Unmanned aerial vehicle high-precision locking strike method under complex background

Abstract

The invention discloses an unmanned aerial vehicle high-precision locking striking method under a complex background, and belongs to the technical field of unmanned aerial vehicle striking. The method specifically comprises the following steps: S1, obtaining an unmanned aerial vehicle target, S2, inputting a target slice obtained in S1 into a weak point extraction network, if a weak point coordinate is successfully extracted, selecting a coordinate with the maximum confidence coefficient as a strike point, and if the weak point coordinate is not successfully extracted, determining that extraction fails, and re-operating a target detection algorithm to obtain a new target slice; the high-precision locking and striking method for the unmanned aerial vehicle under the complex background has the advantages that the unmanned aerial vehicle target can be intelligently locked at high precision under the complex background, and the target unmanned aerial vehicle striking efficiency can be improved, so that the high-precision locking and striking method for the unmanned aerial vehicle under the complex background can be improved, and the high-precision locking and striking method for the unmanned aerial vehicle under the complex background can be applied to the unmanned aerial vehicle. The comprehensive treatment effect of the laser damage system can be improved.

Description

Technical Field

[0001] The present invention specifically relates to a high-precision locking and striking method for a UAV under a complex background, and belongs to the technical field of UAV striking. Background Art

[0002] In recent years, drone technology has developed rapidly, and consumer drones have been widely used, such as low-altitude photogrammetry, power line inspection, public security inspection, and photography by individual enthusiasts. While drones are widely used, they also bring a lot of hidden dangers to public safety, such as secretly photographing confidential places, invading airports, and delivering prohibited items, which are common, posing a serious threat to national, military and public security. For this reason, drones need to be destroyed during illegal flight. As a hard-kill method for drones, high-energy laser destruction systems use target detection and tracking technology and high-energy laser beam far-field control technology to achieve continuous high-energy burning of drone targets, thereby achieving physical destruction of drone targets. This method is silent, accurate and efficient, and has no electromagnetic pollution to the environment, but this method has high requirements for target tracking accuracy. The existing destruction system is very easy to miss the target under complex backgrounds (such as buildings, trees, mountains, thick clouds, etc.) and when there is a large fire cover, which makes it more difficult to destroy or even fails to strike. Summary of the invention

[0003] In order to solve the above problems, the present invention proposes a high-precision locking and striking method for UAVs under complex backgrounds, which can intelligently lock UAV targets with high precision under complex backgrounds and improve the comprehensive disposal effect of the laser damage system.

[0004] The high-precision locking and striking method of a drone under a complex background of the present invention is as follows: S1: UAV target acquisition. After receiving the external locking and tracking command, the visual device captures the tracking scene image in real time. In order to make the target imaging details clear and convenient for algorithm processing, the target imaging scale is required to be Not less than 100 pixels * 100 pixels, and the scene image is sent to the intelligent processing unit through the image acquisition card; the intelligent processing unit detects the drone target through the target detection algorithm. If the drone target is successfully detected, the detection result with the highest confidence is selected as the tracking target. If the drone target is not successfully detected, it is determined that the detection has failed, and the target detection algorithm continues to detect the scene image sent to the image acquisition card; the drone target detected by the target detection algorithm is the target slice cropped by the detection frame; S2: Input the target slice obtained in S1 into the weak point extraction network. If the weak point coordinates are successfully extracted, the coordinates with the highest confidence are selected as the strike point. If the weak point coordinates are not successfully extracted, the extraction is judged to have failed, and the target detection algorithm is re-run to obtain a new target slice. S3: Initialization of target tracking network: Use the weak point coordinates of the target detection frame and its corresponding target slice to initialize the target tracking network. The target tracking network is initialized. S4: Dynamically adjust the focal length of the visual device to maintain the size consistency of the collected on-site images; measure the distance in real time through the rangefinder equipped with the visual device, and adjust the camera focal length of the visual device in real time according to the distance measurement results; S5: The drone continues to track the target, and the target tracking network continuously updates the coordinates of the target bounding box and the coordinates of the weak points on the fuselage; S6: Strike the weak points of the drone target. After receiving the external laser strike command, the servo turntable and the quick-reflection mirror will deflect in response to the drone target continuously tracked by S5. At the same time, the laser beam will continue to burn the weak points of the drone through the quick-reflection mirror. S7, the target drone strike is completed. When the intelligent processing unit detects that the characteristic map response value corresponding to the target center point position is less than 0.01, it is confirmed that the target is lost, the tracking and strike are stopped, and the target tracking network is initialized again; when the stop light and cancel tracking instructions are received, the intelligent processing unit stops the target tracking network initialization operation, and then controls the servo turntable and the fast reflection mirror to stop deflection, and completes the target drone strike; the high-precision locking and striking method of drones under complex backgrounds of the present invention realizes rapid and intelligent extraction of drone target weaknesses by constructing a deep convolution model algorithm; at the same time, a twin network model algorithm is constructed to achieve continuous locking and tracking of drone targets, and ensure that the laser strike point is always focused on the weak points of the drone target, thereby achieving rapid disposal of the drone target.

[0005] The high-precision locking and striking method for unmanned aerial vehicles under complex backgrounds of the present invention needs to track and process the images through a GPU intelligent processing device when processing on-site images, and in order to ensure the processing efficiency, the tracking camera video stream data on the tracking turntable can be acquired with low latency through an image acquisition card, and the latency is controlled within 20ms; when the intelligent processing unit is operating, the image processing can be completed within 15ms, and the target's miss amount information can be converted into angle control information, and the miss amount information can be tracked and pushed to the servo control system and the fast reflex mirror system in real time; the servo control system drives the servo turntable to respond quickly; the servo turntable realizes tracking of the unmanned aerial vehicle target, ensuring that the unmanned aerial vehicle target is located in the exact center of the tracking video image; the fast reflex mirror system drives the fast reflex mirror to respond quickly; and the high-energy laser beam pointing is controlled in real time to ensure that the laser strike point is stably focused on the weak point of the unmanned aerial vehicle target body.

[0006] Furthermore, the real-time adjustment of the focal length of the camera equipped with the visual device according to the ranging result is as follows: the focal length value of the camera at the time of initialization of the known target tracking network and the distance value , then the focal length value and the distance value The conversion relationship is as follows: Formula 1; Since the camera focal length value at the initialization time of the target tracking network and the distance value is a known condition, and the distance value It can be obtained in real time through the rangefinder. According to formula 1, the focal length value can be calculated ; and according to the focal length value Adjust the camera focus of the vision device in real time.

[0007] Furthermore, the target detection algorithm is a YOLO series algorithm; if yolov5 is used for detection, in the detection result, if the confidence level is greater than 0.5, it is assumed that the target is successfully detected.

[0008] Furthermore, the working process of the weak point extraction network is as follows: The target slice obtained by S1 is scaled to a size of 64*64*3, and then input into the backbone network for feature extraction. In order to improve the video processing speed and achieve an effective frame rate, the backbone network selects lightweight convolutional neural networks such as MobileNetV3 or GhostNet, and the output of the backbone convolutional neural network is an 8*8*3 feature map; the backbone network outputs an 8*8*3 feature map, and the feature map is used to extract the weak points of the fuselage; according to actual combat experience, the rotor of the drone can be easily penetrated by high-energy lasers quickly, and hitting the rotor can cause the drone to fall quickly, so the rotor of the drone can be selected as the weak point of the fuselage. Since the extraction of weak points is affected by factors such as camera imaging effects, background interference, and foreground occlusion, the weak point extraction network can obtain and output a set of recommended rankings of the weak point positions of the fuselage. When the weak point extraction network is trained, the loss function is used for training evaluation and model optimization. The loss function is defined as follows: Formula 2; As shown in Formula 2, during the training phase, the data annotation points are regressed by calculating the loss function. Specifically, during the training phase, the input image is divided into 8*8 grids. When a data annotation point falls into a certain grid, this grid is responsible for predicting this annotation point. The output feature map generates 8*8 groups of prediction values. Each group of prediction values ​​contains x, y coordinates and confidence conf, where Represents grid coordinates, when the grid contains annotation points , otherwise 0, On the contrary, Represents the weight parameter when calculating coordinate loss, Represents the weight parameter of the grid when calculating the confidence loss if it does not contain the annotation point. represents the coordinate prediction value, represents the true value of the coordinate, represents the confidence prediction value, Represents the true confidence value when the grid contains the annotation point , otherwise it is 0; in the inference stage, the weak points of the input image are predicted. In the inference stage, the backbone network can predict the weak points of the input image and obtain 64 sets of prediction values. The prediction confidence is evaluated by calculating the loss function. The prediction results are considered credible, and after mapping the prediction results back to the initial image slice coordinate values, they are sorted from large to small according to the confidence level; finally, a set of recommended sorting of fuselage weak point locations is output.

[0009] Furthermore, the target tracking network can realize continuous tracking of the drone target and continuous pointing to the key point area on the drone. The target tracking network is specifically as follows: S51: Establish a twin network with the same structure as the weak point extraction network, and set the frame number of the acquired image to , and the image tracking box size ; Initialize the twin network, the initial frame is known , target bounding box and target key points, the initial frame is preprocessed based on the target bounding box and target key point coordinates, the initial frame preprocessing includes cropping and scaling, and a 32*32*3 key point frame and a 127*127*3 template frame are obtained; in order to avoid the violent jitter of the tracking target center point due to the continuous change of the target tracking frame size during target tracking, the focal length of the visual device is dynamically adjusted to maintain the consistency of the size of the collected on-site image; that is, the prerequisite requirement for the use of the target tracking network, the scale of the tracked target image will not change significantly, and a fixed-size tracking frame size is used To limit; S52: Input the key point frame and template frame obtained in S51 into the backbone network of the twin network. The backbone network of the twin network uses a 5-layer convolutional network to output feature maps of 2*2*256 and 14*14*256 respectively. and ; S53: Start target tracking, knowing the previous frame The target bounding box coordinates are based on this coordinate position for the current frame Perform preprocessing, including cropping and scaling, to obtain a search frame of 303*303*3; S54: Input the search frame obtained in S53 into the backbone network of the twin network and output a 36*36*256 feature map ; When the weak point of the target is continuously affected by the laser focus, it will be covered by fire. At this time, if the same backbone network parameters are used for feature extraction of the key point frame, template frame and search frame, it is difficult for the backbone network to learn the similarity between the two features of whether the target is covered by fire. Therefore, during model training, the key point frame, template frame and search frame need to learn different backbone network parameters respectively; therefore, when extracting the features of the key point frame, template frame and search frame, they are completed by the backbone networks of their respective twin networks respectively; S55: and Input to the KeyPoint branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 27*27*1 response map representing the similarity between different positions and the initial key point frame; S56: and Input to the Classification branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 17*17*1 response map, representing whether the target information is contained at different positions; S57: and Input to the Centerness branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 17*17*1 response map, representing the probability of the target center at different positions in the image; S58: First multiply the elements of the Classification branch response map and the Centerness branch response map, then take out the location of the maximum response value, map it back to the original map coordinates, and get the final target center point coordinates. ); S59: Map the KeyPoint branch response map back to the original image coordinates, given the previous frame The target key point coordinates of the previous frame Coordinate offset between center point coordinates ,by ) is the search center, , obtain the coordinates of the location where the maximum response value in the search area is located, and obtain the final key point prediction result, which includes the bounding box coordinates and the coordinate values ​​of the weak point of the fuselage.

[0010] Furthermore, the deflection response of the servo turntable and the fast reflection mirror is specifically as follows: Calculate the servo miss distance based on the camera parameters of the visual device, the target center point coordinates, and the target weak point coordinates and the fast-reflection mirror miss , servo miss distance and the fast-reflection mirror miss This is the correction amount that the servo turntable and the quick-reflection mirror need to make; the servo turntable and the quick-reflection mirror deflect in response, so that the servo turntable Continuously track the target center, the fast-reflection mirror Continue to accurately point to the drone's weak points.

[0011] Compared with the prior art, the high-precision locking and striking method for drones under complex backgrounds of the present invention achieves rapid disposal of drone targets by quickly and intelligently extracting the weak points of drone targets, continuously locking and tracking drone targets, and ensuring that the laser striking point is always focused on the weak points of drone targets; it can achieve intelligent locking and striking of drone targets under complex backgrounds such as buildings, trees, mountains, and large fire coverage during strikes. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 This is a block diagram of the overall workflow module of the UAV high-precision locking and striking method of the present invention.

[0013] Figure 2 The figure is a schematic diagram of the overall working process of the high-precision locking and striking method for unmanned aerial vehicles of the present invention.

[0014] Figure 3 It is a schematic diagram of the network workflow of vulnerability extraction of the present invention.

[0015] Figure 4 This is a schematic diagram of the connection structure of each module of the target tracking network of the present invention.

[0016] Figure 5 It is a schematic diagram of the target tracking network workflow of the present invention. DETAILED DESCRIPTION

[0017] Embodiment 1: like Figures 1 to 5 The high-precision locking and striking method of a UAV under a complex background is shown, and the method is specifically as follows: S1: UAV target acquisition. After receiving the external locking and tracking command, the visual device captures the tracking scene image in real time. In order to make the target imaging details clear and convenient for algorithm processing, the target imaging scale is required to be Not less than 100 pixels * 100 pixels, and the scene image is sent to the intelligent processing unit through the image acquisition card; the intelligent processing unit detects the drone target through the target detection algorithm. If the drone target is successfully detected, the detection result with the highest confidence is selected as the tracking target. If the drone target is not successfully detected, it is determined that the detection has failed, and the target detection algorithm continues to detect the scene image sent to the image acquisition card; the drone target detected by the target detection algorithm is the target slice cropped by the detection frame; S2: Input the target slice obtained in S1 into the weak point extraction network. If the weak point coordinates are successfully extracted, the coordinates with the highest confidence are selected as the strike point. If the weak point coordinates are not successfully extracted, the extraction is judged to have failed, and the target detection algorithm is re-run to obtain a new target slice. S3: Initialization of target tracking network: Use the weak point coordinates of the target detection frame and its corresponding target slice to initialize the target tracking network. The target tracking network is initialized. S4: Dynamically adjust the focal length of the visual device to maintain the size consistency of the collected on-site images; measure the distance in real time through the rangefinder equipped with the visual device, and adjust the camera focal length of the visual device in real time according to the distance measurement results; S5: The drone continues to track the target, and the target tracking network continuously updates the coordinates of the target bounding box and the coordinates of the weak points on the fuselage; S6: Strike the weak points of the drone target. After receiving the external laser strike command, the servo turntable and the quick-reflection mirror will deflect in response to the drone target continuously tracked by S5. At the same time, the laser beam will continue to burn the weak points of the drone through the quick-reflection mirror. S7, the target drone strike is completed. When the intelligent processing unit detects that the characteristic map response value corresponding to the target center point position is less than 0.01, it is confirmed that the target is lost, the tracking and strike are stopped, and the target tracking network is initialized again; when the stop light and cancel tracking instructions are received, the intelligent processing unit stops the target tracking network initialization operation, and then controls the servo turntable and the fast reflection mirror to stop deflection, and completes the target drone strike; the high-precision locking and striking method of drones under complex backgrounds of the present invention realizes rapid and intelligent extraction of drone target weaknesses by constructing a deep convolution model algorithm; at the same time, a twin network model algorithm is constructed to achieve continuous locking and tracking of drone targets, and ensure that the laser strike point is always focused on the weak points of the drone target, thereby achieving rapid disposal of the drone target.

[0018] The high-precision locking and striking method for unmanned aerial vehicles under complex backgrounds of the present invention needs to track and process the images through a GPU intelligent processing device when processing on-site images, and in order to ensure the processing efficiency, the tracking camera video stream data on the tracking turntable can be acquired with low latency through an image acquisition card, and the latency is controlled within 20ms; when the intelligent processing unit is operating, the image processing can be completed within 15ms, and the target's miss amount information can be converted into angle control information, and the miss amount information can be tracked and pushed to the servo control system and the fast reflex mirror system in real time; the servo control system drives the servo turntable to respond quickly; the servo turntable realizes tracking of the unmanned aerial vehicle target, ensuring that the unmanned aerial vehicle target is located in the exact center of the tracking video image; the fast reflex mirror system drives the fast reflex mirror to respond quickly; and the high-energy laser beam pointing is controlled in real time to ensure that the laser strike point is stably focused on the weak point of the unmanned aerial vehicle target body.

[0019] The real-time adjustment of the focal length of the camera equipped with the visual device according to the ranging result is as follows: The focal length value of the camera at the time of initialization of the known target tracking network and the distance value , then the focal length value and the distance value The conversion relationship is as follows: Formula 1; Since the camera focal length value at the initialization time of the target tracking network and the distance value is a known condition, and the distance value It can be obtained in real time through the rangefinder. According to formula 1, the focal length value can be calculated ; and according to the focal length value Adjust the camera focus of the vision device in real time.

[0020] The target detection algorithm is a YOLO series algorithm; if yolov5 is used for detection, in the detection result, if the confidence level is greater than 0.5, it is assumed that the target is successfully detected.

[0021] The specific working process of the weak point extraction network is as follows: The target slice obtained by S1 is scaled to a size of 64*64*3, and then input into the backbone network for feature extraction. In order to improve the video processing speed and achieve an effective frame rate, the backbone network selects lightweight convolutional neural networks such as MobileNetV3 or GhostNet. The output of the backbone convolutional neural network is an 8*8*3 feature map; the backbone network outputs an 8*8*3 feature map, which is used to extract the weak points of the fuselage. According to actual combat experience, the rotor of a drone can be easily penetrated by a high-energy laser quickly, and hitting the rotor can cause the drone to fall quickly. The rotor of the drone is selected as the weak point of the fuselage, such as Figure 2The white dots shown are the weak points of the fuselage. Since the weak point extraction is affected by factors such as camera imaging effects, background interference, and foreground occlusion, the weak point extraction network can obtain and output a set of recommended rankings of the weak point positions of the fuselage. When the weak point extraction network is trained, the loss function is used for training evaluation and model optimization. The loss function is defined as follows: Formula 2; As shown in Formula 2, during the training phase, the data annotation points are regressed by calculating the loss function. Specifically, during the training phase, the input image is divided into 8*8 grids. When a data annotation point falls into a certain grid, this grid is responsible for predicting this annotation point. The output feature map generates 8*8 groups of prediction values. Each group of prediction values ​​contains x, y coordinates and confidence conf, where Represents grid coordinates, when the grid contains annotation points , otherwise 0, On the contrary, Represents the weight parameter when calculating coordinate loss, Represents the weight parameter of the grid when calculating the confidence loss if it does not contain the annotation point. represents the coordinate prediction value, represents the true value of the coordinate, represents the confidence prediction value, Represents the true confidence value when the grid contains the annotation point , otherwise it is 0; in the inference stage, the weak points of the input image are predicted. In the inference stage, the backbone network can predict the weak points of the input image and obtain 64 sets of prediction values. The prediction confidence is evaluated by calculating the loss function. The prediction results are considered credible, and after mapping the prediction results back to the initial image slice coordinate values, they are sorted from large to small according to the confidence level; finally, a set of recommended sorting of fuselage weak point locations is output.

[0022] The target tracking network can realize continuous tracking of the drone target and continuous pointing to the key point area on the drone. The target tracking network is specifically as follows: S51: Establish a twin network with the same structure as the weak point extraction network, and set the frame number of the acquired image to , and the image tracking box size ; Initialize the twin network, the initial frame is known , target bounding box and target key points, the initial frame is preprocessed based on the target bounding box and target key point coordinates, the initial frame preprocessing includes cropping and scaling, and a 32*32*3 key point frame and a 127*127*3 template frame are obtained; in order to avoid the violent jitter of the tracking target center point due to the continuous change of the target tracking frame size during target tracking, the focal length of the visual device is dynamically adjusted to maintain the consistency of the size of the collected on-site image; that is, the prerequisite requirement for the use of the target tracking network, the scale of the tracked target image will not change significantly, and a fixed-size tracking frame size is used To limit; S52: Input the key point frame and template frame obtained in S51 into the backbone network of the twin network. The backbone network of the twin network uses a 5-layer convolutional network to output feature maps of 2*2*256 and 14*14*256 respectively. and ; S53: Start target tracking, knowing the previous frame The target bounding box coordinates are based on this coordinate position for the current frame Perform preprocessing, including cropping and scaling, to obtain a search frame of 303*303*3; S54: Input the search frame obtained in S53 into the backbone network of the twin network and output a 36*36*256 feature map ; When the weak point of the target is continuously affected by the laser focus, it will be covered by fire. At this time, if the same backbone network parameters are used for feature extraction of the key point frame, template frame and search frame, it is difficult for the backbone network to learn the similarity between the two features of whether the target is covered by fire. Therefore, during model training, the key point frame, template frame and search frame need to learn different backbone network parameters respectively; therefore, when extracting the features of the key point frame, template frame and search frame, they are completed by the backbone networks of their respective twin networks respectively; S55: and Input to the KeyPoint branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 27*27*1 response map representing the similarity between different positions and the initial key point frame; S56: and Input to the Classification branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 17*17*1 response map, representing whether the target information is contained at different positions; S57: and Input to the Centerness branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 17*17*1 response map, representing the probability of the target center at different positions in the image; S58: First multiply the elements of the Classification branch response map and the Centerness branch response map, then take out the location of the maximum response value, map it back to the original map coordinates, and get the final target center point coordinates. ); S59: Map the KeyPoint branch response map back to the original image coordinates, given the previous frame The target key point coordinates of the previous frame Coordinate offset between center point coordinates ,by ) is the search center, , obtain the coordinates of the location where the maximum response value in the search area is located, and obtain the final key point prediction result, which includes the bounding box coordinates and the coordinate values ​​of the weak point of the fuselage.

[0023] The deflection response of the servo turntable and the fast-reflection mirror is as follows: Calculate the servo miss distance based on the camera parameters of the visual device, the target center point coordinates, and the target weak point coordinates and the fast-reflection mirror miss , servo miss distance and the fast-reflection mirror miss This is the correction amount that the servo turntable and the quick-reflection mirror need to make; the servo turntable and the quick-reflection mirror deflect in response, so that the servo turntable Continuously track the target center, the fast-reflection mirror Continue to accurately point to the drone's weak points.

[0024] The above embodiments are only preferred implementations of the present invention, so any equivalent changes or modifications made according to the structures, features and principles described in the scope of application of the present invention are included in the scope of application of the present invention.

Claims

1. A high-precision locking and striking method for UAVs in complex backgrounds, characterized by: The method is specifically as follows: S1: UAV target acquisition. After receiving the external locking and tracking command, the visual device captures the tracking scene image in real time and sends the scene image to the intelligent processing unit through the image acquisition card; The intelligent processing unit detects the drone target through the target detection algorithm. If the drone target is successfully detected, the detection result with the highest confidence is selected as the tracking target. If the drone target is not successfully detected, the detection is judged to have failed, and the target detection algorithm is used to continue to detect the on-site image sent by the image acquisition card. The drone target detected by the target detection algorithm is a target slice cropped by the detection frame; S2: Input the target slice obtained in S1 into the weak point extraction network. If the weak point coordinates are successfully extracted, the coordinates with the highest confidence are selected as the strike point. If the weak point coordinates are not successfully extracted, the extraction is judged to have failed, and the target detection algorithm is re-run to obtain a new target slice. S3: Initialization of target tracking network: Use the weak point coordinates of the target detection frame and its corresponding target slice to initialize the target tracking network. The target tracking network is initialized. S4: Dynamically adjust the focal length of the visual device to maintain the size consistency of the collected on-site images; measure the distance in real time through the rangefinder equipped with the visual device, and adjust the camera focal length of the visual device in real time according to the distance measurement results; S5: The drone continues to track the target, and the target tracking network continuously updates the coordinates of the target bounding box and the coordinates of the weak points on the fuselage; S6: Strike the weak points of the drone target. After receiving the external laser strike command, the servo turntable and the quick-reflection mirror will deflect in response to the drone target continuously tracked by S5. At the same time, the laser beam will continue to burn the weak points of the drone through the quick-reflection mirror. S7. Complete the target drone strike. When the intelligent processing unit detects that the characteristic map response value corresponding to the target center point position is less than 0.01, it confirms that the target is lost, stops tracking and striking, and re-initializes the target tracking network. When receiving the stop light and cancel tracking instructions, the intelligent processing unit stops the target tracking network initialization operation, and then controls the servo turntable and fast reflection mirror to stop deflection, completing the target drone strike.

2. The high-precision locking and striking method for UAVs under complex backgrounds according to claim 1 is characterized in that: The real-time adjustment of the focal length of the camera equipped with the visual device according to the ranging result is as follows: The focal length value of the camera at the time of initialization of the known target tracking network and the distance value , then the focal length value and the distance value The conversion relationship is as follows: Formula 1; Since the camera focal length value at the initialization time of the target tracking network and the distance value is a known condition, and the distance value It can be obtained in real time through the rangefinder. According to formula 1, the focal length value can be calculated ; and according to the focal length value Adjust the camera focus of the vision device in real time.

3. The high-precision locking and striking method for UAVs under complex backgrounds according to claim 1 is characterized in that: The target detection algorithm is a YOLO series algorithm; in the detection result, if the confidence level is greater than 0.5, it is assumed that the target is successfully detected.

4. The high-precision locking and striking method for UAVs under complex backgrounds according to claim 1 is characterized in that: The specific working process of the weak point extraction network is as follows: The target slice obtained by S1 is scaled to a size of 64*64*3, and then input into the backbone network for feature extraction. The backbone network outputs a 8*8*3 feature map, which is used to extract the weak points of the fuselage. When the weak point extraction network is trained, the loss function is used for training evaluation and model optimization. The loss function is defined as follows: Formula 2: As shown in Formula 2, during the training phase, the data annotation points are regressed by calculating the loss function. Specifically, during the training phase, the input image is divided into 8*8 grids. When a data annotation point falls into a certain grid, this grid is responsible for predicting this annotation point. The output feature map generates 8*8 groups of prediction values. Each group of prediction values ​​contains x, y coordinates and confidence conf, where Represents grid coordinates, when the grid contains annotation points , otherwise 0, On the contrary, Represents the weight parameter when calculating coordinate loss, Represents the weight parameter of the grid when calculating the confidence loss if it does not contain the annotation point. represents the coordinate prediction value, represents the true value of the coordinate, represents the confidence prediction value, Represents the true confidence value when the grid contains the annotation point , otherwise it is 0; in the inference stage, the weak points of the input image are predicted. In the inference stage, the backbone network can predict the weak points of the input image and obtain 64 sets of prediction values. The prediction confidence is evaluated by calculating the loss function. The prediction results are considered credible, and after mapping the prediction results back to the initial image slice coordinate values, they are sorted from large to small according to the confidence level; finally, a set of recommended sorting of fuselage weak point locations is output.

5. The high-precision locking and striking method for UAVs under complex backgrounds according to claim 4 is characterized in that: The backbone network is a MobileNetV3 or GhostNet convolutional neural network.

6. The high-precision locking and striking method for UAVs under complex backgrounds according to claim 1 is characterized in that: The target tracking network is as follows: S51: Establish a twin network with the same structure as the weak point extraction network, and set the frame number of the acquired image to , and the image tracking box size ; Initialize the twin network, the initial frame is known , target bounding box and target key points, preprocessing the initial frame based on the target bounding box and target key point coordinates, wherein the initial frame preprocessing includes cropping and scaling to obtain a 32*32*3 key point frame and a 127*127*3 template frame; S52: Input the key point frame and template frame obtained in S51 into the backbone network of the twin network. The backbone network of the twin network uses a 5-layer convolutional network to output feature maps of 2*2*256 and 14*14*256 respectively. and ; S53: Start target tracking, knowing the previous frame The target bounding box coordinates are based on this coordinate position for the current frame Perform preprocessing, including cropping and scaling, to obtain a search frame of 303*303*3; S54: Input the search frame obtained in S53 into the backbone network of the twin network and output a 36*36*256 feature map ; The key point frame, template frame and search frame feature extraction are completed through the backbone network of their respective twin networks; S55: and Input to the KeyPoint branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 27*27*1 response map representing the similarity between different positions and the initial key point frame; S56: and Input to the Classification branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 17*17*1 response map, representing whether the target information is contained at different positions; S57: and Input to the Centerness branch, which includes a cross-correlation layer and a feature extraction layer, and outputs a 17*17*1 response map, representing the probability of the target center at different positions in the image; S58: First multiply the elements of the Classification branch response map and the Centerness branch response map, then take out the location of the maximum response value, map it back to the original map coordinates to obtain the final target center point coordinates ( ); S59: Map the KeyPoint branch response map back to the original image coordinates, given the previous frame The target key point coordinates of the previous frame Coordinate offset between center point coordinates ,by ) is the search center, , obtain the coordinates of the location where the maximum response value in the search area is located, and obtain the final key point prediction result, which includes the bounding box coordinates and the coordinate values ​​of the weak point of the fuselage.

7. The high-precision locking and striking method for UAVs under complex backgrounds according to claim 1 is characterized in that: The deflection response of the servo turntable and the fast reflection mirror is as follows: Calculate the servo miss distance based on the camera parameters of the visual device, the target center point coordinates, and the target weak point coordinates and the fast-reflection mirror miss The servo turntable and the fast mirror deflect in response, so that the servo turntable Continuously track the target center, the fast-reflection mirror Continue to accurately point to the drone's weak points.

Images