0101] shown by the dashed small rectangle in

[0102] According to an embodiment of the present invention, in step S42, the camera coordinate data is the three-dimensional coordinate data of the camera in the coordinate system. The plane coordinate corresponding to the camera coordinate data is the coordinate data of the first two dimensions in the three-dimensional coordinate data, that is, the x-axis coordinate data and the y-axis coordinate data. Both the first coordinate data and the predicted coordinate data are three-dimensional coordinate data, and their corresponding plane coordinate data are also the coordinate data of the first two dimensions in the three-dimensional coordinate data, that is, the x-axis coordinate data and the y-axis coordinate data. The plane coordinate corresponding to the camera coordinate data can be connected with the plane coordinate corresponding to the first coordinate data to obtain the first connection line, and the plane coordinate corresponding to the camera coordinate data can be connected with the plane coordinate corresponding to the predicted coordinate data to obtain the second connection line. The first connection line and the second connection line can be used to determine whether there is an obstacle blocking between the camera and the drone. If in the two-dimensional plane, the first connection line and the second connection line do not cross an obstacle, then in the three-dimensional space, there is also no obstacle blocking between the drone and the camera, and using two-dimensional coordinates can simplify the operation and save computing resources.

[0103] According to an embodiment of the present invention, in step S43, the obstacle area is the range of the area where the obstacle is located. For example, the area where obstacles such as trees in the risk area are located. The no-fly zone of the drone may also block the field of view of the camera. Therefore, the no-fly zone of the drone can also be used as the obstacle area, that is, the third range can include the first range.

[0103] According to an embodiment of the present invention, in step S44, the first connection line ( Figure 3The solid line connection between the camera in the figure and the solid small rectangle corresponding to the first coordinate data) has no intersection with the third range, which can indicate that there is no obstacle blocking between the current position of the drone and the camera. The second connection ( Figure 3 The dashed line connection between the camera in the figure and the dashed small rectangle corresponding to the predicted coordinate data) has no intersection with the third range, which can indicate that there is no obstacle blocking between the predicted coordinate data of the drone and the camera. Since both the first connection and the second connection have no intersection with the third range, the camera can capture the drone for multiple consecutive moments without being interfered by obstacles and can be used as a candidate camera.

[0104] According to an embodiment of the present invention, according to Figure 2 the distribution of the cameras in the figure, when the drone is in a risk area, there may be multiple candidate cameras, that is, there are multiple cameras that can capture the drone for multiple consecutive moments. However, only 2-3 cameras are needed to locate the coordinates of the drone. Therefore, the most suitable camera for shooting can be selected from multiple candidate cameras. For example, the camera closest to the drone and with the clearest shot is used as the first camera. The distribution of the cameras can also be Figure 2 different, but there may still be multiple candidate cameras, and screening is required. If the number of candidate cameras is small (for example, there are many obstacles, resulting in fewer cameras that can capture the drone for multiple consecutive moments), for example, only 2 or 3, then all candidate cameras can be directly determined as the first camera. If there is only 1 candidate camera, or even no candidate camera, the number of predicted coordinate data of the drone can be reduced (for example, the first preset number is reduced by one each time, so that the predicted coordinate data is reduced by one), thereby reducing the number of the second connections, and further relaxing the screening conditions for candidate cameras, that is, in fewer consecutive moments, there can be more cameras that can continuously capture the drone without being interfered by obstacles. If there is still no candidate camera or only 1 candidate camera, the number of predicted coordinate data can be continuously reduced until the first preset number is reduced to 0, or until two or more candidate cameras are found.

[0105] According to an embodiment of the present invention, in step S45, as described above, only 2-3 cameras are needed to locate the coordinates of the drone. Therefore, the most suitable first camera for shooting can be selected from multiple candidate cameras. According to the camera coordinate data, the first coordinate data, the predicted coordinate data, and the third range of the candidate cameras, the preference coefficient of the candidate cameras is determined, including: determining the preference coefficient C of the i-th candidate camera according to formula (2) S,i ,

[0106]

[0107] where L 1,iis the distance between the first coordinate data and the camera coordinate data of the i-th undetermined camera, l 3,i is the minimum plane distance between the plane coordinate corresponding to the first coordinate data and the third range, L 1,i,p,j is the distance between the predicted coordinate data at the future j-th moment and the camera coordinate data of the i-th undetermined camera, l 3,i,p,j is the minimum plane distance between the plane coordinate corresponding to the predicted coordinate data at the future j-th moment and the third range, γ is a coefficient less than 1 and greater than 0, N1 is the first preset quantity, j ≤ N1, and both j and N1 are positive integers.

[0108] According to an embodiment of the present invention, in formula (2), is the ratio of the minimum plane distance between the plane coordinate corresponding to the first coordinate data at the current moment and the third range and the distance between the first coordinate data and the camera coordinate data of the i-th undetermined camera. The larger the minimum plane distance between the plane coordinate corresponding to the first coordinate data at the current moment and the third range, the smaller the probability that the drone will be blocked by an obstacle at the next moment. In other words, the higher the probability of continuously capturing the drone for multiple moments. The smaller the distance between the first coordinate data and the camera coordinate data of the i-th undetermined camera, the clearer the picture of the drone captured by the camera and the less likely it is to be interfered by other objects flying in the air. For example, if there are birds flying near the drone, if the camera is too far from the drone, the size of the drone in the camera's picture is small and difficult to distinguish, and there may be a situation of confusing the birds and the drone and then tracking and photographing the birds. Therefore, the smaller the distance between the first coordinate data and the camera coordinate data of the i-th undetermined camera, the higher the probability that the camera can clearly and stably capture the drone. Therefore, can comprehensively describe the possibility that the i-th camera can continuously and clearly and stably track and photograph the drone for multiple moments. The higher this ratio, the better the tracking effect on the drone.

[0109] According to an embodiment of the present invention, has a meaning similar to and can represent the possibility that the undetermined camera can still continuously and clearly and stably track and photograph the drone for multiple moments at the future j-th moment. The higher this ratio, the better the tracking effect of the undetermined camera on the drone at that moment. However, the predicted coordinate data of the drone in the future is obtained through a position prediction model and may have errors. Therefore, may also have errors. Therefore, when analyzing the possibility that the undetermined camera can still continuously and clearly and stably track and photograph the drone for multiple moments at the future j-th moment at the current moment, this possibility will have a certain discount, and the greater the time gap between the future j-th moment and the current moment, the smaller the discount. That is, a discount needs to be given to Set a smaller coefficient. γ can be used as the discount rate coefficient, which can be the accuracy rate of the trained position prediction model. For example, when validating the trained position prediction model in the validation set, the historical coordinate data in the validation set can be subtracted from the coordinate data predicted by the trained position prediction model to obtain a difference vector, and the norm of the difference vector can be calculated. The ratio of the norm to the norm of the historical coordinate data can be used as the prediction accuracy rate of the trained position prediction model for this coordinate. The prediction accuracy rates of the trained position prediction model for multiple coordinates can be averaged to obtain the accuracy rate of the trained position prediction model, that is, the above-mentioned discount rate coefficient (for example, γ = 0.95). Then, for the current moment, analyze the possibility that the to-be-determined camera at the j-th future moment can continuously and clearly and stably track and photograph the UAV for multiple consecutive moments, that is, discount the possibility that the to-be-determined camera at the j-th future moment can continuously and clearly and stably track and photograph the UAV for j times. It is the sum of the possibilities that the to-be-determined cameras at the first preset number of future moments can continuously and clearly and stably track and photograph the UAV for the current moment. Then it represents the possibility that the to-be-determined camera at each moment can continuously, clearly and stably track and photograph the UAV, which can be used as the preference coefficient of the to-be-determined camera. Similarly, the preference coefficient of each camera can be obtained. The higher the preference coefficient, the more stable the shooting and tracking quality of the camera, and it can be preferentially selected as the first camera.

[0110] According to an embodiment of the present invention, in step S46, 2 or 3 (i.e., the second preset number) to-be-determined cameras with the highest preference coefficients can be selected as the first camera. And in step S47, determine the orientation of the first camera. For example, if it takes a preset time for the first camera to adjust its orientation, the predicted coordinate data of the UAV at the first future moment can be determined, and the direction of the vector between the camera coordinate data of the first camera and the predicted coordinate data can be determined, which is the orientation of the first camera. For example, the yaw angle of the first camera can be taken as the orientation angle of the projection of the vector on the xoy plane, and the pitch angle of the first camera can be taken as the angle between the vector and the xoy plane.

[0111] In this way, the possibility that the to-be-determined camera can continuously, clearly and stably track and photograph the UAV at each moment can be determined by the ratio of the minimum planar distance between the planar coordinates corresponding to the first coordinate data and the third range and the distance between the first coordinate data and the camera coordinate data of the i-th to-be-determined camera, and the ratio of the minimum planar distance between the planar coordinates corresponding to the predicted coordinate data at the future j-th moment and the third range and the distance between the predicted coordinate data at the future j-th moment and the camera coordinate data of the i-th to-be-determined camera. When solving the possibility that the to-be-determined camera can continuously, clearly and stably track and photograph the UAV at future moments, considering the accuracy of the position prediction model, a discount rate is set for the future possibility, so as to improve the accuracy of the preference coefficient, provide an accurate objective standard for screening the first camera, improve the accuracy and stability of tracking and photographing the UAV, and reduce the probability of losing the target.

[0112] According to an embodiment of the present invention, the method further includes: when at least one first camera loses sight of the UAV, or when the first preset number of future moments ends, re-determine the first camera. When the first camera loses sight of the UAV, it is necessary to replace the first camera to photograph the UAV, or when the first preset number of future moments ends, it is expected that the position of the UAV has changed greatly, and the first camera can be re-determined to photograph the UAV. The method of re-determining the first camera is similar to the above and will not be elaborated here.

[0113] According to an embodiment of the present invention, in step S5, if the radar does not lose the target, the first coordinate data determined by the radar is used to determine the flight trajectory of the UAV, so as to determine the risk of the UAV. If the radar loses the target, the coordinates of the UAV in the coordinate system can be calculated using the coordinates of the UAV in the video frame captured by the first camera, and then the flight trajectory of the UAV is determined, so as to determine the risk of the UAV. Among them, the UAV no-fly zone can be an area where UAVs are prohibited from staying or photographing. The longer the UAV stays in the risk area, the more pictures of the UAV no-fly zone it may capture, and the greater the risk of the UAV. And if the flight trajectory of the UAV can surround the boundary of the UAV no-fly zone, the pictures of the UAV no-fly zone captured by the UAV are more comprehensive, and the risk of the UAV is greater.

[0114] According to an embodiment of the present invention, in step S5, according to at least one of the first coordinate data and the video frame captured by the first camera, the duration of the UAV entering the risk area, and the first range of the UAV no-fly zone, determine the risk coefficient of the UAV, including:

[0115] Step S51, if the radar does not lose the target, use the first coordinate data as the positioning data of the UAV;

[0116] Step S52, if the radar loses the target, determine the positioning data of the UAV according to the image coordinates of the UAV in the video frames captured by multiple first cameras and the calibration parameters of the multiple first cameras.

[0117] Step S53, determine the risk coefficient of the UAV according to the positioning data of the UAV at multiple moments after entering the risk area, the duration of the UAV entering the risk area, and the first range of the no-fly zone of the UAV.

[0118] According to an embodiment of the present invention, in step S51, as described above, if the radar does not lose the target, use the more accurate radar positioning data (i.e., the first coordinate data) as the positioning data of the UAV. In step S52, if the radar loses the target, determine the positioning data of the UAV according to the image coordinates of the UAV in the video frames captured by multiple first cameras, the calibration parameters of the multiple first cameras, and the camera coordinate data of the multiple first cameras. For example, each camera can be calibrated before being put into use. For example, a checkerboard calibration board is used for calibration. Triangulation is performed based on the calibration parameters (internal parameters and external parameters) of at least 2 first cameras and the image coordinates of the UAV, so as to obtain the coordinates of the UAV in the coordinate system, that is, the positioning data.

[0119] According to an embodiment of the present invention, in step S53, determine the risk coefficient of the UAV according to the positioning data of the UAV at multiple moments after entering the risk area, the duration of the UAV entering the risk area, and the first range of the no-fly zone of the UAV, including:

[0120] Step S531, determine the centroid plane coordinates of the first range;

[0121] Step S532, obtain the third connection line between the centroid plane coordinates and the plane coordinates corresponding to the positioning data at each moment;

[0122] Step S533, determine the first intersection point between the third connection line and the projection edge of the first range;

[0123] Step S534, obtain the first plane distance between the first intersection point and the plane coordinates corresponding to the positioning data at each moment;

[0124] Step S535, fit the first plane distance corresponding to each moment with each moment to obtain a plane distance function;

[0125] Step S536, obtain the first distance between the first intersection point corresponding to each moment and the first intersection point corresponding to the moment when the radar detects the UAV entering the risk area on the projection edge of the first range;

[0126] Step S537, determine the first ratio of the first distance to the total length of the projection edge of the first range.

[0127] Step S538: Fit the first ratio corresponding to each moment with each moment to obtain an observed ratio function;

[0128] Step S539: Determine the risk coefficient of the UAV according to the observed ratio function, the planar distance function, and the duration of the UAV entering the risk area.

[0129] According to an embodiment of the present invention, in step S531, the centroid plane coordinates of the first range are the centroid of the projection of the first range (spatial range) on the xoy plane ( Figure 2 the range shown by the rectangular dotted line in the figure). In step S532, after the UAV enters the risk area, multiple moments have elapsed. The planar coordinates (i.e., the x-axis coordinate and the y-axis coordinate) corresponding to the positioning data at each moment can be connected to the centroid plane coordinates to obtain a third connection line corresponding to each moment. In step S533, the third connection line intersects the projection edge of the first range, and the coordinates of the first intersection point can be determined. In step S534, the first planar distance can be used as the planar projection distance between the UAV and the UAV no-fly area. In step S535, the first planar distance corresponding to each moment and each moment can be fitted by interpolation or polynomial fitting to obtain a planar distance function, which is used to describe the variation law of the planar projection distance between the UAV and the UAV no-fly area with time.

[0130] Figure 4 Exemplarily, a schematic diagram of the third connection line according to an embodiment of the present invention is shown.

[0131] According to an embodiment of the present invention, the connection line between the planar coordinates corresponding to the positioning data and the centroid plane coordinates is the third connection line. The intersection point of the third connection line and the projection edge of the first range (the edge of the projection of the first range on the xoy plane, i.e., Figure 4 the rectangular dotted line in the figure) is the first intersection point. In step S536, the first distance between the first intersection point at each moment and the first intersection point corresponding to the initial moment when the UAV enters the risk area on the projection edge of the first range can be determined. As Figure 4 shown, the position of the solid rectangular frame labeled 1 represents the planar coordinates corresponding to the positioning data at the initial moment when the UAV enters the risk area, and the position of the solid rectangular frame labeled 2 represents the planar coordinates corresponding to the positioning data at the second moment when the UAV enters the risk area. Based on the above method, the first intersection points corresponding to the two can be obtained, and the first distance between the two first intersection points on the projection edge of the first range is the distance traveled by the first intersection point corresponding to the initial moment along the projection edge to the first intersection point corresponding to the second moment, that is, Figure 4 the total length of the part modified to a solid line in the rectangular dotted line in the figure. Based on a similar method, the first distance corresponding to each moment can be determined.

[0132] According to an embodiment of the present invention, in step S537, a first ratio of the first distance to the total length of the projection edge of the first range can be determined. For example, the ratio of the total length of the part modified to a solid line in the rectangular dotted line to the perimeter of the rectangular frame. Similarly, the first ratio corresponding to each moment can be determined. And in step S538, the first ratio corresponding to each moment is fitted with each moment to obtain an observed ratio function, which is used to describe the variation law of the first ratio with time, and can also be used to describe the variation law of the ratio of the UAV observing the no-fly zone of the UAV with time.

[0133] According to an embodiment of the present invention, in step S539, according to the observed ratio function, the plane distance function, and the duration of the UAV entering the risk area, the risk coefficient of the UAV is determined, including: determining the risk coefficient R of the UAV according to formula (3),

[0134]

[0135] wherein, f o (T) is the observed ratio function, f l (T) is the plane distance function, T is the time variable, T t is the duration of the UAV entering the risk area, r is the radius of the risk area, is the average plane moving speed of the UAV at multiple moments after the UAV enters the risk area, 0 ≤ T ≤ T t , and w1 and w2 are preset weights.

[0136] According to an embodiment of the present invention, in formula (3), at the moment T after the UAV enters the risk area, the plane projection distance between the UAV and the no-fly zone of the UAV can be represented by f l (T), and the ratio of the flight trajectory surrounding the UAV that can surround the no-fly zone of the UAV is the first ratio, which can be used to represent the ratio of the UAV observing the no-fly zone of the UAV, and can be represented by f o (T). The higher f o (T) is, the more comprehensive the UAV's observation of the no-fly zone of the UAV is, and the greater the risk of this UAV is, which can be used to represent the risk of the comprehensiveness of the UAV's observation of the no-fly zone of the UAV. The smaller f l (T) is, the closer the UAV is to the no-fly zone of the UAV, and the clearer the UAV can observe the no-fly zone of the UAV, and the greater the risk of this UAV is. Therefore, it can be represented by the risk of the clarity of the UAV's observation of the no-fly zone of the UAV.

[0137] According to an embodiment of the present invention, both of the above two risks can be accumulated over time. For example, for the risk of the comprehensiveness of the UAV's observation of the no-fly zone, even if the UAV stays at a position without moving, fo (T) remains unchanged, but the drone can still obtain more information about the no-fly zone of the drone at the same position as time increases. For example, it can take more images of the no-fly zone of the drone and understand more details in the images. Therefore, the risk of the comprehensiveness of the observation of the no-fly zone of the drone can accumulate over time. On the other hand, at the first moment, the risk value of the comprehensiveness of the observation of the no-fly zone of the drone is 15%. At the second moment, the risk value of the comprehensiveness of the observation of the no-fly zone of the drone is 20%. At the third moment, the drone moves in the opposite direction, resulting in the risk value of the comprehensiveness of the observation of the no-fly zone of the drone dropping to 13%. This does not mean that the total risk of the comprehensiveness of the observation of the no-fly zone of the drone decreases. Instead, it means that the drone repeatedly observes the same part of the no-fly zone of the drone, that is, the total risk of the comprehensiveness of the observation of the no-fly zone of the drone is increasing. Therefore, the risk of the comprehensiveness of the observation of the no-fly zone of the drone can accumulate over time.

[0138] According to an embodiment of the present invention, on the other hand, the risk of the clarity of the observation of the no-fly zone of the drone also accumulates over time. For example, the flight process of the drone is to first approach the no-fly zone of the drone and then move away from the no-fly zone of the drone, resulting in the risk value of the clarity of the observation of the no-fly zone of the drone increasing first and then decreasing. However, this does not mean that the total risk of the clarity of the observation of the no-fly zone of the drone decreases. Instead, it means that the drone repeatedly observes the no-fly zone of the drone from far to near and then from near to far, and can obtain more images of the no-fly zone of the drone and understand more details in the images, that is, the total risk of the clarity of the observation of the no-fly zone of the drone by the drone is increasing. Therefore, the risk of the clarity of the observation of the no-fly zone of the drone accumulates over time.

[0139] According to an embodiment of the present invention, the instantaneous risk values of the above two risks at the same moment can be weighted and summed. Among them, the two preset weights can be set to equal values, for example, both equal to 0.5, or they can be set to unequal values. The present invention does not limit this. After the weighted summation, the instantaneous value of the observation risk of the no-fly zone of the drone can be obtained. By integrating this instantaneous value, the risk of observing the no-fly zone of the drone since the drone enters the risk area can be obtained. Furthermore, the longest time for the drone to cross the risk area at its average speed is That is, the time for the drone to cross the risk area along the diameter of the risk area. The longer the time for the drone to enter the risk area, the longer the time it can observe the no-fly zone of the drone. The ratio of the duration of the drone entering the risk area to the longest time for the drone to cross the risk area at its average speed. The larger this ratio is, the greater the possibility that the drone is not passing by the risk area but observing the no-fly zone of the drone. When this ratio exceeds 1, it can be considered that the drone deliberately stays in the risk area to observe the no-fly zone of the drone. Therefore, It can be used to describe the possibility that the drone specifically observes the no-fly zone of the drone rather than passing by, that is, to describe the risk of the drone deliberately staying. By multiplying the risk of the drone deliberately staying by the risk of the drone observing the no-fly zone of the drone, the risk coefficient of the drone can be obtained. The higher this risk coefficient is, the more likely it is that the drone deliberately stays in the risk area to observe the no-fly zone of the drone, and the more information is observed. That is, it indicates that the risk of the drone is higher, and it also indicates that the risk of information leakage of the information prohibited from being photographed by the drone is higher. Among them, the ratio of the displacement to the time of the time periods between multiple adjacent moments after the drone enters the risk area can be calculated, that is, the speeds of multiple time periods, and the average of the speeds of multiple time periods is obtained

[0140] In this way, through integration, the risk of the clarity of the drone's observation of the no-fly zone of the drone and the risk of the comprehensiveness of the drone's observation of the no-fly zone of the drone can be processed to obtain the risk of the drone observing the no-fly zone of the drone, and the ratio of the duration of the drone entering the risk area to the longest time for the drone to cross the risk area at its average speed is used to represent the possibility that the drone deliberately stays to observe the no-fly zone of the drone, so as to obtain the risk coefficient of the drone, which can accurately represent the risk that the drone more deliberately stays in the risk area to observe the no-fly zone of the drone and observes more information, and accurately represent the risk of information leakage of the information prohibited from being photographed in the no-fly zone of the drone.

[0141] According to an embodiment of the present invention, in step S6, when the risk coefficient is higher than the preset risk threshold, the alarm level is determined. As the time for the drone to enter the risk area continues to increase and the positions passed by the drone continue to increase, its risk coefficient will become higher and higher until it is higher than the preset risk threshold (for example, 3), then it can be determined that the risk of the drone is relatively high, and its risk coefficient can be updated in real time, and the alarm level can be determined. For example, when the risk coefficient is higher than 3 but less than or equal to 5, it is a low alarm level; when the risk coefficient is higher than 5 but less than or equal to 10, it is a medium alarm level; when the risk coefficient is higher than 10, it is a high alarm level.

[0142] According to an embodiment of the present invention, in step S7, when the risk coefficient is higher than the preset risk threshold, the recorded image of the drone can be selected according to the first coordinate data. Step S7 may include: immediately re-determining the first camera, and determining the re-determined first camera as the second camera to capture the recorded image. That is, the method of selecting the second camera is the same as that of selecting the first camera. A camera that is close to the drone, unobstructed, and is expected to continuously capture the drone at multiple future moments can be selected as the second camera, so as to obtain more recorded images of the drone, and further obtain more information about the drone for further investigation and evidence collection.

[0143] According to an embodiment of the present invention, in step S8, countermeasures can be determined according to the alarm level. For example, in the case where the drone leaves the risk area after reaching the low alarm level at most, the recorded image can be saved for subsequent observation. In the case where the drone leaves the risk area after reaching the medium alarm level at most, the recorded image can be saved, and the relevant management agency of the drone can be reported for investigation, and the operator of the drone can be reminded. In the case where the drone still does not leave the risk area after reaching the high alarm level at most, measures such as radio interference can be taken to prevent it from continuing to fly, or the drone can be driven away, and the recorded image can be saved, the relevant management agency of the drone can be reported for investigation, and the operator of the drone can be reminded, etc. Thus, the safety of the no-fly zone of the drone is improved.

[0144] The low-altitude UAV anti-reversal method based on radar positioning and image recognition technology according to an embodiment of the present invention can, after the UAV enters the risk area and approaches the UAV no-fly zone, detect the coordinate data of the UAV through radar, and when there is a risk of the radar losing the target, promptly select a suitable camera to photograph the UAV, thereby reducing the risk of losing the target. It can also, after the UAV enters the risk area, judge the risk of the UAV in real time and reduce the interference to the passing risk-free UAVs. When judging whether there is a risk of the radar losing the target, it can be judged whether the UAV entering the risk area is leaving the risk area. If the UAV has not left the risk area, it is judged whether the UAV is at a relatively low altitude and the altitude is still decreasing through judgment conditions, and the time required for the UAV to reach the lower limit of the detection altitude of the radar is relatively short, so as to judge whether there is a risk of the radar losing the target, and reserve a calibration duration for the camera to track and photograph the UAV, so as to accurately track the UAV, improve the tracking accuracy, and reduce the probability of losing the target. When selecting the first camera, the ratio of the minimum planar distance between the planar coordinates corresponding to the first coordinate data and the third range and the distance between the first coordinate data and the camera coordinate data of the i-th undetermined camera, and the ratio of the minimum planar distance between the planar coordinates corresponding to the predicted coordinate data at the future j-th moment and the third range and the distance between the predicted coordinate data at the future j-th moment and the camera coordinate data of the i-th undetermined camera can be used to determine the possibility that the undetermined camera at each moment can continuously, clearly and stably track and photograph the UAV. When solving the possibility that the undetermined camera at the future moment can continuously, clearly and stably track and photograph the UAV, considering the accuracy of the position prediction model, a discount rate is set for the future possibility, so as to improve the accuracy of the preference coefficient, provide an accurate objective standard for screening the first camera, improve the accuracy and stability of tracking and photographing the UAV, and reduce the probability of losing the target. When determining the risk coefficient of the UAV, the risks of the UAV's observation clarity of the UAV no-fly zone and the risks of the UAV's observation comprehensiveness of the UAV no-fly zone can be processed by integration to obtain the risk of the UAV observing the UAV no-fly zone, and the possibility of the UAV deliberately staying to observe the UAV no-fly zone is represented by the ratio of the duration of the UAV entering the risk area to the longest time required for the UAV to cross the risk area at its average speed, so as to obtain the risk coefficient of the UAV, which can accurately represent the risk that the UAV is more deliberately staying in the risk area to observe the UAV no-fly zone and observing more information, and accurately represent the risk of information leakage of the prohibited shooting information in the UAV no-fly zone.

[0145] The present invention can be a method, device, system and / or computer program product. The computer program product can include a computer-readable storage medium having thereon computer-readable program instructions for performing various aspects of the present invention.

[0146] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and illustrated in the embodiments. Without departing from the said principles, the embodiments of the present invention can be deformed or modified in any way.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or equivalently replace some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A low-altitude UAV anti-reflection method based on radar positioning and image recognition technologies, characterized in that, Including: Taking the radar as the coordinate origin, establishing a coordinate system, and setting a no-fly zone and a risk zone for unmanned aerial vehicles (UAVs) in the coordinate system. Among them, the no-fly zone for UAVs is located within the risk zone, and the second range of the risk zone is larger than the first range of the no-fly zone for UAVs; When the radar detects that a UAV enters the risk zone, obtaining the first coordinate data of the UAV in the coordinate system; According to the first coordinate data, determining whether there is a risk of the radar losing the target; If there is a risk of the radar losing the target, then according to the first coordinate data and the camera coordinate data of multiple cameras, determining the first camera to be turned on and the orientation of the first camera, where the multiple cameras are set at multiple positions outside the no-fly zone for UAVs and within the risk zone; According to at least one of the first coordinate data and the video frames captured by the first camera, the duration of the UAV entering the risk zone, and the first range of the UAV no-fly zone, determining the risk coefficient of the UAV; In the case where the risk coefficient is higher than a preset risk threshold, determining the alarm level; According to the first coordinate data, selecting a second camera to capture a record image of the UAV; According to the alarm level, determining countermeasures; 2. The anti - reverse method for low - altitude UAVs based on radar positioning and image recognition technology according to claim 1, wherein, According to the first coordinate data, determining whether there is a risk of the radar losing the target, including: Determine the planar coordinates (x t , y t ) of the drone at the current moment and the planar coordinates (x t-1 , y t-1 ) at the previous moment according to the first coordinate data, where the current moment is the t-th moment after the drone enters the risk area; If obtain the altitude coordinate z of the UAV at the current moment t and the altitude coordinate z at the previous moment t-1 ; According to the formula Determine judgment conditions C1, C2, and C3, where h ll is the lower limit of the detection height of the radar, δ1 is a coefficient greater than 1, Δt is the time difference between the current moment and the previous moment, and n d is the number of moments corresponding to the preset camera calibration duration; When the judgment conditions C1, C2, and C3 are all satisfied simultaneously, determining that there is a risk of the radar losing the target; 3. The low-altitude UAV anti-reflection method based on radar positioning and image recognition technology according to claim 1, characterized in that, If there is a risk of the radar losing the target, then according to the first coordinate data and the camera coordinate data of multiple cameras, determining the first camera to be turned on and the orientation of the first camera, including: Inputting the first coordinate data at multiple moments after the UAV enters the risk zone into a trained position prediction model to obtain predicted coordinate data for the first preset number of future moments; Determining the first connection line between the plane coordinates corresponding to the camera coordinate data and the plane coordinates corresponding to the first coordinate data, and determining the second connection line between the plane coordinates corresponding to the camera coordinate data and the plane coordinates corresponding to the predicted coordinate data; Determining the third range of the obstacle area within the risk zone, where the third range includes the first range; Among the multiple cameras, determining the to-be-determined cameras for which both the first connection line and the second connection line have no intersection with the third range; According to the camera coordinate data of the to-be-determined cameras, the first coordinate data, the predicted coordinate data, and the third range, determining the preference coefficient of the to-be-determined cameras; According to the preference coefficients of the respective to-be-determined cameras, determining the first camera; According to the camera coordinate data of the first camera and the first coordinate data, determining the orientation of the first camera; 4. The anti-reversal method for low-altitude unmanned aerial vehicles based on radar positioning and image recognition technology according to claim 3, wherein According to the camera coordinate data of the to-be-determined cameras, the first coordinate data, the predicted coordinate data, and the third range, determining the preference coefficient of the to-be-determined cameras, including: According to the formula Determine the optimization coefficient C of the i-th to-be-determined camera S,i , where L 1,i is the distance between the first coordinate data and the camera coordinate data of the i-th to-be-determined camera, and l 3,i is the minimum planar distance between the planar coordinate corresponding to the first coordinate data and the third range, and L 1,i,p,j is the distance between the predicted coordinate data at the future j-th moment and the camera coordinate data of the i-th to-be-determined camera, and l 3,i,p,j is the minimum planar distance between the planar coordinate corresponding to the predicted coordinate data at the future j-th moment and the third range, γ is a coefficient less than 1 and greater than 0, N1 is the first preset quantity, j ≤ N1, and both j and N1 are positive integers.

5. The anti-reverse method for low-altitude unmanned aerial vehicles based on radar positioning and image recognition technology according to claim 1, characterized in that, According to at least one of the first coordinate data and the video frames captured by the first camera, the duration of the UAV entering the risk zone, and the first range of the UAV no-fly zone, determining the risk coefficient of the UAV, including: If the radar does not lose the target, the first coordinate data is used as the positioning data of the UAV; If the radar loses the target, the positioning data of the UAV is determined according to the image coordinates of the UAV in the video frames captured by multiple first cameras and the calibration parameters of the multiple first cameras; The risk coefficient of the UAV is determined according to the positioning data of the UAV at multiple moments after entering the risk area, the duration of the UAV entering the risk area, and the first range of the no-fly zone of the UAV.

6. The anti-reversal method for low-altitude unmanned aerial vehicles based on radar positioning and image recognition technology according to claim 5, wherein Determining the risk coefficient of the UAV according to the positioning data of the UAV at multiple moments after entering the risk area, the duration of the UAV entering the risk area, and the first range of the no-fly zone of the UAV includes: Determine the centroid plane coordinates of the first range; Obtain the third connection line between the centroid plane coordinates and the plane coordinates corresponding to the positioning data at each moment; Determine the first intersection point of the third connection line and the projection edge of the first range; Obtain the first plane distance between the first intersection point and the plane coordinates corresponding to the positioning data at each moment; Fit the first plane distances corresponding to each moment with each moment to obtain a plane distance function; Obtain the first distance between the first intersection point corresponding to each moment and the first intersection point corresponding to the moment when the radar detects the UAV entering the risk area on the projection edge of the first range; Determine the first ratio of the first distance to the total length of the projection edge of the first range; Fit the first ratios corresponding to each moment with each moment to obtain an observation ratio function; Determine the risk coefficient of the UAV according to the observation ratio function, the plane distance function, and the duration of the UAV entering the risk area.

7. The low-altitude UAV anti-reversal method based on radar positioning and image recognition technology according to claim 6, characterized in that, Determining the risk coefficient of the UAV according to the observation ratio function, the plane distance function, and the duration of the UAV entering the risk area includes: According to the formula Determine the risk coefficient R of the UAV, where f o (T) is the observation ratio function, f l (T) is the planar distance function, T is the time variable, T t is the duration for the UAV to enter the risk area, r is the radius of the risk area, is the average planar movement speed of the UAV at multiple moments after the UAV enters the risk area, 0 ≤ T ≤ T t , w1 and w2 are preset weights.

8. The anti-reverse method for low-altitude unmanned aerial vehicles based on radar positioning and image recognition technology according to claim 3, wherein The method further includes: When at least one first camera loses sight of the UAV, or at the end of the first preset number of future moments, re-determine the first camera.

9. The anti - reflection method for low - altitude UAVs based on radar positioning and image recognition technology according to claim 8, wherein, Selecting a second camera to capture a record image of the UAV according to the first coordinate data includes: Immediately re-determine the first camera, and determine the re-determined first camera as the second camera to capture the record image.

Images