A method and apparatus for identifying a UAV, an electronic device, and a computer program product

Abstract

The present disclosure provides a UAV identification method and device, electronic equipment and computer program product, relating to the technical field of UAVs. The UAV identification method of the present disclosure comprises: acquiring a radio frequency signal; preprocessing the radio frequency signal to obtain initial time-frequency data; performing denoising processing on the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain target time-frequency data; extracting signal features in the target time-frequency data, and determining the corresponding UAV type based on the signal features.

Description

Technical Field

[0001] This specification relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to a UAV identification method, device, electronic device, and computer program product. Background Technology

[0002] With the rapid development of drone technology, drones are widely used in logistics, surveying, agriculture, security, and many other fields. However, this also brings many security risks, such as unauthorized drone flights and malicious airspace intrusions. Therefore, the supervision and identification of drones is particularly important. Drones can communicate wirelessly with other devices (such as remote control devices) via radio frequency signals. Different types of drones have significant differences in signal characteristics, and the type of drone can be identified by analyzing the signal characteristics in the radio frequency signal. However, drone radio frequency signals are susceptible to environmental noise interference, and current denoising algorithms tend to introduce too much background noise, affecting the accuracy of drone identification.

[0003] In view of this, some embodiments of this specification provide a drone identification method, apparatus, electronic device, and computer program product, which are designed to effectively remove background noise and improve the accuracy of drone type identification. Summary of the Invention

[0004] This specification provides one or more embodiments of a drone identification method, the method comprising: acquiring a radio frequency signal; preprocessing the radio frequency signal to obtain initial time-frequency data; denoising the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain target time-frequency data; extracting signal features from the target time-frequency data, and determining the corresponding drone type based on the signal features.

[0005] According to one or more embodiments of this specification, a method for denoising initial time-frequency data based on the average amplitude of a radio frequency signal in initial time-frequency data to obtain target time-frequency data includes: extracting noise time-frequency data corresponding to noise periods from the initial time-frequency data based on a first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data; generating average background time-frequency data based on a second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency point in the noise time-frequency data; and denoising the initial time-frequency data based on the average background time-frequency data to obtain target time-frequency data; wherein the amplitude value is used to characterize the signal strength of the radio frequency signal at a specific time point and a specific frequency point.

[0006] According to one or more embodiments of this specification, a method for extracting noise time-frequency data corresponding to a noise period from initial time-frequency data is provided based on a first average value of the amplitude values ​​of the radio frequency signal at different frequency points at different time points in the initial time-frequency data. The method includes: determining a first average value of the amplitude values ​​of the radio frequency signal at different frequency points at different time points in the initial time-frequency data; filtering each first average value based on a first preset threshold and the magnitude of each first average value; determining the time point corresponding to the filtered first average value as a noise period; and extracting noise time-frequency data corresponding to the noise period from the initial time-frequency data.

[0007] According to one or more embodiments of this specification, a method for generating average background time-frequency data based on a second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency in noise time-frequency data includes: determining a second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency in noise time-frequency data; determining the amplitude value corresponding to each frequency point at each time point based on the second average value corresponding to each frequency point; and generating average background time-frequency data.

[0008] According to one or more embodiments of this specification, the method extracts signal features from target time-frequency data and determines the corresponding UAV type based on the signal features, including: acquiring image transmission signal features from target time-frequency data, the image transmission signal features including periodic features of the image transmission signal; matching the image transmission signal features with signal features in a first preset feature library to determine the corresponding UAV type.

[0009] According to one or more embodiments of this specification, the method extracts signal features from target time-frequency data and determines the corresponding UAV type based on the signal features, including: when there is a valid radio frequency signal in the target time-frequency data within a preset frequency range, compressing the target time-frequency data within the preset frequency range; matching the compressed target time-frequency data with time-frequency data in a second preset feature library to determine the corresponding UAV type.

[0010] According to one or more embodiments of this specification, a method for extracting signal features from target time-frequency data and determining the corresponding UAV type based on the signal features includes: extracting frequency-hopping block data from the target time-frequency data; matching the frequency of the frequency-hopping block data with a preset frequency set; when the matching is successful, obtaining the frequency-hopping signal features of the frequency-hopping block data, wherein the frequency-hopping signal features include at least one of the following features of the frequency-hopping signal: dwell time, signal bandwidth, time domain interval, and frequency domain interval; and matching the frequency-hopping signal features with signal features in a third preset feature library to determine the corresponding UAV type.

[0011] According to one or more embodiments of this specification, after acquiring the radio frequency signal, the method further includes: extracting identification information from the radio frequency signal and matching it with information in a preset information database; when the match is successful, determining the corresponding drone type.

[0012] According to one or more embodiments of this specification, a method for acquiring radio frequency signals includes: acquiring set target acquisition parameters, the target acquisition parameters including at least one of a target center frequency, a target bandwidth, and a target sampling rate; and acquiring radio frequency signals acquired by a detection device based on the target acquisition parameters.

[0013] One or more embodiments of this specification also provide a drone identification device, the device comprising: an acquisition module for acquiring radio frequency signals; a preprocessing module for preprocessing the radio frequency signals to obtain initial time-frequency data; a denoising module for denoising the initial time-frequency data based on the average amplitude of the radio frequency signals in the initial time-frequency data to obtain target time-frequency data; and a type determination module for extracting signal features from the target time-frequency data and determining the corresponding drone type based on the signal features.

[0014] This specification also provides an electronic device in one or more embodiments, characterized in that it includes: one or more processors; and a memory for storing one or more programs, which, when executed by one or more processors, can implement the drone identification method described in some embodiments of this specification.

[0015] One or more embodiments of this specification also provide a computer program product, including a computer program that, when at least a portion of the computer program is executed by a processor, can implement the drone identification method described in some embodiments of this specification. Attached Figure Description

[0016] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. The same numbers in the drawings denote the same structures or steps.

[0017] Figure 1 This is an exemplary flowchart of a drone identification method according to some embodiments of this specification.

[0018] Figure 2 This is a schematic diagram of the initial time-frequency data of a regular image transmission signal according to some embodiments of this specification.

[0019] Figure 3 This is a schematic diagram of the initial time-frequency data of an irregular image transmission signal according to some embodiments of this specification.

[0020] Figure 4This is a schematic diagram of the initial time-frequency data of a frequency hopping signal according to some embodiments of this specification.

[0021] Figure 5 This is a schematic diagram of a timing sequence according to some embodiments of this specification.

[0022] Figure 6 This is a schematic diagram of a frequency domain sequence according to some embodiments of this specification.

[0023] Figure 7 This is a schematic diagram of the target time-frequency data of a regular image transmission signal according to some embodiments of this specification.

[0024] Figure 8 This is a schematic diagram of the target time-frequency data of a frequency hopping signal according to some embodiments of this specification.

[0025] Figure 9 This is an exemplary flowchart illustrating another drone identification method according to some embodiments of this specification.

[0026] Figure 10 This is an exemplary block diagram of a drone identification device according to some embodiments of this specification. Detailed Implementation

[0027] To more clearly illustrate the technical solutions of the embodiments in this specification, the embodiments will be described in detail below with reference to the accompanying drawings. Obviously, the content described below are some examples or embodiments of this specification. For those skilled in the art, without creative effort, the technical solutions or means disclosed in this specification can be applied to other scenarios based on this technical content.

[0028] It should be understood that the terms "system," "device," "unit," and / or "module" used in this specification are a method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0029] Unless otherwise specified, the technical terms used to describe components, elements, etc. in this specification are not singular but may include plural. Generally speaking, terms such as "comprising" or "including" only indicate that explicitly identified steps, elements, or components are included, and these steps, elements, and components do not constitute an exclusive list, as the described method or apparatus may also include other steps or components.

[0030] This specification uses flowcharts to illustrate the operational steps performed by the apparatus or system of related embodiments. However, unless otherwise specified, the order in which these steps are described should not be construed as a limitation on the order of execution. Those skilled in the art can adjust the order of these steps based on the knowledge and information conveyed by the embodiments in this specification. Such adjustments include, but are not limited to, reversing the order of steps, merging multiple steps, and splitting a step.

[0031] Unmanned Aerial Vehicles (UAVs), also known as unmanned aircraft or drones, are unmanned aerial vehicles controlled by radio remote control equipment or their own programmed control devices. They are widely used in logistics, surveying, agriculture, security, and many other fields. While the widespread use of UAVs brings convenience, it also brings many safety hazards, such as unauthorized UAV flights and malicious airspace intrusions. Therefore, the supervision and identification of UAVs is particularly important. UAV identification can be based on UAV type; for example, UAV type can include UAV model or UAV classification. Accurately identifying the type of UAV can not only determine whether the UAV is flying legally but also assess its performance parameters.

[0032] In some related embodiments, drones can broadcast or communicate wirelessly with other devices via radio frequency (RF) signals. For example, a drone can communicate wirelessly with a remote control device that controls it via RF signals. Different types of drones exhibit significant differences in signal characteristics. For instance, different types of drones differ in parameters such as signal modulation methods, frequency ranges, and signal modes (e.g., frequency hopping mode, fixed frequency mode). The type of drone can be identified by analyzing these signal characteristics in the RF signal. In some related embodiments, a threshold can be set to separate the drone's RF signal from background noise; however, drone RF signals are susceptible to environmental noise interference, and current denoising algorithms tend to introduce excessive background noise, affecting the accuracy of drone identification.

[0033] Therefore, some embodiments of this specification propose a drone identification method, which obtains initial time-frequency data by preprocessing the radio frequency signals collected by the detection device, denoising the initial time-frequency data based on the average amplitude of the radio frequency signals in the initial time-frequency data to obtain target time-frequency data, extracting signal features from the target time-frequency data, and determining the corresponding drone type based on the signal features, thereby effectively removing background noise and improving the accuracy of drone type identification.

[0034] Figure 1 This is an exemplary flowchart of a drone identification method according to some embodiments of this specification. Figure 1 The illustrated process 100 can be executed by a terminal device, such as an electronic detection device. In some embodiments, process 100 can be implemented by a drone identification device 1000 deployed on the terminal device. Figure 1 As shown, in some embodiments, process 100 may include the following steps.

[0035] Step 110: Acquire radio frequency signal. In some embodiments, step 110 may be implemented by acquisition module 1010.

[0036] In some embodiments, radio frequency (RF) signals refer to modulated high-frequency electromagnetic wave signals, typically ranging from 300 kHz to 300 GHz. In some embodiments, RF signals can be acquired using an RF acquisition board in a terminal device. The RF acquisition board can preset multiple acquisition parameters, including center frequency, bandwidth, and sampling rate. For example, the RF acquisition board can preset multiple center frequencies such as 400 MHz, 900 MHz, 1.4 GHz, 2.4 GHz, and 5.2 GHz. Users can set target acquisition parameters according to their needs to customize the analysis of RF signals in specific frequency bands. Target acquisition parameters can include at least one of a target center frequency, target bandwidth, and target sampling rate.

[0037] In some embodiments, the terminal device can acquire set target acquisition parameters and acquire radio frequency signals based on the target acquisition parameters. For example, the set target center frequency can be 2.4 GHz, the set target bandwidth can be 40 MHz, and the set target sampling rate can be 200 MS / s. The terminal device can acquire and obtain radio frequency signals based on the above-set target acquisition parameters.

[0038] Step 120: Preprocessing is performed based on the radio frequency signal to obtain initial time-frequency data. In some embodiments, step 120 can be implemented by the preprocessing module 1020.

[0039] In some embodiments, preprocessing based on the radio frequency (RF) signal may include performing a short-time Fourier transform (STFT) on the RF signal to convert it into initial time-frequency data. In some embodiments, the initial time-frequency data can be characterized by a time-spectrum graph, which graphically represents the distribution of the RF signal in time and frequency. For example, the horizontal axis of the initial time-spectrum graph can represent time, the vertical axis can represent frequency, and the brightness of each pixel in the initial time-spectrum graph can represent the amplitude value of the RF signal at that pixel (corresponding to the time and frequency points). In other embodiments, the initial time-frequency data can also be characterized by matrix data corresponding to the initial time-spectrum graph, so that the calculation results of the initial time-frequency data are stored in matrix form, where each element of the matrix can represent the amplitude value of the RF signal at a specific time and frequency point.

[0040] In some embodiments, the acquired radio frequency signals may include image transmission signals, which can be used to transmit image or video data between the drone and other devices (such as remote control devices that control the drone). Image transmission signals may include regular image transmission signals and irregular image transmission signals. Regular image transmission signals typically use fixed frequency bands to transmit image or video data, while irregular image transmission signals may use dynamic frequency bands, dynamic time intervals, or other randomized strategies to transmit image or video data during transmission. Figure 2 This is a schematic diagram of the initial time-frequency data of a regular image transmission signal according to some embodiments of this specification, such as... Figure 2 As shown, the regular image transmission signal uses a fixed frequency band to transmit data, presenting regular narrowband frequency stripes in the vertical direction, such as... Figure 2 The rectangular stripes in the signal indicate that the center frequency and bandwidth of the signal remain approximately constant at different time points. Figure 3 This is a schematic diagram of the initial time-frequency data of an irregular image transmission signal according to some embodiments of this specification, such as... Figure 3 As shown, unlike regular narrowband frequency stripes, irregular image transmission signals appear as non-rectangular patterns with multiple peaks within the signal bandwidth, such as... Figure 3 As shown, there are three wave peaks above the hill shape, with relatively stable frequencies, while the frequencies of the wave peaks at the hill shape change over time.

[0041] In some embodiments, the acquired radio frequency signals may include frequency hopping signals, which can be used to transmit control commands, such as flight direction commands, altitude adjustment commands, etc., between the drone and other devices (such as remote control devices that control the drone). Figure 4 This is a schematic diagram of the initial time-frequency data of a frequency-hopping signal according to some embodiments of this specification, such as... Figure 4 As shown, frequency hopping signals can switch between multiple frequency points in a short period of time.

[0042] In some embodiments, the initial time-frequency data can be compressed to reduce its storage space. For example, the resolution of the initial time-frequency spectrogram can be compressed to reduce its storage space.

[0043] Step 130: Denoise the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain the target time-frequency data. In some embodiments, step 130 can be implemented by the denoising processing module 1030.

[0044] In some embodiments, noise time-frequency data corresponding to the noise period can be extracted from the initial time-frequency data based on the first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point. The following describes in detail how to obtain the noise time-frequency data.

[0045] In some embodiments, a first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data can be determined first. The amplitude value is used to characterize the signal strength of the radio frequency signal at a specific time point and a specific frequency point. For each time point, the first average value of the amplitude values ​​corresponding to different frequency points can be the first average value of the amplitude values ​​of the radio frequency signal corresponding to different frequency points at the same time point; in other words, the first average value can be the average value of the amplitude values ​​corresponding to the radio frequency signal in the frequency domain direction (or frequency direction). For example, suppose the time points... The corresponding frequency points are respectively , , Radio frequency signal at time point Frequency point , , The corresponding amplitude values ​​are S( ), S ( ), S ( Then the radio frequency signal at time point Frequency point , , The first average of the corresponding amplitude values ​​is = Furthermore, according to the aforementioned determined time points... Frequency point , , The corresponding first average value The method calculates each time point Different frequency points The first average of the corresponding amplitude values This generates a time series of average amplitude values. Figure 5 This is a schematic diagram of a timing sequence according to some embodiments of this specification. For example... Figure 5 As shown, the horizontal axis can represent time. The vertical axis can represent the first average value. First average Over time The changes can be as follows Figure 5 As shown.

[0046] In some embodiments, after determining the first average value of each amplitude value corresponding to different frequency points of the radio frequency signal at each time point based on the initial time-frequency data, the first average values ​​can be filtered based on a first preset threshold and the magnitude of each first average value, and the time point corresponding to the filtered first average value can be determined as a noise period. According to one example, the first preset threshold can be a preset percentage, and the first average values ​​can be sorted, for example, in ascending order. The time point corresponding to the first average value ranking first in the sorted sequence (a preset percentage) can be determined as a noise period. For example, the first average values ​​are sorted in ascending order as follows: , , , , , , , , , If the first preset threshold is 50%, then the average value of the top 50% of the sorted sequence will be used. , , , , corresponding time point , , , , The time period is identified as a noise period. According to another example, the first preset threshold can be a preset average amplitude value, and time points corresponding to the first average value below the first preset threshold can be identified as noise periods. For example, assume each first average value... The first average values ​​of the values ​​below the first preset threshold are respectively , , , , Then the first average value can be used. , , , , corresponding time point , , , , This period was identified as a noisy time.

[0047] In some embodiments, noise time-frequency data corresponding to a determined noise period can be extracted from the initial time-frequency data. For example, taking the foregoing embodiment as an example, the noise period can be... , , , , The corresponding initial time-frequency data is extracted as noise time-frequency data.

[0048] In some embodiments, average background time-frequency data can be generated based on the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency in the noise time-frequency data. The following describes in detail how to generate average background time-frequency data.

[0049] In some embodiments, a second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency point in the noise time-frequency data can be determined first. The amplitude value is used to characterize the signal strength of the radio frequency signal at a specific time point and a specific frequency point. For each frequency point, the second average value of the amplitude values ​​corresponding to different time points can be the second average value of the amplitude values ​​of the radio frequency signal at different time points at the same frequency point; in other words, the second average value can be the average value of the amplitude values ​​corresponding to the radio frequency signal in the time domain (or time direction). For example, assuming the frequency point... The corresponding time points are respectively , , Radio frequency signals at frequency points Time point , , The corresponding amplitude values ​​are S( ), S ( ), S ( Then the radio frequency signal at the frequency point Time point , , The second average of the corresponding amplitude values ​​is = Furthermore, the frequency points are determined according to the above method. Time point , , The corresponding second average value The method calculates each frequency point. Different time points The second average value of each corresponding amplitude value This generates a frequency domain sequence of the average amplitude. Figure 6 This is a schematic diagram of a frequency domain sequence according to some embodiments of this specification. For example... Figure 6 As shown, the horizontal axis can represent frequency. The vertical axis can represent the second average value. Second average With frequency The changes can be as follows Figure 6 As shown.

[0050] In some embodiments, after determining the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency point in the noise time-frequency data, the amplitude value corresponding to each frequency point at each time point can be determined based on the second average value corresponding to each frequency point, and average background time-frequency data can be generated. For example, the frequency domain sequence pair generated based on the second average value can be copied in the time domain direction to generate average background time-frequency data. For instance, the generated frequency domain sequence pair can be, for example, (…). , ), ( , )……( , This allows the frequency domain sequence pair to be copied in the time domain, ensuring that the frequency and amplitude values ​​at each time point are identical to the original frequency domain sequence pair, thus generating average background time-frequency data. For example, at time point... The corresponding frequency and amplitude values ​​include ( , ), ( , )……( , ), time point The corresponding frequency and amplitude values ​​also include ( , ), ( , )……( , ), where m is any natural number greater than 1. Furthermore, average background time-frequency data can be generated based on each time point, the frequency point corresponding to each time point, and the amplitude value corresponding to each time point and frequency point. The average background time-frequency data can be represented by the average background time-frequency spectrum or by the matrix data corresponding to the average background time-frequency spectrum.

[0051] In some embodiments, the initial time-frequency data can be denoised based on the average background time-frequency data to obtain the target time-frequency data. For example, the initial time-frequency data can be the matrix data corresponding to the initial time-frequency spectrum, and the average background time-frequency data can be the matrix data corresponding to the average background time-frequency spectrum. The target time-frequency data can be obtained by subtracting the matrix data corresponding to the initial time-frequency spectrum from the matrix data corresponding to the average background time-frequency spectrum, thereby removing background noise from the initial time-frequency data. In some embodiments, the target time-frequency data can also be binarized based on a second preset threshold to further denoise the target time-frequency data and highlight the effective signal. For example, the target time-frequency data can be the target time-frequency spectrum. A second preset threshold can be set for the target time-frequency spectrum. If the brightness value corresponding to a pixel is greater than or equal to the second preset threshold, it is marked as "1" and represented in white in the target time-frequency spectrum. If the brightness value corresponding to a pixel is less than the second preset threshold, it is marked as "0" and represented in black in the target time-frequency spectrum. After binarization, the target time-frequency spectrum can be converted from a grayscale image to a black-and-white binary image. Figure 7 This is a schematic diagram of target time-frequency data for a regular image transmission signal according to some embodiments of this specification. For example, after denoising the initial time-frequency data of the regular image transmission signal, the obtained target time-frequency data of the regular image transmission signal can be as follows: Figure 7 As shown. Figure 8 This is a schematic diagram of target time-frequency data for a frequency-hopping signal according to some embodiments of this specification. For example, after denoising the initial time-frequency data of the frequency-hopping signal, the obtained target time-frequency data of the frequency-hopping signal can be as follows: Figure 8 As shown.

[0052] In some embodiments, morphological processing can also be performed on the target time-frequency data. This morphological processing can involve locally modifying the target time-frequency data using a sliding window. For example, morphological processing can include dilation and erosion. Dilation can fill holes or gaps in the target time-frequency data to enhance signal connectivity. Erosion can remove noise and glitches from the target time-frequency data, making signal boundaries clearer. Morphological processing can repair damaged signals in the target time-frequency data to reveal a complete signal.

[0053] Step 140: Extract signal features from the target time-frequency data and determine the corresponding UAV type based on the signal features. In some embodiments, step 140 can be implemented by the first type determination module 1040.

[0054] In some embodiments, the acquired radio frequency signals can be regular image transmission signals, irregular image transmission signals, or frequency hopping signals, and different discrimination methods can be used to determine the corresponding drone type for different radio frequency signals.

[0055] In some embodiments, the method for identifying regular image transmission signals can acquire image transmission signal features from target time-frequency data and match these features with signal features in a first preset feature library to determine the corresponding UAV type. The image transmission signal features may include periodic characteristics of the image transmission signal. The following describes in detail how to identify the UAV type corresponding to a regular image transmission signal.

[0056] In some embodiments, rectangular signal blocks (such as...) in the target time-frequency data can be identified first. Figure 7 The white rectangular area in the target time-frequency data can be used as an example. For instance, the rectangular signal block in the target time-frequency data can be identified using the rectangle detection module in OpenCV. Then, the identified rectangular signal blocks are filtered based on a preset bandwidth to select regular image transmission signals. For example, rectangular signal blocks with a bandwidth greater than 10MHz are identified as regular image transmission signals. Further, the image transmission signal characteristics in the target time-frequency data are obtained, where the image transmission signal characteristics can include the periodic characteristics of the regular image transmission signal. For example, the regular image transmission signal can be converted into a time sequence based on the time points of its appearance and disappearance, and used as the periodic characteristics of the regular image transmission signal. For example, if the time point of the regular image transmission signal's appearance is recorded as "01" and the time point of its disappearance is recorded as "00", then the periodic characteristics of the regular image transmission signal can be represented as 01 01 00 00 01 00 01 01 00 01.

[0057] In some embodiments, the first preset feature library can pre-store different drone types and their corresponding signal features. These signal features can be preset time-series sequences. By comparing the periodic features of the obtained regular image transmission signal with the preset time-series sequences in the first preset feature library for similarity determination, and comparing this similarity with a determination threshold, when the similarity is greater than the determination threshold, it is determined that the periodic features of the regular image transmission signal successfully match the preset time-series sequences in the first preset feature library, thereby determining the corresponding drone type. For example, if the preset signal feature corresponding to drone type A in the first preset feature library is 01 01 00 01 01 00 01 00 01, and the periodic feature of the regular image transmission signal is 01 01 00 00 01 0001 01 00 01, then the cosine distance between the two feature sequences can be calculated, resulting in a similarity of 90%. This similarity of 90% is greater than the determination threshold of 80%, indicating that the two feature sequences successfully match, thus determining that the drone type corresponding to the target time-frequency data is A.

[0058] In some embodiments, for the method of identifying irregular image transmission signals, it can first be determined whether there is a valid radio frequency signal within a preset frequency range. When a valid radio frequency signal exists within the preset frequency range for the target time-frequency data, the target time-frequency data within the preset frequency range can be compressed, and the compressed target time-frequency data can be matched with the time-frequency data in a second preset feature library to determine the corresponding UAV type. The following describes in detail how to identify the UAV type corresponding to irregular image transmission signals.

[0059] In some embodiments, the presence of a valid radio frequency (RF) signal within a preset frequency range can be determined based on the duty cycle of the signal within that preset frequency range. The preset frequency range can be the frequency range corresponding to an FM signal or a Wi-Fi enhancement signal. For example, the target time-frequency data within the preset frequency range can be extracted, and the duty cycle of the signal within that range can be calculated. For instance, the duty cycle can be determined by calculating the ratio of the area of ​​the signal region to the area of ​​the overall region within the preset frequency range. If the duty cycle is within a preset threshold range, a valid RF signal is determined to exist within the preset frequency range. Further, the resolution of the target time-frequency data within the preset frequency range is compressed, and then the compressed target time-frequency data within the preset frequency range is matched with time-frequency data in a second preset feature library. In some embodiments, the second preset feature library can pre-store different drone types and their corresponding preset time-frequency data. The preset time-frequency data can be a preset time-spectrum map corresponding to an FM signal or a Wi-Fi enhancement signal. For example, the template matching module in OpenCV can be used to match the compressed target time-frequency data within the preset frequency range with the preset time-frequency data in the second preset feature library. When a match is successful, the drone type corresponding to the preset time-frequency data is output.

[0060] In some embodiments, the method for identifying frequency-hopping signals can first extract frequency-hopping block data from the target time-frequency data, match the frequency of the frequency-hopping block data with a preset frequency set, and when a match is successful, obtain the frequency-hopping signal features of the frequency-hopping block data, and match the frequency-hopping signal features with signal features in a third preset feature library to determine the corresponding UAV type. The following describes in detail how to identify the UAV type corresponding to a frequency-hopping signal.

[0061] In some embodiments, frequency hopping block data in the target time-frequency data can be extracted first by rectangle detection. For example, a rectangle detection module in OpenCV can be used to identify rectangular signal blocks in the target time-frequency data. Then, the identified rectangular signal blocks are filtered based on preset dimensions (such as the length and width of the rectangle) to select frequency hopping block data, and the positions of the frequency hopping block data are marked in the target time-frequency data. Then, the frequency of the frequency hopping block data is matched with a preset frequency set. For example, assuming the frequencies of the frequency hopping block data are 2401MHz, 2403MHz, and 2405MHz, and the preset frequency set is 2401MHz, 2402MHz, 2403MHz, 2404MHz, 2405MHz, and 2406MHz, if the frequency of the frequency hopping block data is within the preset frequency set, then the match is successful. Further, the frequency hopping signal characteristics of the frequency hopping block data are obtained. The frequency hopping signal characteristics may include at least one of the following characteristics of the frequency hopping signal: dwell time, signal bandwidth, time domain interval, and frequency domain interval. The dwell time can be the length of time the frequency hopping signal continuously transmits at a single frequency point, for example... Figure 8 The horizontal length of a single rectangular block. Signal bandwidth can be the spectral width occupied by a frequency-hopping signal at a single point in time, for example... Figure 8 The height of a single rectangular block in the vertical direction. The time-domain interval can be the time interval between adjacent frequency-hopping blocks, for example... Figure 8 The horizontal spacing between adjacent rectangular blocks. Frequency domain spacing can be the frequency difference between adjacent frequency-hopping blocks, for example... Figure 8 The vertical spacing between adjacent rectangular blocks. In some embodiments, the third preset feature library may pre-store different UAV types and their corresponding signal features. These signal features may include a preset dwell time, a preset signal bandwidth, a preset time-domain interval, or a preset frequency-domain interval. Frequency-hopping signal features can be matched with signal features in the third preset feature library to determine the corresponding UAV type. For example, the dwell time of frequency-hopping block data can be matched with a preset dwell time in the third preset feature library; the signal bandwidth of frequency-hopping block data can be matched with a preset signal bandwidth in the third preset feature library; the time-domain interval of frequency-hopping block data can be matched with a preset time-domain interval in the third preset feature library; and the frequency-domain interval of frequency-hopping block data can be matched with a preset frequency-domain interval in the third preset feature library. When any one or more of these matches successfully, or when all matches successfully, the corresponding UAV type is output.

[0062] In some embodiments, the communication protocol of some drones also carries identification information, and the drone type can be determined by identifying the identification information in the radio frequency signal, thereby improving the accuracy of drone type identification. Figure 9 This is an exemplary flowchart illustrating another drone identification method according to some embodiments of this specification. Figure 9 The illustrated process 900 can be executed by a terminal device, such as an electronic detection device. In some embodiments, process 900 can be implemented by a drone identification device 1000 deployed on a processing device or a terminal device. Figure 9 As shown, in some embodiments, process 900 may include the following steps.

[0063] Step 910: Acquire radio frequency signal. In some embodiments, step 910 may be implemented by acquisition module 1010.

[0064] This step is similar to the previous one. For a detailed explanation of this step, please refer to the description in step 110 above. It will not be repeated here.

[0065] Step 920: Extract the identification information from the radio frequency signal and match it with the information in the preset information database. When a match is successful, determine the corresponding drone type. In some embodiments, step 920 can be implemented by the second type determination module 1050.

[0066] In some embodiments, the identification information in the radio frequency signal may include a unique hardware identifier for the drone or an identifier actively reported by the drone during flight. For example, the identification information may include a Drone ID or a Remote ID. The Drone ID may be a unique hardware identifier embedded at the drone's factory, used to identify the drone hardware itself. For example, the Drone ID may be the serial number or SN code of a DJI drone, which the DJI drone can actively broadcast via radio frequency signals during flight. The Remote ID may be a remote identification code actively broadcast by the drone during flight. For example, the Remote ID may be a remote identification code that international and national standards require drones to actively broadcast to verify the legitimacy of the drone's identity. For example, the Remote ID may contain information such as an operator registration code, manufacturer code, and product category code. The identification methods for these two types of identification information are described below.

[0067] In some embodiments, for Drone ID identification, the received radio frequency signal can be preprocessed, decoded to obtain drone ID protocol information, and the corresponding drone type can be determined based on the device_type field in the drone ID protocol information. For example, the acquired radio frequency signal can be digitized and then input to a deinterleaver to deinterleave the signal, eliminating interleaving interference during transmission and restoring the original signal order. The deinterleaver can use a specific permutation algorithm to reorder the scattered bit sequence at the receiving end, restoring the original order and ensuring signal integrity. Further, the deinterleaved signal can be soft-output decoded. For example, the deinterleaved signal can be used as soft input and input to two parallel decoders, each decoding the signal. The decoded signal can then be further input to two convolutionals for smoothing, reducing noise and improving the signal-to-noise ratio. Furthermore, the convolutionally processed signal undergoes iterative decoding. This iterative decoding process, through multiple iterations, continuously optimizes the decoding result, improving the accuracy and reliability of the decoding, thereby resolving the drone ID protocol information. The drone ID protocol information may include a device_type field. In some embodiments, a preset device_type field corresponding to different drone types can be pre-stored in a preset drone type database. By comparing the device_type field with the preset device_type field in the preset drone type database, the corresponding drone type can be determined.

[0068] In some embodiments, for Remote ID identification, the received radio frequency signal can be preprocessed, segmented, and parsed to obtain basic ID message information. The corresponding drone type can then be determined based on the UAS ID field in the basic ID message information. For example, the acquired radio frequency signal can be digitized and then filtered, denoised, and enhanced to remove noise and interference components, improving the signal-to-noise ratio. Further, the preprocessed signal is segmented and parsed according to the Remote ID protocol format. The Remote ID protocol may include basic ID messages, position vector messages, operational description messages, system messages, etc. After segmentation and parsing, basic ID message information can be obtained, which may include a UAS ID field. In some embodiments, a preset UAS ID field corresponding to different drone types can be pre-stored in a preset drone type database. By comparing the UAS ID field with the preset UAS ID field in the preset drone type database, the corresponding drone type can be determined.

[0069] In some embodiments, after step 920, step 120 in process 100 may continue to be executed. In other embodiments, step 920 may also be executed in parallel with step 120.

[0070] By extracting identification information from radio frequency signals and matching it with information in a pre-set database, the corresponding drone type can be determined, thus providing broader coverage for drone type identification and improving the accuracy of drone identification.

[0071] This manual also provides a drone identification device. Figure 10 This is an exemplary block diagram of a drone identification device according to some embodiments of this specification. Figure 10 As shown, in some embodiments, the drone identification device 1000 may include an acquisition module 1010, a preprocessing module 1020, a noise reduction processing module 1030, and a first type determination module 1040.

[0072] Acquisition module 1010 is used to acquire radio frequency signals.

[0073] The preprocessing module 1020 is used to preprocess the radio frequency signal to obtain initial time-frequency data.

[0074] The noise reduction module 1030 is used to perform noise reduction processing on the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain the target time-frequency data.

[0075] The first type determination module 1040 is used to extract signal features from the target time-frequency data and determine the corresponding UAV type based on the signal features.

[0076] In some optional embodiments, the acquisition module 1010 can also be used to acquire set target acquisition parameters, including at least one of target center frequency, target bandwidth and target sampling rate; and acquire radio frequency signals acquired by the detection device based on the target acquisition parameters.

[0077] In some optional embodiments, the denoising module 1030 may also be used to: extract noise time-frequency data corresponding to the noise period in the initial time-frequency data based on the first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data; generate average background time-frequency data based on the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency point in the noise time-frequency data; and perform denoising processing on the initial time-frequency data based on the average background time-frequency data to obtain target time-frequency data; wherein the amplitude value is used to characterize the signal strength of the radio frequency signal at a specific time point and a specific frequency point.

[0078] In some optional embodiments, the noise reduction processing module 1030 can also be used to determine the first average value of each amplitude value corresponding to different frequency points of the radio frequency signal at each time point in the initial time-frequency data; filter each first average value based on a first preset threshold and the magnitude of each first average value; determine the time point corresponding to the filtered first average value as the noise period; and extract the noise time-frequency data corresponding to the noise period in the initial time-frequency data.

[0079] In some optional embodiments, the noise reduction processing module 1030 can also be used to determine the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency point in the noise time-frequency data; determine the amplitude value corresponding to each frequency point at each time point based on the second average value corresponding to each frequency point, and generate average background time-frequency data.

[0080] In some optional embodiments, the first type determination module 1040 can also be used to acquire image transmission signal features in the target time-frequency data, the image transmission signal features including the periodicity features of the image transmission signal; match the image transmission signal features with signal features in a first preset feature library to determine the corresponding UAV type.

[0081] In some optional embodiments, the first type determination module 1040 can also be used to compress the target time-frequency data within the preset frequency range when there is a valid radio frequency signal in the target time-frequency data within the preset frequency range; and match the compressed target time-frequency data with the time-frequency data in the second preset feature library to determine the corresponding UAV type.

[0082] In some optional embodiments, the first type determination module 1040 can also be used to: extract frequency hopping block data from the target time-frequency data; match the frequency of the frequency hopping block data with a preset frequency set; when the match is successful, obtain the frequency hopping signal features of the frequency hopping block data, the frequency hopping signal features including at least one of the following features of the frequency hopping signal: dwell time, signal bandwidth, time domain interval, frequency domain interval; match the frequency hopping signal features with signal features in a third preset feature library to determine the corresponding UAV type.

[0083] In some optional embodiments, the drone identification device 1000 may further include a second type determination module 1050, which is used to extract the identification information in the radio frequency signal after acquiring the radio frequency signal, and match it with the information in the preset information database. When the match is successful, the corresponding drone type is determined.

[0084] For more information on each module, please refer to [link / reference]. Figures 1-9 The relevant explanations will not be repeated here. It should be understood that... Figure 10The apparatus and modules shown can be implemented in various ways. For example, in some embodiments, the apparatus and modules can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by an appropriate instruction execution device, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the methods and apparatus described above can be implemented using computer-executable instructions and / or included in the control code of a processor, such as in the memory of a disk, CD, or DVD-ROM. The apparatus and modules described in this specification can be implemented not only by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips or transistors, or programmable hardware devices such as field-programmable gate arrays or programmable logic devices, but also by software, for example, executed by various types of processors, or by a combination of the aforementioned hardware circuitry and software (e.g., firmware).

[0085] It should be noted that the above description of the device and its modules is for convenience only and should not be construed as limiting this specification to the embodiments described. It is understood that those skilled in the art, after understanding the principle of the device, may arbitrarily combine the various modules without departing from this principle to form sub-devices connected to other modules. Alternatively, some modules may be split to obtain more modules or multiple units under a single module. Such modifications are all within the scope of this specification.

[0086] Some embodiments of this specification also provide an electronic device, including one or more processors; and a memory for storing one or more programs, which, when executed by the one or more processors, enable the one or more processors to implement this specification. Figures 1-9 The method shown.

[0087] Some embodiments of this specification also provide a computer program product, including computer instructions that, when at least a portion of the computer instructions are executed by a processor, can implement this specification. Figures 1-9 The method is illustrated. In some embodiments, the computer program product may relate only to computer instructions, which may be carried on a storage medium or processing device. In other embodiments, the computer program product may also be a storage medium or processing device containing the aforementioned computer instructions. The processing device may include one or more processors, and the storage medium.

[0088] In some embodiments, the processor may be a combination of one or more of the following processors: central processing unit (CPU), application-specific integrated circuit (ASIC), application-specific instruction set processor (ASIP), graphics processing unit (GPU), physical processing unit (PPU), digital signal processor (DSP), field-programmable gate array (FPGA), programmable logic device (PLD), programmable logic controller (PLC), reduced instruction set computer (RISC), and microprocessor.

[0089] In some embodiments, the storage medium may include one or more combinations of the following: mass storage, removable storage, volatile read-write memory, and read-only memory (ROM). Exemplary mass storage may include disks, optical disks, solid-state drives, etc. Exemplary removable storage may include flash drives, floppy disks, optical disks, memory cards, compressed hard disks, magnetic tapes, etc. Exemplary volatile read-write memory may include random access memory (RAM). Exemplary RAM may include dynamic random access memory (DRAM), dual data rate synchronous dynamic random access memory (DDRSDRAM), static random access memory (SRAM), silicon controlled retrieval memory (T-RAM), and zero-capacitance memory (Z-RAM), etc. Exemplary read-only memory may include masked read-only memory (MROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), compressed hard disk read-only memory (CD-ROM), and digital multifunction hard disk read-only memory, etc.

[0090] The beneficial effects that the embodiments of this specification may bring include, but are not limited to: (1) obtaining initial time-frequency data by preprocessing the radio frequency signals collected by the detection device, obtaining target time-frequency data by denoising the initial time-frequency data based on the average amplitude of the radio frequency signals in the initial time-frequency data, extracting the signal features in the target time-frequency data, and determining the corresponding UAV type based on the signal features, thereby effectively removing background noise and improving the accuracy of UAV type identification; (2) determining the corresponding UAV type by extracting the identification information in the radio frequency signal and matching it with the information in the preset information database, thereby covering the identification of UAV types more broadly and improving the accuracy of UAV identification; (3) expanding the range of UAV type identification by determining the UAV type of irregular image transmission signals, and covering the identification of multiple UAV types more broadly; (4) determining the UAV type of frequency hopping signal based on the frequency hopping signal features, the frequency hopping signal features including at least one of dwell time, signal bandwidth, time domain interval, and frequency domain interval, thereby improving the accuracy of UAV type identification based on frequency hopping signals. It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects may be any one or a combination of the above, or any other possible beneficial effects.

[0091] The basic concepts have been described above. It is obvious that the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, various modifications, improvements, and corrections may be made to this specification by those skilled in the art. Such modifications, improvements, and corrections are taught in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

Claims

1. A method for identifying unmanned aerial vehicles (UAVs), characterized in that, The method includes: Acquire radio frequency signals; Preprocessing is performed on the radio frequency signal to obtain initial time-frequency data; The initial time-frequency data is denoised based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain the target time-frequency data. Extract signal features from the target time-frequency data, and determine the corresponding UAV type based on the signal features; The step of denoising the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain the target time-frequency data includes: Based on the first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data, noise time-frequency data corresponding to the noise period is extracted from the initial time-frequency data. Average background time-frequency data is generated based on the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency in the noise time-frequency data. The initial time-frequency data is denoised based on the average background time-frequency data to obtain the target time-frequency data. The amplitude value is used to characterize the signal strength of the radio frequency signal at a specific time point and a specific frequency point.

2. The method according to claim 1, characterized in that, The step of extracting noise time-frequency data corresponding to the noise period from the initial time-frequency data based on the first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data includes: Determine the first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data; Based on a first preset threshold and the magnitude of each first average value, the first average values ​​are filtered, and the time point corresponding to the filtered first average value is determined as the noise period. The noise time-frequency data corresponding to the noise period is extracted from the initial time-frequency data.

3. The method according to claim 1, characterized in that, The step of generating average background time-frequency data based on the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency in the noise time-frequency data includes: Determine the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency point in the noise time-frequency data; Based on the second average value corresponding to each frequency point, the amplitude value corresponding to each frequency point at each time point is determined, and the average background time-frequency data is generated.

4. The method according to claim 1, characterized in that, The step of extracting signal features from the target time-frequency data and determining the corresponding UAV type based on the signal features includes: The image transmission signal features in the target time-frequency data are obtained, and the image transmission signal features include the periodicity features of the image transmission signal; The image transmission signal features are matched with signal features in the first preset feature library to determine the corresponding UAV type.

5. The method according to claim 1, characterized in that, The step of extracting signal features from the target time-frequency data and determining the corresponding UAV type based on the signal features includes: When the target time-frequency data contains a valid radio frequency signal within a preset frequency range, the target time-frequency data within the preset frequency range is compressed. The compressed target time-frequency data is matched with the time-frequency data in the second preset feature library to determine the corresponding UAV type.

6. The method according to claim 1, characterized in that, The step of extracting signal features from the target time-frequency data and determining the corresponding UAV type based on the signal features includes: Extract the frequency hopping block data from the target time-frequency data; The frequency of the frequency hopping block data is matched with a preset frequency set; When a match is successful, the frequency hopping signal characteristics of the frequency hopping block data are obtained. The frequency hopping signal characteristics include at least one of the following characteristics of the frequency hopping signal: dwell time, signal bandwidth, time domain interval, and frequency domain interval. The frequency hopping signal features are matched with signal features in a third preset feature library to determine the corresponding UAV type.

7. The method according to claim 1, characterized in that, The acquisition of radio frequency signals includes: Obtain the set target acquisition parameters, wherein the target acquisition parameters include at least one of the target center frequency, target bandwidth, and target sampling rate; The radio frequency signal acquired by the detection device is obtained based on the target acquisition parameters.

8. A drone identification device, characterized in that, The device includes: The acquisition module is used to acquire radio frequency signals; The preprocessing module is used to preprocess the radio frequency signal to obtain initial time-frequency data; The noise reduction module is used to perform noise reduction processing on the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain the target time-frequency data; The type determination module is used to extract signal features from the target time-frequency data and determine the corresponding UAV type based on the signal features; The step of denoising the initial time-frequency data based on the average amplitude of the radio frequency signal in the initial time-frequency data to obtain the target time-frequency data includes: Based on the first average value of the amplitude values ​​of the radio frequency signal at different frequency points at each time point in the initial time-frequency data, noise time-frequency data corresponding to the noise period is extracted from the initial time-frequency data. Average background time-frequency data is generated based on the second average value of the amplitude values ​​of the radio frequency signal at different time points at each frequency in the noise time-frequency data. The initial time-frequency data is denoised based on the average background time-frequency data to obtain the target time-frequency data. The amplitude value is used to characterize the signal strength of the radio frequency signal at a specific time point and a specific frequency point.

9. An electronic device, characterized in that, include: One or more processors; A memory for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, It includes a computer program that, when at least a portion of the computer program is executed by a processor, enables the implementation of the method as described in any one of claims 1 to 7.

Images