Anti-drone investigation method and device, electronic equipment and computer readable medium

Abstract

Embodiments of the present disclosure disclose anti-UAV detection methods, devices, electronic equipment and computer readable media. A specific embodiment of the method comprises: performing object feature extraction on millimeter wave radar data collected for a preset electronic fence airspace to generate an object feature information set; performing radio frequency signal detection to generate an object radio frequency information set; determining object radio frequency information in the object radio frequency information set that meets a preset signal screening condition as candidate radio frequency information to obtain a candidate radio frequency information set; performing infrared detection to obtain an infrared detection result set; performing interference elimination on object feature information in the object feature information set that represents non-UAV targets to generate a UAV feature information set; performing anti-UAV interference based on the UAV feature information set, and performing anti-UAV tracing to generate a UAV detection result set. This embodiment can improve the detection accuracy of UAV targets.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, particularly to the fields of drone signal recognition and drone control, and specifically to anti-drone reconnaissance methods, apparatus, electronic devices, and computer-readable media. Background Technology

[0002] With the rapid development and widespread adoption of drone technology, the problem of drone misuse (e.g., controlling drones to fly into no-fly zones) has become increasingly prominent. Currently, various anti-drone identification technologies exist, such as using lidar and visible light cameras to detect drone targets during anti-drone reconnaissance, and interfering with drones by continuously operating signal jamming. However, these methods are susceptible to environmental interference, leading to misidentification of targets (e.g., mistaking birds for drones). Summary of the Invention

[0003] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0004] Some embodiments of this disclosure provide counter-drone reconnaissance methods, apparatuses, electronic devices, and computer-readable media to address the technical problems mentioned in the background section above.

[0005] In a first aspect, some embodiments of this disclosure provide a method for countering unmanned aerial vehicle (UAV) reconnaissance. The method includes: extracting object features from millimeter-wave radar data collected against a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes object location coordinates; performing radio frequency (RF) signal reconnaissance based on the object feature information group to generate an object RF information group, wherein the object RF information includes the type of object signal; and determining the object RF information in the object RF information group that meets preset signal filtering conditions as candidate RF information to obtain a candidate RF information group, wherein the preset signal filtering conditions include the object RF information containing a table... The system identifies targets that have not issued signals of a preset communication type, including positioning, image transmission, and control types. Infrared detection is performed based on the candidate radio frequency information group to obtain an infrared detection result group, where the infrared detection results characterize the infrared features of the detected target. Using the infrared detection result group, interference is removed from the target feature information group that represents non-UAV targets to generate a UAV feature information group. Anti-UAV interference is performed based on the UAV feature information group, and anti-UAV tracing is performed using the UAV feature information group and the target radio frequency information group to generate a UAV reconnaissance result group.

[0006] Secondly, some embodiments of this disclosure provide an anti-drone reconnaissance device, comprising: a drone feature extraction unit configured to extract object features from millimeter-wave radar data collected against a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes object location coordinates; a radio frequency signal detection unit configured to perform radio frequency signal detection based on the object feature information group to generate an object radio frequency information group, wherein the object radio frequency information includes the type of object signal; and a determination unit configured to determine the object radio frequency information in the object radio frequency information group that meets preset signal filtering conditions as candidate radio frequency information, thereby obtaining a candidate radio frequency information group, wherein the preset signal filtering conditions include the object radio frequency information including The system includes: an identifier for an object signal that has not emitted a preset communication type, which includes positioning, image transmission, and control types; an infrared detection unit configured to perform infrared detection based on the aforementioned candidate radio frequency information group to obtain an infrared detection result group, wherein the infrared detection result represents the infrared characteristics of the detected target; an interference elimination unit configured to use the aforementioned infrared detection result group to eliminate interference from the object feature information representing non-UAV targets in the aforementioned object feature information group to generate a UAV feature information group; and an anti-UAV jamming unit configured to perform anti-UAV jamming based on the aforementioned UAV feature information group and to perform anti-UAV tracing using the aforementioned UAV feature information group and the aforementioned object radio frequency information group to generate a UAV reconnaissance result group.

[0007] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0008] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0009] The various embodiments of this disclosure have the following beneficial effects: the anti-drone reconnaissance methods of some embodiments of this disclosure can improve the accuracy of target reconnaissance. Here, the reason for misjudgment of detected targets is that drone target detection using methods such as lidar and visible light cameras is easily affected by environmental interference. Based on this, the anti-drone reconnaissance methods of some embodiments of this disclosure first extract object features from millimeter-wave radar data collected against a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes the object's position coordinates. Here, compared to lidar detection, by introducing millimeter-wave radar, utilizing its stronger penetration capability in adverse weather conditions, the stability of anti-drone reconnaissance under different weather conditions is improved, thereby reducing the impact of weather conditions on the detection results. Then, radio frequency signal reconnaissance is performed based on the above object feature information group to generate an object radio frequency information group, wherein the object radio frequency information includes the type of object signal. Next, the radio frequency (RF) information of objects meeting the preset signal filtering conditions in the aforementioned object RF information group is identified as candidate RF information, resulting in a candidate RF information group. The preset signal filtering conditions include that the object RF information contains an object signal identifier indicating that it has not emitted a preset communication type. The preset communication types include: positioning type, image transmission type, and control type. Here, considering that millimeter-wave radar has stronger penetration capabilities but relatively lower resolution, RF signal reconnaissance is performed to further filter the detected targets. Then, infrared detection is performed based on the aforementioned candidate RF information group to obtain an infrared detection result group, where the infrared detection results represent the infrared characteristics of the detected target. Finally, using the aforementioned infrared detection result group, interference is removed from the object feature information group representing non-UAV targets in the aforementioned object feature information group to generate a UAV feature information group. Here, considering the possibility of false detections (e.g., misidentifying birds as drones), and the significant differences between the infrared characteristics of drones and birds, infrared detection is introduced on top of the aforementioned millimeter-wave radar + radio frequency signal reconnaissance to further distinguish suspicious targets (i.e., candidate radio frequency information), thereby further eliminating interference. In practice, considering the small features of the target in images captured by visible light cameras, leading to a high risk of false detection, this application extracts the infrared characteristics of the target through infrared detection. Furthermore, based on the differences in infrared characteristics between the targets (e.g., birds and drones), more accurate target differentiation can be achieved. This further improves the stability of reconnaissance under various weather conditions. Finally, anti-drone jamming is performed based on the aforementioned drone feature information group, and anti-drone tracing is performed using the aforementioned drone feature information group and the aforementioned target radio frequency information group to generate a drone reconnaissance result group. This improves the accuracy of drone target detection, facilitating precise anti-drone jamming. Attached Figure Description

[0010] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0011] Figure 1 This is a flowchart of some embodiments of the anti-drone reconnaissance method according to the present disclosure; Figure 2 This is a schematic diagram of morphological feature extraction.

[0012] Figure 3 These are schematic diagrams illustrating the structure of some embodiments of the anti-drone reconnaissance device according to this disclosure; Figure 4 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0013] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0014] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0015] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0018] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] Figure 1 A flow 100 of some embodiments of the anti-drone reconnaissance method according to the present disclosure is shown. The anti-drone reconnaissance method includes the following steps: Step 101: Extract object features from millimeter-wave radar data collected in the preset electronic fence airspace to generate object feature information groups.

[0020] In some embodiments, the entity executing the anti-drone reconnaissance method (e.g., a computing device) can extract object features from millimeter-wave radar data collected over a pre-defined electronic fence airspace via wired or wireless means to generate an object feature information set, wherein the object feature information includes object location coordinates. Here, the electronic fence airspace can be a pre-defined area that restricts drone flight. Furthermore, the object feature information can characterize the detected object suspected to be a drone.

[0021] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (Ultra Wide Band) connections, and other currently known or future wireless connection methods.

[0022] It should be noted that the aforementioned computing devices can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed on the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0023] In some optional implementations of certain embodiments, the aforementioned executing entity extracts object features from millimeter-wave radar data collected in a preset electronic fence airspace to generate an object feature information group, including: Step S1 involves preprocessing the millimeter-wave radar data to obtain processed radar data. First, random noise in the millimeter-wave radar data is removed using a median filtering algorithm (3×3 window size). Then, Kalman filtering is used to smooth fluctuations in the millimeter-wave radar echo signal, resulting in smoothed radar data. Next, the smoothed radar data is converted into a 3D point cloud format, while redundant data outside the electronic fence airspace is removed, yielding the processed radar data. Here, the 3D point cloud format may include horizontal, vertical, and timestamp coordinates.

[0024] Step S2 involves performing target tracking and detection on the processed radar data to generate a detection identifier group and a corresponding object position coordinate group. First, the processed radar data is segmented using the CA-CFAR (Cell Averaging Constant False Alarm Rate Algorithm) algorithm to obtain multiple tracking targets. Here, the false alarm rate can be set to 10. -6 This process is designed to eliminate the influence of ground clutter from the data. Then, Kalman filtering is used to predict the trajectory of each tracked target, resulting in a set of target trajectories. Here, the state model of the tracked target can be represented using a Constant Acceleration Model (CA). Simultaneously, a detection label is assigned to each tracked target using the Joint Probabilistic Data Association (JPDA) algorithm to avoid target overlap. Finally, the polar coordinate parameters (distance, azimuth, and elevation angles) of the tracked target are transformed to the camera coordinate system of the high-resolution infrared device to obtain the object's position coordinates. Thus, the detection label set and object position coordinate set corresponding to each tracked target can be obtained. Here, the object position coordinates can be the position coordinates of the tracked target corresponding to the current timestamp.

[0025] Step S3: Based on the processed radar data and the object position coordinate group, feature extraction is performed on the detected targets corresponding to each detection marker in the detection marker group to obtain a target feature information group. Specifically, for each piece of processed radar data, the following steps are performed: First, the instantaneous velocity, acceleration, and heading angle change rate of the target are calculated as target feature information based on the object position coordinates corresponding to the processed radar data at multiple time points.

[0026] Step S4 involves performing object recognition on the feature information in the aforementioned target feature information group to generate an object feature information group. Specifically, the target feature information can be input into a pre-trained SVM (Support Vector Machine) model for UAV recognition to obtain target confidence. Then, the detection identifier corresponding to a target confidence score greater than a preset confidence threshold is determined as the object identifier. Finally, the object identifier, its corresponding target feature information, and the object's current position coordinates are determined as the object feature information.

[0027] In practice, extracting object features from millimeter-wave radar data using the methods described above can be used to initially eliminate interference. However, considering the small size of the detected targets and the insufficient accuracy of millimeter-wave radar data, false detections are still possible. Therefore, further screening is required.

[0028] Step 102: Perform radio frequency signal detection based on the object feature information group to generate the object radio frequency information group.

[0029] In some embodiments, the execution entity may perform radio frequency signal detection based on the object feature information group to generate an object radio frequency information group. The object radio frequency information includes the type of the object signal.

[0030] In some optional implementations of certain embodiments, the execution entity performs radio frequency signal detection based on the object feature information group to generate an object radio frequency information group, including: Step S1 involves sending a communication request to the detection object corresponding to each object feature information in the aforementioned object feature information group to obtain the object identity identifier, thus obtaining an object identity identifier group. The communication request can be a request to verify the object identity identifier. Here, after sending a communication request to the detection object, if no response including the object identity identifier is received within a preset time (e.g., 5 seconds), an empty object identity identifier is generated. An empty object identity identifier indicates that the object does not possess a tag permitting flight within the electronic fence airspace, and is therefore a violating object. Furthermore, after receiving the object identity identifier, further authentication can be performed to ensure its accuracy. Thus, drones permitted to fly within the electronic fence airspace can be identified. Therefore, the remaining detection targets can be further detected.

[0031] Step S2: The object feature information in the above object feature information group that has an empty object identity identifier is identified as the detection target information, thus obtaining the detection target information group.

[0032] Step S3: For each piece of detection target information in the above detection target information group and the object position coordinates included in the above detection target information, perform the following first processing step: The first step involves controlling the signal detection device to collect signals from the location coordinates of the aforementioned object, thereby obtaining a communication signal group. Specifically, the signal detection device can be controlled to collect signals directionally from the location coordinates of the object to obtain the communication signal group.

[0033] The second step involves preprocessing the aforementioned communication signal group to obtain the processed communication signal group. Firstly, a variational mode decomposition algorithm can be used to preprocess the communication signal to filter out background noise and co-channel interference, resulting in the processed communication signal group.

[0034] The third step involves verifying the consistency between the processed communication signal group and the detected target information to obtain a verification result. This can be achieved by determining the target distance value corresponding to each processed communication signal in the processed communication signal group using signal strength ranging. Then, the distance difference between the target distance value and the object's location coordinates is determined. Next, the processed communication signal with the smallest distance difference (less than or equal to a preset distance threshold) is selected as the target communication signal. Finally, the signal type corresponding to the target communication signal is detected to determine the type of object signal it belongs to. Here, the object signal type can include positioning, image transmission, and control types. If the object signal type corresponding to the detected target information does not include the positioning type, a verification result indicating a failed verification is generated. If the distance difference is greater than the preset distance threshold, it is determined that the detected target information and the processed communication signal correspond to different targets, and a verification result indicating a failed verification is also generated. Conversely, if the distance difference is less than or equal to the preset distance threshold and the smallest distance difference is found, it indicates that the target communication signal and the processed communication signal correspond to the same object, thus generating a verification result indicating a passed verification. If the type of the object signal corresponding to the detected target information includes positioning type (and / or image transmission type, control type), then a verification result indicating that the characterization verification has passed is generated.

[0035] In practice, to avoid further signal interference, directional signal detection equipment is used for signal collection. While this significantly reduces interference from other signals, multiple signal sources may still exist in the same direction within the scene, leading to the collection of multiple communication signals by the directional signal. Therefore, in the consistency verification step described above, identifying the signal distance can be used to accurately filter out objects matching the target information, provided that the target signal has been collected. Furthermore, if the verification fails, it indicates that the target unit has not transmitted a signal. Therefore, further screening is needed to accurately filter out non-UAV units.

[0036] The fourth step involves generating an object signal identifier that represents the preset communication type emitted by the target, in response to the verification result indicating that the verification has passed. This object signal identifier, along with the aforementioned object location coordinates, is then used to determine the object's radio frequency information. The corresponding object signal type can be used as the object signal identifier.

[0037] The fifth step is to generate an object signal identifier indicating that the verification failed, which indicates that the target did not emit a preset communication type, and to determine the object signal identifier and the above-mentioned object position coordinates as object radio frequency information.

[0038] In practice, drones need to transmit and receive various signals during flight. Positioning signals are used for drone positioning and route navigation; image transmission signals are used to transmit images captured by the drone to the control terminal; and control signals allow the drone to receive control signals from the control terminal. Therefore, based on signal identification, an object emitting positioning (and / or image transmission and control) signals can be identified as a drone. However, some drones do not need to connect to a positioning system but instead navigate using a pre-programmed route and their own inertial measurement unit. Thus, they do not need to transmit positioning signals, meaning no positioning-type communication signals are detected, and they also do not transmit any of the signal types corresponding to their intended use. Therefore, for objects whose verification results indicate failure, further refinement using infrared detection is still necessary.

[0039] Step 103: Select the radio frequency information of objects that meet the preset signal screening conditions in the object radio frequency information group as candidate radio frequency information to obtain the candidate radio frequency information group.

[0040] In some embodiments, the execution entity may determine the object radio frequency information in the object radio frequency information group that meets the preset signal filtering conditions as candidate radio frequency information to obtain a candidate radio frequency information group. The preset signal filtering conditions are that the object radio frequency information includes an object signal identifier that does not transmit a preset communication type. The preset communication type includes: positioning type, image transmission type and control type.

[0041] Step 104: Perform infrared detection based on the candidate radio frequency information group to obtain the infrared detection result group.

[0042] In some embodiments, the aforementioned execution entity can perform infrared detection based on the aforementioned candidate radio frequency information group to obtain an infrared detection result group. Specifically, infrared detection information whose matching degree with the candidate radio frequency information exceeds a preset matching degree can be selected from the database by looking up a table, and this information is used as the infrared detection result.

[0043] In some optional implementations of certain embodiments, the execution entity performs infrared detection based on the candidate radio frequency information group to obtain an infrared detection result group, including: For each candidate radio frequency information in the above candidate radio frequency information group, perform the following second processing step: Step S1 involves controlling a high-resolution infrared device to acquire images of the area containing the object's location coordinates, as included in the candidate radio frequency information, to obtain a high-resolution infrared image. First, the azimuth and elevation angles are generated using the direction from the high-resolution infrared device (high-resolution infrared camera) to the object's location coordinates. Then, the distance between the high-resolution infrared device and the object's location coordinates is used as the observation distance. Next, the azimuth, elevation, and observation distance are used as parameters to control the pan-tilt unit of the high-resolution infrared device for target calibration. Target calibration may also include adjusting the angular distance so that the target occupies more than a preset proportion in the image (e.g., 5%). This facilitates the observation of the target's heat source details. Finally, the high-resolution infrared device acquires images of the object's location coordinates, as included in the candidate radio frequency information, to obtain a high-resolution infrared image.

[0044] Step S2 involves extracting morphological features from the high-resolution infrared image. First, median filtering removes salt-and-pepper noise from the high-resolution infrared image. Then, an adaptive histogram equalization algorithm enhances the contrast of the high-resolution infrared image, improving target contour features and preventing blurring due to similar ambient temperatures. Next, an adaptive thresholding algorithm combined with a morphological algorithm segments the processed infrared image, resulting in segmented target region groups. Then, a feature point extraction algorithm extracts the coordinates of key points in each segmented target region, resulting in key point coordinate groups. Finally, these key point coordinate groups are used to define the morphological features of a single frame.

[0045] As an example, feature point extraction algorithms may include, but are not limited to, at least one of the following: SIFT (Scale-Invariant Feature Transform) algorithm, FAST corner detection, BRIEF descriptor, etc. See also... Figure 2 The diagram illustrates morphological feature extraction. The region containing the target object 202 (taking a drone as an example) in the processed infrared image 201 is segmented using an adaptive threshold segmentation algorithm combined with a morphological algorithm, resulting in segmented target region group 203. Then, a feature point extraction algorithm is used to extract the coordinates of key points in each segmented target region, resulting in key point coordinate group 204: [(x1, y1), (x2, y2), (x3, y3), (x4, y4), (x5, y5), (x6, y6)]. Here, the number of key point coordinates is illustrative; more key point coordinates can be detected.

[0046] Step S3 involves identifying heat source features in the high-resolution infrared image to generate heat source features. Specifically, the high-resolution infrared image is segmented according to the segmented target region to obtain segmented infrared regions. Then, the heat values ​​at the locations of the multiple key points within the segmented infrared regions are determined as heat source features.

[0047] Step S4: Based on the aforementioned morphological features and heat source features, an infrared detection marker is generated as the infrared detection result. This infrared detection marker is used to characterize whether the detected target is a drone. Next, the aforementioned morphological features and heat source features are normalized to obtain normalized morphological features and normalized heat source features. Here, the normalization interval can be [0, 1]. Then, the coordinate values ​​of each key point and the corresponding heat value are arranged in the same order to form a one-dimensional feature vector. Finally, the one-dimensional feature vector is input to a preset lightweight multilayer perceptron to generate the infrared detection result. Here, the lightweight multilayer perceptron may include an input layer, one hidden layer, and an output layer. Furthermore, the dimension of the hidden layer is less than three times the dimension of the input layer to avoid model redundancy. A linear rectified function is used as the activation function.

[0048] Optionally, infrared detection markers can be generated through the following steps: First, the coordinates of each key point in the morphological features can be transformed from the image coordinate system to the camera coordinate system, resulting in a transformed coordinate set. Here, the transformed key points in the transformed coordinate set can be connected according to a preset key point order to obtain a key point vector set. Then, error statistics are performed on the key point vector sets between consecutive frames to obtain the morphological error value. Here, the error values ​​corresponding to two adjacent frames can be used as the basic error value, and the mean of each basic error value is determined as the morphological error value. Specifically, the sum of the differences between the corresponding key point vectors between two frames can be determined as the error value between two adjacent frames. Second, the standard deviation and extreme value difference of each heat value in the above heat source features are determined. Here, the larger the standard deviation of heat, the more uneven the distribution of heat difference at each key point coordinate position. The larger the extreme value difference of heat, the greater the difference in heat difference at each key point coordinate position. Finally, in response to the determination that the morphological error value is less than a preset error threshold, the thermal standard deviation is greater than a preset standard deviation threshold, or the extreme value of the thermal difference is greater than a preset extreme value, an infrared detection label characterizing the detected target as a drone is generated. In response to the determination that the morphological error value is greater than or equal to a preset error threshold, the thermal standard deviation is less than or equal to a preset standard deviation threshold, and the extreme value of the thermal difference is less than a preset extreme value, an infrared detection label characterizing the detected target as not being a drone is generated. Here, the determination that the detected target is not a drone requires all three of the above conditions to be met simultaneously; otherwise, the detected target is determined to be a drone.

[0049] Furthermore, drones vary greatly in type; for example, a drone may have 1 to 18 propellers, while a bird has only one pair of wings. Therefore, after extracting the thermal key point structures of drones and birds, a clear structural comparison can be made to classify drone units into non-drone units.

[0050] In practice, considering that non-drone units (such as birds) rely on wing flapping and have relatively uniform heat source distribution, while drone units rely on fixed-wing propellers and have a more concentrated heat source distribution—that is, heat sources exist only at the propulsion system location, with no significant heat sources in other areas of the propeller—heat source feature identification and morphological feature extraction can effectively distinguish whether a target is a drone. Specifically, considering that there is a certain detection error in the keypoint coordinates of a single frame, and that the morphological changes of fixed drone units are much smaller than those of non-drone units, generating a base error value for consecutive frame images can not only determine whether there are significant morphological changes in the target in consecutive frame images, but also eliminate the detection error of individual keypoint coordinates to a certain extent. Furthermore, because morphological feature extraction and heat source feature identification share the same keypoint coordinates, the calibration error between features is also avoided. Therefore, the accuracy of target identification can be greatly improved.

[0051] Step 105: Using the infrared detection result group, remove the object feature information representing non-UAV targets from the object feature information group to generate the UAV feature information group.

[0052] In some embodiments, the execution entity may use the infrared detection result group to remove interference from the object feature information representing non-UAV targets in the object feature information group in order to generate a UAV feature information group.

[0053] In some optional implementations of certain embodiments, the execution entity utilizes the infrared detection result group to remove interference from the object feature information representing non-UAV targets in the object feature information group to generate a UAV feature information group, including: Step S1: Identify the infrared detection markers representing non-UAVs included in the infrared detection result group as interference markers, thus obtaining the interference marker group. The interference markers represent non-UAV units.

[0054] Step S2 involves removing object feature information from the aforementioned object feature information group that meets the interference removal conditions, resulting in a UAV feature information group. The UAV feature information represents the corresponding detection target as a UAV. The interference removal conditions include at least one of the following: the object identity identifier corresponding to the object feature information is not empty; or the object feature information matches an interference identifier in the aforementioned interference identifier group. Here, matching the object feature information with an interference identifier in the interference identifier group indicates that the object feature information is a falsely detected non-UAV unit, and therefore it is removed. Furthermore, removing object feature information with corresponding object identity identifiers ensures that interference is only applied to UAVs with empty object identity identifiers.

[0055] Step 106: Perform anti-drone interference based on the drone feature information group, and perform anti-drone tracing using the drone feature information group and the target radio frequency information group, so as to generate a drone reconnaissance result group.

[0056] In some embodiments, the aforementioned executing entity may perform anti-drone interference based on the aforementioned drone feature information group, and perform anti-drone tracing using the aforementioned drone feature information group and the aforementioned object radio frequency information group, so as to generate a drone reconnaissance result group.

[0057] In some optional implementations of certain embodiments, the execution entity performs anti-drone interference based on the aforementioned drone feature information group and performs anti-drone tracing using the aforementioned drone feature information group and the aforementioned target radio frequency information group to generate a drone reconnaissance result group, including: For each drone feature in the above drone feature information group, perform the following interference steps: Step S1 involves controlling a radio jamming device matching the aforementioned UAV characteristic information to interfere with the signal of the UAV corresponding to the aforementioned UAV characteristic information, and controlling a high-resolution infrared device to acquire an image of the UAV corresponding to the aforementioned UAV characteristic information, thereby obtaining an image of the UAV after interference. First, the radio jamming device closest to the target's location coordinates corresponding to the UAV characteristic information can be selected as the matching radio jamming device. Then, the radio jamming device is activated according to preset frequency band parameters and preset frequency parameters to interfere with the signal of the UAV corresponding to the aforementioned UAV characteristic information. Here, by acquiring an image of the UAV under signal interference, the UAV's reaction state after being interfered with can be determined. For example, the target UAV may return to base or land after being interfered with.

[0058] Step S2 involves performing drone attitude recognition on the disturbed drone image to generate a drone attitude identifier. This involves using the aforementioned keypoint extraction algorithm to extract the coordinates of each keypoint in the disturbed drone image, obtaining a current keypoint coordinate set. Next, the minimum bounding rectangle of each current keypoint coordinate in the current keypoint coordinate set is determined. Then, the orientation vector of the minimum bounding rectangle is determined as the current drone heading vector. Finally, based on the current drone heading vector and parameters such as the drone's velocity, acceleration, and altitude over a continuous time period, a drone attitude identifier is generated according to preset judgment conditions. Here, the drone attitude identifier can include: an attitude identifier representing normal drone flight, an attitude identifier representing drone return-to-home, and an attitude identifier representing drone landing.

[0059] As an example, the criteria for determining normal flight attitude may include: [The angle between the current UAV heading vector and all historical heading vectors within 3 seconds is ≤ 10° (i.e., no significant deviation in heading, and the direction remains consistent); and the overall change in the heading vector within 3 seconds (the difference between the maximum and minimum angles) is ≤ 15°, with no directional turning trend. The average speed within 3 seconds is within the UAV's preset cruise speed range (e.g., 5-15 m / s), and the instantaneous speed fluctuation is ≤ ±2 m / s (no sudden acceleration or deceleration); the absolute value of the vertical speed is ≤ 0.5 m / s (no significant altitude gain or loss). The altitude change within 3 seconds is ≤ ±1 m, and the altitude fluctuation trend is stable, with no continuous upward or downward characteristics]. The criteria for determining the return attitude indicator may include: [The angle between the current heading vector and the heading vector at the initial moment within 3 seconds is ≥ 30°, and the heading vector gradually approaches the "preset return point direction" within 3 seconds (the angle between the heading vector and the return point direction decreases by ≥ 5° every 0.5 seconds); the angle between the heading vectors at any adjacent moment within 3 seconds is ≤ 15°, and the turning is smooth without sudden changes. The average horizontal speed is stable at 5-8 m / s (return cruise speed) within 3 seconds, and the instantaneous speed fluctuation is ≤ ±1.5 m / s; the absolute value of the vertical speed is ≤ 1 m / s, and the altitude remains stable or slowly decreases (descent rate ≤ 1 m / s), without drastic ascent or descent. The altitude change within 3 seconds is ≤ ±3 meters, and the overall state is "small fluctuation + slow descent" or "completely stable", without continuous ascent or rapid descent characteristics]. The criteria for determining landing attitude can include: [The angle between the current heading vector and all historical heading vectors within 3 seconds is ≤ 15° (the heading is basically fixed, without significant turning); there is no convergence trend toward a specific return point, and the heading remains relatively stationary or slightly adjusted. The horizontal speed shows a continuous decreasing trend within 3 seconds, with the average horizontal speed gradually decreasing from the cruising range to ≤ 3 m / s; and the horizontal speed in the last second is ≤ 1.5 m / s (low-speed gliding or hovering when approaching the ground); the vertical speed is negative (downward), and the absolute value is stable at 2-3 m / s (uniform descent, without sudden drops). The altitude continues to decrease within 3 seconds, with a total descent of ≥ 5 meters (or the descent reaches 20% of the initial altitude); the rate of altitude descent is stable at 2-3 m / s, without sudden changes in fluctuation exceeding ±0.8 m / s].

[0060] Step S3: In response to determining that the UAV attitude identifier represents the UAV's return flight, the UAV's return flight route is collected based on the UAV feature information group. Specifically, the implementation method described in step 101 can be used to extract UAV return flight feature information matching the UAV's feature information from the millimeter-wave radar data at the current moment. Here, the UAV return flight feature information may include the UAV's positioning coordinates at consecutive time points. Furthermore, the fitted curve of each UAV positioning coordinate in the UAV return flight feature information can be used to determine the UAV's return flight route.

[0061] Step S4: Based on the aforementioned UAV return route, generate source tracing terminal information, and determine the aforementioned source tracing terminal information, the aforementioned UAV return route, and the aforementioned UAV feature information as the UAV reconnaissance result. The source tracing terminal information includes the source tracing terminal reference coordinates. Here, Kalman filtering can be used to predict the return route of the target UAV (corresponding to the aforementioned UAV feature information) to generate a predicted route and a route confidence score. Secondly, when the route confidence score is greater than a preset confidence threshold, the intersection point of the predicted route and the ground plane in the camera coordinate system can be determined as the source tracing terminal reference coordinates. Finally, the predicted route can also be added to the UAV reconnaissance result.

[0062] Optionally, in response to determining the aforementioned UAV attitude markers representing UAV landing, a UAV backtracking route is generated based on the aforementioned UAV feature information group and the aforementioned object radio frequency information group. Specifically, firstly, UAV feature information and object radio frequency signals at the current time point are extracted from the aforementioned UAV feature information group and the aforementioned object radio frequency information group, respectively serving as the current feature information and the current radio frequency signal. Then, combining the UAV feature information and object radio frequency signals from historical time periods, the corresponding UAV position coordinates and UAV distance values ​​at each time point are extracted. Next, the correspondence between each UAV position coordinate and the UAV distance value is determined through the correspondence between time points. Then, the UAV position coordinates are corrected using each UAV distance value to obtain a corrected UAV coordinate group. Here, the UAV position coordinates can be corrected so that the distance values ​​of the adjusted UAV position coordinates are less than a preset range of the corresponding UAV distance values. Finally, the fitted curves of each adjusted UAV position coordinate are determined as the UAV backtracking route.

[0063] Furthermore, based on the aforementioned UAV backtracking route, source tracing terminal information can be generated, and the aforementioned source tracing terminal information, the aforementioned UAV backtracking route, and the aforementioned UAV feature information can be determined as the UAV reconnaissance result. Specifically, Kalman filtering can be used to predict the UAV backtracking route to obtain the predicted route. Secondly, the intersection point of the predicted route and the ground plane can be determined as the reference coordinates of the source tracing terminal. Therefore, the source tracing terminal information, UAV feature information, UAV backtracking route, and target trajectory can be determined as the UAV reconnaissance result.

[0064] Optionally, the aforementioned executing entity can also invoke the monitoring equipment corresponding to the coordinates of the aforementioned tracing terminal to obtain monitoring video for the target time period. Here, the target time period can be generated by reverse calculation based on the flight speed of the target drone and the predicted route. Secondly, image recognition can be performed on consecutive frames of video images in the monitoring video using a preset network model to generate image recognition results. Here, the image recognition results can include a representation of the drone control terminal identifier and the recognition confidence level. Specifically, during the recognition process, when the recognition confidence level exceeds a preset recognition confidence level threshold, the location of the identified target holding and operating the drone control terminal can be determined as a supplementary query location. Simultaneously, with authorization, and in conjunction with an identity verification system, the target holding and operating the drone control terminal can be scanned for personal information to generate target identity information. Finally, the target identity information and the supplementary query location can be added to the drone reconnaissance results.

[0065] In practice, step 106 and its related content can be used to trace the source of drones that have intruded into electronic fence airspace. Here, the core of the above implementation method lies in shifting the source tracing challenge from the "signal layer" to the "behavioral layer." First, by identifying drone characteristic information and interfering with the signals of illegally flying drones, not only can "black flight" by drones be quickly stopped, but the corresponding drone control terminal can also be detected in a timely manner based on the drone's feedback. Specifically, after a drone is subjected to signal interference, there are various possible reactions, and the above implementation method provides corresponding processing for different reactions: For drones that return on their own, their return route can be identified to trace the drone control terminal. For drones that land, their historical flight trajectory can be used to trace the corresponding terminal location, and combined with corresponding monitoring video, this can provide referenceable source tracing coordinates for the early warning terminal.

[0066] Optionally, the aforementioned implementing entity may also include the following steps: Step S1: Generate an anti-drone mapping map based on the above drone reconnaissance result group. Specifically, the target drone return route (or drone backtracking route) and source terminal coordinates included in each drone reconnaissance result group can be transformed from the camera coordinate system to the map coordinate system to obtain the anti-drone mapping map.

[0067] Step S2: Based on the aforementioned anti-drone mapping map and the aforementioned drone reconnaissance result group, send an anti-drone warning signal to the target device. Specifically, the aforementioned anti-drone mapping map and the aforementioned drone reconnaissance result group can be used as warning information to send the anti-drone warning signal to the target terminal (e.g., a reconnaissance and monitoring terminal in an electronic fence airspace).

[0068] The various embodiments of this disclosure have the following beneficial effects: the anti-drone reconnaissance methods of some embodiments of this disclosure can improve the accuracy of target reconnaissance. Here, the reason for misjudgment of detected targets is that drone target detection using methods such as lidar and visible light cameras is easily affected by environmental interference. Based on this, the anti-drone reconnaissance methods of some embodiments of this disclosure first extract object features from millimeter-wave radar data collected against a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes the object's position coordinates. Here, compared to lidar detection, by introducing millimeter-wave radar, utilizing its stronger penetration capability in adverse weather conditions, the stability of anti-drone reconnaissance under different weather conditions is improved, thereby reducing the impact of weather conditions on the detection results. Then, radio frequency signal reconnaissance is performed based on the above object feature information group to generate an object radio frequency information group, wherein the object radio frequency information includes the type of object signal. Subsequently, the radio frequency information of objects meeting the preset signal filtering conditions in the aforementioned object radio frequency information group is identified as candidate radio frequency information, resulting in a candidate radio frequency information group. The preset signal filtering conditions include object radio frequency information containing an object signal identifier indicating that a preset communication type has not been emitted. The preset communication types include image transmission and control types. Here, considering that millimeter-wave radar has stronger penetration capabilities but relatively lower resolution, radio frequency signal reconnaissance is performed to further filter the detected targets. Then, infrared detection is performed based on the aforementioned candidate radio frequency information group to obtain an infrared detection result group, where the infrared detection results represent the infrared characteristics of the detected target. Using the aforementioned infrared detection result group, interference is eliminated from the object feature information group representing non-UAV targets to generate a UAV feature information group. Here, considering the possibility of false detections (e.g., misidentifying birds as UAVs), and the significant differences between the infrared characteristics of UAVs and birds, infrared detection is introduced on top of the aforementioned millimeter-wave radar + radio frequency signal reconnaissance to further distinguish suspicious detected targets (i.e., candidate radio frequency information), thereby further eliminating interference items. In practice, considering the small feature size of targets in images captured by visible light cameras, which can easily lead to false detections, this application extracts the infrared features of the targets using infrared detection. Furthermore, based on the differences in infrared features between targets (e.g., birds and drones), more accurate target differentiation can be achieved. This further improves the stability of reconnaissance under various weather conditions. Finally, anti-drone jamming is performed based on the aforementioned drone feature information group, and anti-drone tracing is performed using the aforementioned drone feature information group and the aforementioned target radio frequency information group to generate a drone reconnaissance result group. This improves the accuracy of drone target detection, facilitating precise anti-drone jamming.

[0069] Further reference Figure 3 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of an anti-drone reconnaissance device, which are similar to... Figure 1 Corresponding to the method embodiments shown, this anti-drone reconnaissance device can be specifically applied to various electronic devices.

[0070] like Figure 3 As shown, some embodiments of the anti-drone reconnaissance device 300 include: a drone feature extraction unit 301, a radio frequency signal detection unit 302, a determination unit 303, an infrared detection unit 304, an interference elimination unit 305, and an anti-drone jamming unit 306. The drone feature extraction unit 301 is configured to extract object features from millimeter-wave radar data collected against a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes object location coordinates; the radio frequency signal detection unit 302 is configured to perform radio frequency signal detection based on the object feature information group to generate an object radio frequency information group, wherein the object radio frequency information includes the type of object signal; the determination unit 303 is configured to determine the object radio frequency information in the object radio frequency information group that meets preset signal filtering conditions as candidate radio frequency information, obtaining a candidate radio frequency information group, wherein the preset signal filtering conditions are that the object radio frequency information includes object signals that represent not transmitting a preset communication type. The identification system includes preset communication types such as positioning, image transmission, and control. An infrared detection unit 304 is configured to perform infrared detection based on the aforementioned candidate radio frequency information group to obtain an infrared detection result group, wherein the infrared detection result characterizes the infrared features of the detected target. An interference removal unit 305 is configured to use the aforementioned infrared detection result group to remove interference from the object feature information group representing non-UAV targets in the aforementioned object feature information group, thereby generating a UAV feature information group. An anti-UAV interference unit 306 is configured to perform anti-UAV interference based on the aforementioned UAV feature information group, and to perform anti-UAV tracing using the aforementioned UAV feature information group and the aforementioned object radio frequency information group, thereby generating a UAV reconnaissance result group.

[0071] It is understandable that the units described in the anti-drone reconnaissance device 300 are related to the reference Figure 1 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the anti-drone reconnaissance device 300 and the units contained therein, and will not be repeated here.

[0072] The following is for reference. Figure 4 It illustrates a schematic diagram of the structure of an electronic device (such as a computing device) suitable for implementing some embodiments of the present disclosure. Figure 4The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of this disclosure. Figure 4 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory may include a non-volatile storage medium and internal memory. The non-volatile storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any of the methods described above. The processor provides computational and control capabilities to support the operation of the entire computer device. The internal memory provides an environment for the execution of the computer program in the non-volatile storage medium; when executed by the processor, the computer program causes the processor to perform any of the methods described above. The network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present disclosure and does not constitute a limitation on the computer device to which the present disclosure is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0073] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0074] In one embodiment, the processor is configured to run a computer program stored in a memory to perform the following steps: extracting object features from millimeter-wave radar data collected in a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes object location coordinates; performing radio frequency signal detection based on the object feature information group to generate an object radio frequency information group, wherein the object radio frequency information includes the type of object signal; identifying object radio frequency information in the object radio frequency information group that meets preset signal filtering conditions as candidate radio frequency information to obtain a candidate radio frequency information group, wherein the preset signal filtering conditions are object radio frequency information. The process includes identifying objects that have not emitted a preset communication type, which includes positioning, image transmission, and control types. Infrared detection is performed based on the candidate radio frequency information group to obtain an infrared detection result group, where the infrared detection result represents the infrared characteristics of the detected target. Using the infrared detection result group, interference is removed from the object characteristic information group representing non-UAV targets to generate a UAV characteristic information group. Anti-UAV interference is performed based on the UAV characteristic information group, and anti-UAV tracing is performed using the UAV characteristic information group and the object radio frequency information group to generate a UAV reconnaissance result group.

[0075] This disclosure also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, and the method implemented when the program instructions are executed can be referred to the various embodiments of the methods described above.

[0076] The aforementioned computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. Alternatively, the aforementioned computer-readable storage medium may be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0077] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0078] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for countering drone reconnaissance, characterized in that, include: Object features are extracted from millimeter-wave radar data collected in a pre-defined electronic fence airspace to generate object feature information groups, where the object feature information includes object location coordinates. Radio frequency (RF) signal detection is performed based on the object feature information group to generate an object RF information group, wherein the object RF information includes the type of the object signal. This includes: sending a communication request to the detection object corresponding to each object feature information in the object feature information group to obtain the object identity identifier, thus obtaining an object identity identifier group; identifying object feature information in the object feature information group where the corresponding object identity identifier is empty as detection target information, thus obtaining a detection target information group; for each detection target information in the detection target information group and the object location coordinates included in the detection target information, the following first processing step is performed: controlling the signal detection device to detect the object... Signals are collected from the azimuth corresponding to the location coordinates to obtain a communication signal group; the communication signal group is preprocessed to obtain a processed communication signal group; the processed communication signal group and the detection target information are checked for consistency to obtain a check result; in response to the check result indicating that the check is passed, an object signal identifier indicating that the detection target has emitted a preset communication type is generated, and the object signal identifier and the object location coordinates are determined as object radio frequency information; in response to the check result indicating that the check is failed, an object signal identifier indicating that the detection target has not emitted a preset communication type is generated, and the object signal identifier and the object location coordinates are determined as object radio frequency information. The radio frequency information of the objects that meets the preset signal filtering conditions in the object radio frequency information group is determined as the candidate radio frequency information, and the candidate radio frequency information group is obtained. The preset signal filtering condition is that the object radio frequency information includes an object signal identifier that does not transmit a preset communication type. The preset communication types include: positioning type, image transmission type and control type. Infrared detection is performed based on the candidate radio frequency information group to obtain an infrared detection result group, wherein the infrared detection result characterizes the infrared features of the detected target, including: for each candidate radio frequency information in the candidate radio frequency information group, the following second processing step is performed: controlling a high-resolution infrared device to acquire an image of the area where the object position coordinates included in the candidate radio frequency information are located, to obtain a high-resolution infrared image; extracting morphological features from the high-resolution infrared image to obtain morphological features; identifying heat source features from the high-resolution infrared image to generate heat source features; generating an infrared detection identifier based on the morphological features and the heat source features as the infrared detection result, wherein the morphological features and heat source features are normalized to obtain normalized morphological features and normalized heat source features, and the coordinate values ​​of each key point coordinate and the corresponding heat value are arranged in the same order to form a one-dimensional feature vector, and the one-dimensional feature vector is input to a preset lightweight multilayer sensor to generate an infrared detection result, wherein the infrared detection identifier is used to characterize whether the detected target is a drone; Using the infrared detection result group, the object feature information representing non-UAV targets in the object feature information group is removed to eliminate interference, so as to generate the UAV feature information group; Anti-drone jamming is performed based on the drone feature information group, and anti-drone tracing is performed using the drone feature information group and the target radio frequency information group to generate a drone reconnaissance result group.

2. The anti-drone reconnaissance method according to claim 1, characterized in that, The method further includes: Based on the drone reconnaissance results group, an anti-drone mapping map is generated; Based on the anti-drone mapping map and the drone reconnaissance result group, an anti-drone early warning signal is sent to the target device.

3. The anti-drone reconnaissance method according to claim 1, characterized in that, The step of extracting object features from millimeter-wave radar data collected in a preset electronic fence airspace to generate object feature information groups includes: The millimeter-wave radar data is preprocessed to obtain processed radar data; The processed radar data is subjected to target tracking and detection to generate a detection identifier group and a corresponding object position coordinate group; Based on the processed radar data and the object position coordinate group, feature extraction is performed on the detection targets corresponding to each detection identifier in the detection identifier group to obtain the target feature information group. Object recognition is performed on the feature information in the target feature information group to generate an object feature information group.

4. The anti-drone reconnaissance method according to claim 3, characterized in that, The step of using the infrared detection result group to remove interference from the object feature information representing non-UAV targets in the object feature information group to generate the UAV feature information group includes: The infrared detection identifiers representing non-UAVs included in the infrared detection result group are identified as interference identifiers, thus obtaining the interference identifier group; The object feature information that meets the interference removal conditions in the object feature information group is removed to obtain the UAV feature information group, wherein the UAV feature information represents the detection target as a UAV. The interference removal conditions include at least one of the following: the object identity identifier corresponding to the object feature information is not empty, and the object feature information matches the interference identifier in the interference identifier group.

5. The anti-drone reconnaissance method according to claim 4, characterized in that, The step of performing anti-drone jamming based on the drone feature information group and performing anti-drone tracing using the drone feature information group and the target radio frequency information group to generate a drone reconnaissance result group includes: For each UAV feature in the UAV feature information group, perform the following interference steps: Control a radio jamming device that matches the drone's feature information to jam the signal of the drone corresponding to the drone's feature information, and control a high-resolution infrared device to acquire images of the drone corresponding to the drone's feature information to obtain images of the jammed drone. The drone attitude is identified from the interfered drone image to generate a drone attitude identifier; In response to determining that the UAV attitude identifier represents the UAV's return to home, the UAV's return route is collected based on the UAV feature information group; Based on the drone's return route, traceability terminal information is generated, and the traceability terminal information, the drone's return route, and the drone's feature information are determined as the drone reconnaissance results, wherein the traceability terminal information includes the traceability terminal's reference coordinates.

6. A counter-drone reconnaissance device, characterized in that, include: The UAV feature extraction unit is configured to extract object features from millimeter-wave radar data collected in a preset electronic fence airspace to generate an object feature information group, wherein the object feature information includes object location coordinates. The radio frequency (RF) signal detection unit is configured to perform RF signal detection based on the object feature information group to generate an object RF information group, wherein the object RF information includes the type of the object signal, including: sending a communication request to the detection object corresponding to each object feature information in the object feature information group to obtain the object identity identifier, thereby obtaining an object identity identifier group; determining the object feature information in the object feature information group whose corresponding object identity identifier is empty as detection target information, thereby obtaining a detection target information group; and for each detection target information in the detection target information group and the object position coordinates included in the detection target information, performing the following first processing step: control signal detection. The device collects signals from the azimuth corresponding to the object's position coordinates to obtain a communication signal group; it preprocesses the communication signal group to obtain a processed communication signal group; it performs a consistency check between the processed communication signal group and the detected target information to obtain a check result; in response to the check result indicating that the check passed, it generates an object signal identifier indicating that the detected target emitted a preset communication type, and determines the object signal identifier and the object's position coordinates as object radio frequency information; in response to the check result indicating that the check failed, it generates an object signal identifier indicating that the detected target did not emit a preset communication type, and determines the object signal identifier and the object's position coordinates as object radio frequency information. The determining unit is configured to determine the object radio frequency information in the object radio frequency information group that meets the preset signal filtering conditions as candidate radio frequency information, thereby obtaining a candidate radio frequency information group. The preset signal filtering conditions are that the object radio frequency information includes an object signal identifier that does not transmit a preset communication type. The preset communication types include: positioning type, image transmission type, and control type. An infrared detection unit is configured to perform infrared detection based on the candidate radio frequency information group to obtain an infrared detection result group, wherein the infrared detection result characterizes the infrared features of the detected target, including: for each candidate radio frequency information in the candidate radio frequency information group, performing the following second processing step: controlling a high-resolution infrared device to acquire an image of the area where the object position coordinates included in the candidate radio frequency information are located, to obtain a high-resolution infrared image; extracting morphological features from the high-resolution infrared image to obtain morphological features; identifying heat source features from the high-resolution infrared image to generate heat source features; generating an infrared detection identifier based on the morphological features and the heat source features as the infrared detection result, wherein the morphological features and heat source features are normalized to obtain normalized morphological features and normalized heat source features, and the coordinate values ​​of each key point coordinate and the corresponding heat value are arranged in the same order to form a one-dimensional feature vector, and the one-dimensional feature vector is input to a preset lightweight multilayer sensor to generate an infrared detection result, wherein the infrared detection identifier is used to characterize whether the detected target is a drone. The interference removal unit is configured to use the infrared detection result group to remove interference from the object feature information representing non-UAV targets in the object feature information group to generate a UAV feature information group. The anti-drone jamming unit is configured to perform anti-drone jamming based on the drone feature information group, and to perform anti-drone tracing using the drone feature information group and the target radio frequency information group to generate a drone reconnaissance result group.

7. An electronic device, characterized in that, include: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-5.

8. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the program, when executed by a processor, implements the method as described in any one of claims 1-5.

Images