Unmanned aerial vehicle detection method and device, electronic equipment and storage medium

Abstract

The invention discloses an unmanned aerial vehicle detection method and device, electronic equipment and a storage medium. The method comprises the following steps: when a control signal of a target unmanned aerial vehicle is identified, performing video acquisition on the target unmanned aerial vehicle according to the control signal of the target unmanned aerial vehicle to obtain a target video; performing unmanned aerial vehicle detection on each video image in the target video to obtain an unmanned aerial vehicle detection result of each video image; and determining the flight path of the target unmanned aerial vehicle according to the playing sequence of each video image in the target video and the unmanned aerial vehicle detection result of each video image. The embodiment of the invention can improve the detection accuracy of the unmanned aerial vehicle.

Description

Technical Field

[0001] The present invention relates to the field of machine vision, and in particular to a drone detection method, device, electronic equipment and storage medium. Background Art

[0002] Drones are increasingly being used in key industries and activities, including delivery services, surveillance, and inspections. While this widespread use of technology has greatly promoted industry progress, it has also brought challenges in terms of safety, public security, and privacy protection.

[0003] Given the tiny size, dynamic speed fluctuations, and changing flight environment of micro-UAV targets, relying solely on radar monitoring and passive positioning technology may be interfered by non-target signals, causing frequent false alarms. These technologies may only provide rough location information of UAV targets, and it is difficult to achieve accurate pixel-level tracking of UAV targets. Summary of the invention

[0004] The present invention provides a drone detection method, device, electronic equipment and storage medium, which can improve the drone detection accuracy.

[0005] According to one aspect of the present invention, a method for detecting a drone is provided, comprising: When a control signal of a target UAV is identified, a video of the target UAV is captured according to the control signal of the target UAV to obtain a target video; Performing drone detection on each video image in the target video to obtain a drone detection result for each video image; The flight trajectory of the target UAV is determined according to the playback order of each of the video images in the target video and the UAV detection result of each of the video images.

[0006] According to another aspect of the present invention, there is provided a drone detection device, characterized in that it comprises: A video acquisition module is used to acquire a video of the target UAV according to the control signal of the target UAV when the control signal of the target UAV is identified, so as to obtain a target video; A drone detection module is used to perform drone detection on each video image in the target video to obtain a drone detection result for each video image; The flight trajectory determination module is used to determine the flight trajectory of the target drone according to the playback order of each video image in the target video and the drone detection result of each video image.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the drone detection method described in any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the drone detection method described in any embodiment of the present invention when executed.

[0009] The technical solution of the embodiment of the present invention determines the collection direction of the target UAV according to the control signal of the target UAV, and collects video of the target UAV based on the collection direction to obtain the target video, and locates the target UAV in the video, so as to determine the flight trajectory of the target UAV according to the positioning information, thereby realizing the positioning of the UAV by combining radio signal and image target detection, thereby improving the positioning accuracy of the UAV.

[0010] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 is a flow chart of a drone detection method provided according to an embodiment of the present invention; Figure 2 is a schematic diagram of a scenario for identifying and tracking a drone at low altitude provided by an embodiment of the present invention; Figure 3 is a schematic diagram of a scenario for identifying and tracking a drone according to an embodiment of the present invention; Figure 4 is a schematic diagram of a scenario for identifying and tracking a drone according to an embodiment of the present invention; Figure 5 is a schematic diagram of a scenario for identifying and tracking a drone according to an embodiment of the present invention; Figure 6is a schematic diagram of a target detection result provided according to an embodiment of the present invention; Figure 7 is a structural schematic diagram of a drone detection device provided according to an embodiment of the present invention; Figure 8 Schematic diagram of the structure of an electronic device for implementing the drone detection method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0013] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0015] Figure 1 A flowchart of a drone detection method provided by an embodiment of the present invention. The embodiment of the present invention is applicable to the situation of signal source positioning in a complex environment. The method can be performed by a drone detection device, which can be implemented in the form of hardware and / or software. The drone detection device can be configured in an electronic device with certain data computing capabilities, which can be a terminal device, which can be a laptop, a mobile phone or a tablet computer, etc. The terminal device can be configured in a mobile monitoring vehicle.

[0016] See also Figure 1 The drone detection method shown includes: S101. When a control signal of a target UAV is identified, video of the target UAV is captured according to the control signal of the target UAV to obtain a target video.

[0017] The control signal is a radio signal, and the target drone may be a drone that needs to be detected. For a specific scene space, the target drone may be an interference drone that needs to be controlled; or the target drone is a drone that needs to be controlled. The control signal is used to control the target drone. The video acquisition device is controlled to track and shoot the target drone to obtain the target video. In the target video, there may be a video image that does not include the target drone, at least there is a video image that includes the target drone, and there may be a video image that includes other drones or aircraft other than the target drone.

[0018] In an optional embodiment, the method of performing video capture on the target UAV to obtain the target video based on the control signal of the target UAV includes: orienting the control signal to the signal source to obtain the direction of the target UAV; and performing video capture in the direction of the target UAV to obtain the target video.

[0019] The signal source is oriented according to the control signal to obtain the direction of the signal source, and the direction of the signal source represents the direction of the drone. Video can be collected toward the sky along the direction of the signal source to obtain the target video.

[0020] In fact, there may be multiple redundant drones in the sky, and these redundant drones may be located in different directions. For example, the directions of multiple signal sources can be used as the directions of the target drone, and video capture can be performed separately to obtain the target videos of the drones pointed to by each signal source. The target video of the target drone can then be determined by manually browsing the video. For another example, when determining the direction in which there are multiple signal sources, the user selects the direction of one signal source as the direction of the target drone. For another example, when determining the direction in which there are multiple signal sources, the direction in which there are a large number of signal sources or the signal sources are concentrated is selected and determined as the direction of the target drone.

[0021] In some embodiments, based on the received radio signal, it is detected whether there is a control signal of an unknown drone; when there is a control signal of an unknown drone, it is determined that the control signal of the target drone is identified.

[0022] In some embodiments, a control signal of a currently controlled drone is obtained to determine whether a control signal of a target drone is identified.

[0023] It can be seen that by orienting the control signal source and then acquiring video based on the direction of the target UAV to obtain the target video, the direction of the UAV can be quickly located, and then the video can be acquired in the direction of the UAV to locate the UAV in the target video. After two stages of UAV positioning, the UAV can be quickly located, and pixel-level positioning can be achieved, thereby improving the positioning accuracy of the UAV.

[0024] In an optional embodiment, the radio signal is collected by a radio signal device on a mobile monitoring vehicle, and the target video is collected by a camera on the mobile monitoring vehicle.

[0025] Among them, the mobile monitoring vehicle is used to travel along a preset trajectory and detect radio signals in real time. The radio signal device is used to collect radio signals. The camera is used to collect images and videos. The camera hardware is initialized to prepare for the capture of the video stream; after determining the direction of the target drone, the camera is controlled to start capturing the video stream in the direction of the target drone, and the real-time collected video data is stored to form the target video. The embodiment of the present invention helps the radio monitoring vehicle to continuously and accurately identify and track the drone, and realizes high-precision and low-system-overhead control of the drone.

[0026] It can be seen that by monitoring the radio signals of drones in real time during the mobile monitoring of radio signals, and detecting the target video of drones in real time to locate the target drones, real-time positioning and supervision of drones can be achieved in radio supervision scenarios.

[0027] S102: Perform drone detection on each video image in the target video to obtain a drone detection result for each video image.

[0028] A video image is a frame of a target video. Whether a drone exists in a video image is detected to obtain a drone detection result of the video image. The drone detection result may include at least one detection frame and classification results of each detection frame. The classification result of the detection frame may include the model of the drone.

[0029] S103: Determine the flight trajectory of the target drone according to the playback order of each video image in the target video and the drone detection result of each video image.

[0030] Among them, the target drone to be detected can be determined from the at least one identified drone, so that only the position of the target drone is determined in each video image. According to the position of the identified target drone in each video image and the playback order of each video image, the motion information of the target drone can be determined. According to the motion information of the target drone, the flight trajectory of the target drone can be determined.

[0031] Analyze the video stream to detect and confirm the drone in the target video. Not only can the exact location of the drone be located, but also its model or category can be classified and marked. Through this high-precision identification process, detailed and critical information can be provided for drone monitoring tasks.

[0032] In an optional embodiment, determining the flight trajectory of the target UAV according to the playback order of each of the video images in the target video and the UAV detection results of each of the video images includes: determining the traveling direction of the target UAV according to the playback order of each of the video images in the target video and the position of the target UAV in the UAV detection results of each of the video images; determining the traveling speed of the target UAV according to the acquisition period of the target video and the position of the target UAV in the UAV detection results; determining the flight trajectory of the target UAV according to the position, traveling direction and traveling speed of the target UAV in the UAV detection results.

[0033] Among them, the spatial position of the target drone is determined according to the position of the target drone in the drone detection results of each video image, as well as the angle and spatial position of the camera. According to the spatial position of the target drone and the video playback order, the movement order of each spatial position can be determined, thereby determining the driving direction of the target drone. The acquisition cycle of the target video can refer to the time interval between each video image. According to the time interval between each video image and the spatial position of the target drone in each video image, the driving speed of the target drone moving from the video image in the previous sequence to the video image in the next sequence in each two adjacent video images is determined. After determining the position, driving direction and driving speed of the target drone, the flight trajectory of the target drone can be determined.

[0034] It can be seen that by determining the flight trajectory of the target drone through the information of the target video and the drone detection results, the target drone can be continuously monitored and tracked, and accurate positioning information of the target drone can be provided to achieve precise pixel-level tracking, ensuring the continuity and effectiveness of the monitoring activities and achieving accurate drone monitoring effects.

[0035] The technical solution of the embodiment of the present invention determines the collection direction of the target UAV according to the control signal of the target UAV, and collects video of the target UAV based on the collection direction to obtain the target video, and locates the target UAV in the video, so as to determine the flight trajectory of the target UAV according to the positioning information, thereby realizing the positioning of the UAV by combining radio signal and image target detection, thereby improving the positioning accuracy of the UAV. Figure 2 A flowchart of a drone detection method provided by an embodiment of the present invention.

[0036] See also Figure 2 The drone detection method shown includes: S201. When a control signal of a target UAV is identified, video of the target UAV is captured according to the control signal of the target UAV to obtain a target video.

[0037] The target video can be uploaded and stored in real time. The target video is transmitted through wireless or wired network to ensure the timeliness and integrity of monitoring information, and provide original data for subsequent analysis and processing, so as to facilitate reference and evidence collection.

[0038] S202: For each video image in the target video, input the video image into a trained target detection model for processing, and output a drone detection result of the video image.

[0039] The target detection model is used to perform target detection on the input video image to obtain the drone detection result of the video image. The target detection model is used to detect the detection frame of the drone in the video image and the type of drone in each detection frame.

[0040] In some embodiments, the target detection model can be a YOLO FastR-CNN (You Only Look Once Faster Region-based Convolutional Neural Networks) model. A collection of labeled images of drones of various models and sizes can be collected and constructed to obtain a labeled training data set; wherein the training data set brings together a wealth of drone image data, which covers a wide range of drone models, and each model is sampled from multiple angles, and these images are standardized to ensure that all images are of the same size, and then the drones in each image are labeled in detail. Through the training data set and the loss function, the YOLO FastR-CNN model is trained to obtain the target detection model. By optimizing the hyperparameters of the network, the stability of the gradient descent process is achieved, ensuring that the loss function value drops to the target level, and the model's fit meets the established standards, thereby constructing an efficient target detection model. The YOLO FastR-CNN model achieves higher accuracy and better stability in recognition and tracking performance, providing great convenience for monitoring personnel.

[0041] S203: Determine the flight trajectory of the target drone according to the playback order of each video image in the target video and the drone detection result of each video image.

[0042] Among them, DeepSORT (Deep Simple Online and Realtime Tracking, a drone real-time online multi-target tracking algorithm based on deep learning) can be used to significantly improve the tracking speed while ensuring the tracking accuracy. It effectively solves the balance problem between real-time and accuracy in traditional target tracking technology, is suitable for high-demand practical application scenarios, and can meet the target tracking needs that place equal emphasis on real-time and accuracy.

[0043] The technical solution of the embodiment of the present invention can improve the accuracy of drone detection by identifying drones in video images through a target detection model, and can also quickly realize drone detection and achieve instant analysis and processing of collected images or frequencies.

[0044] In some optional embodiments, before inputting each video image in the target video into a trained target detection model for processing and outputting the drone detection result of the video image, it also includes: inputting each video image in the target video into an image enhancement model for processing, outputting the enhanced video images, and updating each video image.

[0045] Among them, the image enhancement model is used to enhance the video image, specifically to improve the image clarity or resolution.

[0046] It can be seen that by performing image enhancement on the video image before performing target detection on the video image, the image quality can be quickly improved, thereby improving the target detection accuracy of the image.

[0047] In some optional embodiments, the image enhancement model is trained by: acquiring a first image and a target image, the target image being a high-quality image and the first image being a low-quality image; inputting the first image into a generator and outputting a second image; inputting the target image and the second image into a discriminator respectively to obtain a true or false discrimination result of the target image and a true or false discrimination result of the second image; determining the adversarial loss of the discriminator according to the true or false discrimination result of the target image and the true or false discrimination result of the second image, and adjusting the parameters of the discriminator based on the adversarial loss; performing feature extraction on the second image. The method comprises the steps of: extracting the target image to obtain the intermediate features of the target image and the output features of the target image; calculating the difference between the output features of the second image and the output features of the target image to obtain the content loss; calculating the difference between the intermediate features of the second image and the intermediate features of the target image to obtain the perceptual loss; determining the adversarial loss of the generator according to the true and false discrimination result of the second image; adjusting the parameters of the generator according to the adversarial loss, content loss and perceptual loss of the generator to obtain a trained image enhancement model.

[0048] Among them, the image enhancement model can be a generator in a Generative Adversarial Network (GAN). GAN includes: a generator and a discriminator. The generator is used to generate images that are similar to real images and have better quality, trying to "cheat" the discriminator so that it cannot distinguish between generated data and real data. The task of the discriminator is to distinguish between real data and fake data generated by the generator, and to reduce the probability of being deceived by the generator by improving the discrimination ability.

[0049] The first image and the target image are a pair of images with the same content, the first image has a lower quality, and the target image has a higher quality. Both the first image and the target image are real images. The target image can be first acquired, and the target image can be subjected to a quality reduction process such as reducing the resolution to obtain the first image. The second image is an image generated based on the first image. The second image is not a real image. The discriminator is used to classify images and distinguish between real images and generated images. The adversarial loss may include a first difference between the classification result of the first image and the classification result of the real image, a second difference between the classification result of the target image and the classification result of the real image, and a third difference between the classification result of the second image and the classification result of the generated image. The loss values ​​corresponding to the first difference, the second difference, and the third difference can be calculated respectively, and the three loss values ​​can be fused to obtain the adversarial loss.

[0050] The generator is used to generate a higher quality image with the same content as the first image. The loss of the generator may include the fourth difference between the classification result of the second image and the classification result of the real image, the fifth difference between the content of the second image and the content of the target image, and the sixth difference between the intermediate features of the second image and the intermediate features of the target image, etc. The loss values ​​corresponding to the first difference, the second difference, and the third difference may be calculated respectively, and the three loss values ​​may be fused to obtain the loss of the generator.

[0051] The parameters of the generator can be fixed first, and the parameters of the discriminator can be adjusted according to the adversarial loss first. When the adversarial loss is minimized, converged, or the number of iterations is greater than or equal to a preset threshold, the current discriminator is determined as the trained discriminator; then the parameters of the discriminator are fixed, and the loss of the generator is adjusted according to the loss of the generator. When the loss of the generator is minimized, converged, or the number of iterations is greater than or equal to a preset threshold, the current generator is determined as the trained generator, and the trained generator is determined as the trained image enhancement model.

[0052] In one example, the generator usually adopts an encoder-decoder structure. The encoder part consists of a series of convolutional layers, which is used to extract the features of the input low-quality image and gradually compress the image into a low-dimensional feature representation; the decoder part consists of a series of deconvolutional layers (transposed convolutional layers), which gradually upsamples the low-dimensional feature representation and restores it to a high-resolution enhanced image; in the generator, skip connections can also be used to directly connect the features of the encoder part to the corresponding layer of the decoder, which helps to retain the detailed information of the image and make the generated enhanced image more realistic. The discriminator is generally composed of multiple convolutional layers, which are used to extract and discriminate the features of the input image. The discriminator gradually reduces the resolution of the image through convolution operations, while increasing the number of feature channels to extract high-level features of the image; finally, a fully connected layer maps the extracted features to a scalar value, indicating the probability that the input image is a real image.

[0053] Adversarial Loss: Measures the adversarial relationship between the image generated by the generator and the real image. The generator hopes to minimize the adversarial loss so that it is difficult for the discriminator to distinguish between the generated enhanced image and the real high-quality image; while the discriminator hopes to maximize the adversarial loss to accurately distinguish between real and fake images.

[0054] For the generator: For the discriminator: Among them, z is the noise vector input to the generator, x is the real high-quality image, D is the discriminator, and G is the generator.

[0055] Content Loss: It is used to ensure that the generated enhanced image is consistent with the input low-quality image in terms of content, while having high-quality features. Usually, a pre-trained convolutional neural network (such as VGG) is used to extract the features of the image, and then the distance between the generated image and the real image in the feature space is calculated. The commonly used content loss is the mean square error (MSE) loss, and the mathematical expression is: Among them, is a pre-trained convolutional neural network used to extract image features.

[0056] Perceptual Loss: Similar to content loss, but more focused on the perceptual similarity of images. Perceptual loss measures the perceptual difference between images by comparing the features of generated images and real images at different levels of the pre-trained network. Perceptual loss can make the generated images visually closer to real images, enhancing the quality and authenticity of the images.

[0057] The parameters of the discriminator and generator are updated through back propagation to make the images they generate more realistic and meet the requirements. The discriminator and generator are trained alternately until the loss converges or the predetermined number of training rounds is reached.

[0058] It can be seen that by training the generative adversarial network and enhancing the image, the video image can be improved quickly and accurately, so as to perform target detection in high-quality video images and improve the accuracy of target detection. The detail performance of the video image can be optimized, thereby providing more vivid and accurate visual information for the detection of drones. The video frame can be enhanced in detail and noise suppressed, thereby improving the overall quality of the visual data and providing better visual information for the accurate identification and tracking of drones.

[0059] In one example, Figure 2-Figure 5 A schematic diagram of the identification and tracking of a drone is provided. Figure 6 A schematic diagram of target detection results in a video image is provided.

[0060] Figure 7 A schematic diagram of the structure of a drone detection device provided by an embodiment of the present invention. The device can execute a drone detection method, the device can be implemented in the form of hardware and / or software, and the device can be configured in an electronic device that carries a function with certain data computing capabilities.

[0061] See also Figure 7 The drone detection device shown includes: The video acquisition module 701 is used to acquire a video of the target UAV according to the control signal of the target UAV when the control signal of the target UAV is identified, so as to obtain a target video; The drone detection module 702 is used to perform drone detection on each video image in the target video to obtain a drone detection result for each video image; The flight trajectory determination module 703 is used to determine the flight trajectory of the target drone according to the playback order of each video image in the target video and the drone detection result of each video image.

[0062] Optionally, the drone detection module 702 is specifically used for: For each video image in the target video, the video image is input into a trained target detection model for processing, and a drone detection result of the video image is output.

[0063] Optionally, the drone detection device further includes: The image enhancement module is used to input each video image in the target video into a trained target detection model for processing, and before outputting the drone detection result of the video image, input each video image in the target video into an image enhancement model for processing, output each enhanced video image, and update each video image.

[0064] Optionally, the drone detection device further includes: an image enhancement model training module, which is used to: Acquire a first image and a target image, wherein the target image is a high-quality image and the first image is a low-quality image; Input the first image into a generator and output a second image; Inputting the target image and the second image into a discriminator respectively, to obtain a true or false discrimination result of the target image and a true or false discrimination result of the second image; Determining the adversarial loss of the discriminator according to the true and false discrimination result of the target image and the true and false discrimination result of the second image, and adjusting the parameters of the discriminator based on the adversarial loss; Performing feature extraction on the second image to obtain intermediate features of the second image and output features of the second image; Performing feature extraction on the target image to obtain intermediate features of the target image and output features of the target image; Calculating a difference between an output feature of the second image and an output feature of the target image to obtain a content loss; Calculating a difference between an intermediate feature of the second image and an intermediate feature of the target image to obtain a perceptual loss; Determining the adversarial loss of the generator according to the true and false discrimination result of the second image; According to the adversarial loss, content loss and perceptual loss of the generator, the parameters of the generator are adjusted to obtain a trained image enhancement model.

[0065] Optionally, the video acquisition module 701 is specifically used for: Directing the control signal to obtain the direction of the target UAV; Video is collected in the direction of the target UAV to obtain a target video.

[0066] Optionally, the radio signal is collected by a radio signal device on a mobile monitoring vehicle, and the target video is collected by a camera on the mobile monitoring vehicle.

[0067] Optionally, the flight trajectory determination module 703 is specifically configured to: Determine the travel direction of the target drone according to the playback order of each of the video images in the target video and the position of the target drone in the drone detection results of each of the video images; Determine the travel speed of the target drone according to the acquisition period of the target video and the position of the target drone in the drone detection result; The flight trajectory of the target UAV is determined according to the position, driving direction and driving speed of the target UAV in the UAV detection result.

[0068] In the technical solutions of the embodiments of the present invention, the acquisition, storage and application of network traffic etc. involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals. Figure 8 A schematic diagram of the structure of an electronic device 800 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0069] like Figure 8As shown, the electronic device 800 includes at least one processor 801, and a memory connected to the at least one processor 801 in communication, such as a read-only memory (ROM) 802, a random access memory (RAM) 803, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 801 can perform various appropriate actions and processes according to the computer program stored in the ROM 802 or the computer program loaded from the storage unit 808 to the RAM 803. In the RAM 803, various programs and data required for the operation of the electronic device 800 can also be stored. The processor 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0070] Multiple components in the electronic device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0071] The processor 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 801 performs the various methods and processes described above, such as the drone detection method.

[0072] In some embodiments, the drone detection method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the processor 801, one or more steps of the drone detection method described above may be performed. Alternatively, in other embodiments, the processor 801 may be configured to perform the drone detection method in any other appropriate manner (e.g., by means of firmware).

[0073] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0074] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0075] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a RAM, a ROM, an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0076] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0077] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0078] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS (Virtual Private Server) services.

[0079] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0080] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A drone detection method, characterized in that: The method comprises: When a control signal of a target UAV is identified, a video of the target UAV is captured according to the control signal of the target UAV to obtain a target video; Performing drone detection on each video image in the target video to obtain a drone detection result for each video image; The flight trajectory of the target drone is determined according to the playback order of each of the video images in the target video and the drone detection result of each of the video images.

2. The method according to claim 1, characterized in that The performing drone detection on each video image in the target video to obtain a drone detection result of each video image includes: For each video image in the target video, the video image is input into a trained target detection model for processing, and a drone detection result of the video image is output.

3. The method according to claim 2, characterized in that For each video image in the target video, the video image is input into a trained target detection model for processing, and before the drone detection result of the video image is output, the method further includes: For each video image in the target video, each video image is input into an image enhancement model for processing, each enhanced video image is output, and each video image is updated.

4. The method according to claim 3, characterized in that The image enhancement model is trained in the following way: Acquire a first image and a target image, wherein the target image is a high-quality image and the first image is a low-quality image; Input the first image into a generator and output a second image; Inputting the target image and the second image into a discriminator respectively, to obtain a true or false discrimination result of the target image and a true or false discrimination result of the second image; Determining the adversarial loss of the discriminator according to the true and false discrimination result of the target image and the true and false discrimination result of the second image, and adjusting the parameters of the discriminator based on the adversarial loss; Performing feature extraction on the second image to obtain intermediate features of the second image and output features of the second image; Performing feature extraction on the target image to obtain intermediate features of the target image and output features of the target image; Calculating a difference between an output feature of the second image and an output feature of the target image to obtain a content loss; Calculating a difference between an intermediate feature of the second image and an intermediate feature of the target image to obtain a perceptual loss; Determining the adversarial loss of the generator according to the true and false discrimination result of the second image; According to the adversarial loss, content loss and perceptual loss of the generator, the parameters of the generator are adjusted to obtain a trained image enhancement model.

5. The method according to claim 1, characterized in that The step of collecting video of the target UAV according to the control signal of the target UAV to obtain the target video includes: Directing the control signal to obtain the direction of the target UAV; Video is collected in the direction of the target UAV to obtain a target video.

6. The method according to claim 5, characterized in that The radio signal is collected by a radio signal device on a mobile monitoring vehicle, and the target video is collected by a camera on the mobile monitoring vehicle.

7. The method according to claim 1, characterized in that The step of determining the flight trajectory of the target drone according to the playback order of each of the video images in the target video and the drone detection result of each of the video images comprises: Determine the travel direction of the target drone according to the playback order of each of the video images in the target video and the position of the target drone in the drone detection results of each of the video images; Determine the travel speed of the target drone according to the acquisition period of the target video and the position of the target drone in the drone detection result; The flight trajectory of the target UAV is determined according to the position, driving direction and driving speed of the target UAV in the UAV detection result.

8. A drone detection device, characterized in that: include: A video acquisition module is used to acquire a video of the target UAV according to the control signal of the target UAV when the control signal of the target UAV is identified, so as to obtain a target video; A drone detection module is used to perform drone detection on each video image in the target video to obtain a drone detection result for each video image; The flight trajectory determination module is used to determine the flight trajectory of the target drone according to the playback order of each video image in the target video and the drone detection result of each video image.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the drone detection method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the drone detection method according to any one of claims 1 to 7 when executed.

Images