Method and system for detecting micro unmanned aerial vehicle based on target reflectivity
By combining a target reflectivity-based detection method with active laser detection and target reflectivity feature analysis, the problem of identifying micro-UAVs in complex environments has been solved, achieving high-accuracy detection of micro-UAVs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIDIAN UNIV
- Filing Date
- 2025-09-19
- Publication Date
- 2026-06-09
AI Technical Summary
Existing technologies suffer from low signal-to-noise ratio, significant environmental interference, and low recognition accuracy when detecting micro-drones, especially in complex environments where they are difficult to effectively identify.
A target reflectivity-based detection method is adopted. By emitting pulsed lasers, the target signal is synchronously acquired by dual detectors with a common optical path design. Combined with TOF ranging technology and atmospheric attenuation model, the reflectivity value is calculated and corrected. Target motion features are generated through multi-frame time series analysis, a comprehensive feature vector is constructed, and finally, a machine learning model is used for identification.
It improves the recognition accuracy and robustness of micro-drones in complex environments, enhances the ability to identify low-observable targets, and reduces the false alarm rate.
Smart Images
Figure CN121069350B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of micro-drone detection and identification technology, and in particular to a micro-drone detection method and system based on target reflectivity. Background Technology
[0002] With the rapid development of drone technology, especially the widespread application of consumer and industrial micro-drones, their activities in the low-altitude domain are becoming increasingly frequent. These targets typically have characteristics such as low radar cross-section (RCS), small size, low flight altitude, and slow speed, posing a severe challenge to existing detection technologies. Traditional radar detection systems mainly rely on the intensity of the target echo signal and Doppler frequency shift for detection and tracking. However, due to the small physical size of micro-drones and their use of non-metallic materials, their echo signals are extremely weak and easily affected by internal system noise and external environmental clutter, leading to a significant decrease in detection probability and an increase in false alarm rate.
[0003] Furthermore, video target recognition methods based on optical or infrared imaging have been widely introduced into the field of UAV detection in recent years. These methods mainly identify targets by analyzing their geometric shape, motion patterns, or trajectory characteristics, thus compensating to some extent for the shortcomings of radar in detecting small targets at low altitudes. However, existing video recognition algorithms do not fully utilize the electromagnetic reflection characteristics of the target itself and fail to adequately combine its reflection response characteristics at different frequency bands. Therefore, the recognition accuracy remains unsatisfactory when the target has low differentiation from the background or when there are similar-looking interfering objects (such as birds, fallen leaves, etc.).
[0004] On the other hand, complex environmental conditions (such as rain, fog, smoke, and haze) further exacerbate the difficulty of detection. These weather conditions not only cause shielding and scattering effects on optical imaging systems, reducing image quality, but also attenuate the propagation of radar waves, especially for high-band radars such as Ka and W, making it even more difficult to effectively extract the already weak UAV echoes.
[0005] In summary, existing detection methods have significant limitations when dealing with micro-UAVs with low observability: radar systems are constrained by low signal-to-noise ratios and environmental interference, while video recognition methods suffer from insufficient recognition capabilities due to a lack of effective fusion of reflection characteristics. Therefore, there is an urgent need to develop a novel detection method that can comprehensively utilize target reflection characteristics and multimodal sensor data to improve the reliable identification and stable tracking capabilities of micro-UAVs in complex environments. Summary of the Invention
[0006] This invention provides a method and system for detecting micro-UAVs based on target reflectivity, which solves the problem of low-reflectivity targets being difficult to detect in complex environments in the prior art. It enables the identification of micro-UAVs by combining active laser detection with target reflectivity feature analysis.
[0007] In a first aspect, the present invention provides a method for detecting small unmanned aerial vehicles (UAVs) based on target reflectivity, the method comprising:
[0008] A pulsed laser is emitted toward the target area, and the beam is expanded and collimated to cover the detection area.
[0009] Based on the pulsed laser, the target signal is synchronously acquired by dual detectors with a common optical path design to obtain the target's geometric features and reflectivity intensity signal;
[0010] The original reflectivity is calculated based on the reflectivity intensity signal, and the target distance information is obtained. The original reflectivity is then corrected based on the target distance information to obtain a corrected reflectivity value.
[0011] Target motion features are generated through multi-frame temporal analysis, and based on a fusion strategy, the corrected reflectivity value, the target geometric features, and the target motion features are fused to construct a comprehensive target feature vector;
[0012] The target integrated feature vector is input into the classifier, and the recognition result of the micro-drone is output.
[0013] In conjunction with the first aspect, in one possible implementation, the step of synchronously acquiring target signals using dual detectors with a common optical path design based on the pulsed laser to obtain target geometric features and reflectivity intensity signals includes:
[0014] The dual detectors include an imaging detector and a non-imaging detector;
[0015] The imaging detector is a CMOS detector, used to acquire the geometric features of the target;
[0016] The non-imaging detector is an APD avalanche photodiode array, used to acquire the reflectivity intensity signal.
[0017] In conjunction with the first aspect, in one possible implementation, the field-of-view overlap error between the imaging detector and the non-imaging detector is less than or equal to 0.1°, and the time synchronization error is less than or equal to 1μs.
[0018] In conjunction with the first aspect, in one possible implementation, the formula for calculating the original reflectivity is:
[0019] ;
[0020] in, This represents the peak value of the echo intensity in the reflectivity intensity signal. This represents the peak value of the emitted light intensity in the reflectivity intensity signal; This represents the original reflectance.
[0021] In conjunction with the first aspect, in one possible implementation, the step of acquiring target distance information and correcting the original reflectivity based on the target distance information to obtain a corrected reflectivity value includes:
[0022] The distance information of the target is obtained using Time-of-Flight (TOF) ranging technology. ;
[0023] An atmospheric attenuation model is used, based on the distance information. The calculated correction value is used to correct the original reflectance to eliminate the influence of atmospheric attenuation, resulting in a corrected reflectance value.
[0024] In conjunction with the first aspect, in one possible implementation, generating target motion features through multi-frame temporal analysis includes:
[0025] Based on continuous multi-frame data from the imaging detector, the position coordinates of the target at adjacent time points are obtained;
[0026] Calculate the instantaneous velocity and acceleration based on the time difference of the position coordinates;
[0027] Calculate the trajectory direction angle based on the spatial relationship of the position coordinates;
[0028] The position coordinates, instantaneous velocity, acceleration, and trajectory direction angle at adjacent moments are used as the motion features of the target.
[0029] In conjunction with the first aspect, in one possible implementation, the fusion strategy includes:
[0030] The weights of the target geometric features in the fusion process are dynamically adjusted based on the magnitude of the corrected reflectance value. This dynamic adjustment includes: increasing the weights of the target geometric features when the corrected reflectance value is below 5%; and decreasing the weights of the target geometric features when the corrected reflectance value is above 30%.
[0031] Construct a reflectance-target geometric feature correlation factor; wherein, the reflectance-target geometric feature correlation factor includes: a reflectance-size correlation factor and a reflectance-shape correlation factor; the reflectance-target geometric feature correlation factor is expressed as:
[0032] ;
[0033] The reflectivity-shape correlation factor is expressed as:
[0034] ;
[0035] in, Indicates the area of the light spot; Indicates roundness; This indicates the correction value.
[0036] In conjunction with the first aspect, in one possible implementation, the fusion of the corrected reflectivity value, the target geometric features, and the target motion features to construct a comprehensive target feature vector includes:
[0037] Multiple feature parameters are extracted based on the corrected reflectivity value, the target geometric features, and the target motion features;
[0038] Multiple parameters are normalized separately, and the normalized feature parameters are concatenated in a fixed order to obtain the target comprehensive feature vector.
[0039] In conjunction with the first aspect, in one possible implementation, the classifier is a machine learning model-based classifier, including at least one of support vector machines and convolutional neural networks.
[0040] Secondly, the present invention provides a micro-UAV detection system based on target reflectivity, the system comprising: a laser emitting module for emitting pulsed laser to a target area, which is then expanded and collimated to cover the detection area;
[0041] The dual-detector module is used to synchronously acquire target signals based on the pulsed laser through dual detectors designed with a common optical path, thereby obtaining target geometric features and reflectivity intensity signals.
[0042] The signal processing module is used to calculate the original reflectivity based on the reflectivity intensity signal, obtain target distance information, and correct the original reflectivity based on the target distance information to obtain a corrected reflectivity value.
[0043] The feature fusion module is used to generate target motion features through multi-frame temporal analysis, and based on the fusion strategy, fuse the corrected reflectivity value, the target geometric features and the target motion features to construct a comprehensive target feature vector;
[0044] The classification and recognition module is used to input the comprehensive feature vector of the target into the classifier and output the recognition result of the micro-drone.
[0045] One or more technical solutions provided in this invention have at least the following technical effects or advantages:
[0046] This invention employs a pulsed laser emitted towards the target area. After beam expansion and collimation, the laser uniformly covers the entire detection area, effectively increasing the illumination range and improving the signal-to-noise ratio. Based on this pulsed laser, a dual-detector design with a common optical path synchronously acquires target signals, simultaneously obtaining target geometric features and reflectivity intensity signals, and calculates the original reflectivity accordingly. This common optical path design ensures complete spatiotemporal correspondence between the two types of signals, reducing calibration errors. Target distance information is acquired and used to correct the original reflectivity, resulting in a more accurate corrected reflectivity value. This eliminates the influence of distance on reflection intensity, improving the comparability and reliability of reflection features. Target motion features are generated through multi-frame temporal analysis, and based on a fusion strategy, the corrected reflectivity value, target geometric features, and generated motion features are fused to construct a comprehensive feature vector that fully describes the target attributes, enhancing feature representation capabilities. This comprehensive feature vector is input into a classifier, and the recognition result of the micro-UAV is output based on the multi-dimensional fused features, improving the accuracy and robustness of identifying low-observable targets in complex environments. Attached Figure Description
[0047] Figure 1 A flowchart illustrating the steps of a micro-UAV detection method based on target reflectivity provided in an embodiment of the present invention;
[0048] Figure 2 A flowchart illustrating the processing of the target comprehensive feature vector provided in this embodiment of the invention;
[0049] Figure 3 This is a schematic diagram of a micro-UAV detection system based on target reflectivity provided in an embodiment of the present invention. Detailed Implementation
[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0051] This invention provides a method for detecting small unmanned aerial vehicles (UAVs) based on target reflectivity. (See also...) Figure 1 The method includes the following steps S101 to S104.
[0052] S101 emits pulsed laser light toward the target area, which is then expanded and collimated to cover the detection area.
[0053] For example, the wavelength of a pulsed laser can be selected from near-infrared or long-wave infrared.
[0054] S102, based on pulsed laser, uses dual detectors with a common optical path design to synchronously acquire target signals, thereby obtaining target geometric features and reflectivity intensity signals;
[0055] Specifically, in step S102, based on pulsed laser, the target signal is synchronously acquired by dual detectors with a common optical path design to obtain the target geometric features and reflectivity intensity signal. The dual detectors include an imaging detector and a non-imaging detector. The imaging detector is a CMOS detector used to acquire the target geometric features. The non-imaging detector is an APD avalanche photodiode array used to acquire the reflectivity intensity signal.
[0056] Here, the target's geometric features include: spot area (target size), circularity (target shape regularity), and aspect ratio (target contour features).
[0057] Here, the field-of-view overlap error between the imaging detector and the non-imaging detector is less than or equal to 0.1°, and the time synchronization error is less than or equal to 1μs.
[0058] S103, calculate the original reflectivity based on the reflectivity intensity signal, obtain the target distance information, and correct the original reflectivity based on the target distance information to obtain the corrected reflectivity value;
[0059] Specifically, in step S103, target distance information is obtained, and the original reflectivity is corrected based on the target distance information to obtain a corrected reflectivity value, including the following steps S1031 to S1032.
[0060] S1031, obtains the target's distance information through TOF ranging technology. ;
[0061] S1032 employs an atmospheric attenuation model, based on distance information. The calculated correction value is used to correct the original reflectance to eliminate the effect of atmospheric attenuation, resulting in the corrected reflectance value.
[0062] Here, the formula for calculating the original reflectance is:
[0063] ;
[0064] in, This represents the peak value of the echo intensity in the reflectivity intensity signal. This represents the peak value of the emitted light intensity in the reflectivity intensity signal; This represents the original reflectance.
[0065] S104 generates target motion features through multi-frame temporal analysis, and based on a fusion strategy, fuses the corrected reflectivity value, target geometric features, and target motion features to construct a comprehensive target feature vector;
[0066] Specifically, in step S104, target motion features are generated through multi-frame temporal analysis, including the following steps S1041 to S1044.
[0067] S1041, based on continuous multi-frame data from the imaging detector, obtain the target's position coordinates at adjacent time points;
[0068] S1042, calculate instantaneous velocity and acceleration based on the time difference of position coordinates;
[0069] S1043, Calculate the trajectory direction angle based on the spatial relationship of the position coordinates;
[0070] S1044 uses the position coordinates, instantaneous velocity, acceleration, and trajectory direction angle of adjacent time points as the target motion characteristics.
[0071] Here, the fusion strategy includes:
[0072] (1) Adjust the weight of the target geometric features in the fusion dynamically according to the magnitude of the corrected reflectance value; wherein, the dynamic adjustment includes: increasing the weight of the target geometric features when the corrected reflectance value is lower than 5%; and decreasing the weight of the target geometric features when the corrected reflectance value is higher than 30%.
[0073] (2) Construct a reflectance-target geometric feature correlation factor; wherein, the reflectance-target geometric feature correlation factor includes: a reflectance-size correlation factor and a reflectance-shape correlation factor; the reflectance-target geometric feature correlation factor is expressed as:
[0074] ;
[0075] The reflectance-shape correlation factor is expressed as:
[0076] ;
[0077] in, Indicates the area of the light spot; Indicates roundness; This indicates the correction value.
[0078] Specifically, in step S104, the corrected reflectivity value, target geometric features, and target motion features are fused to construct a comprehensive target feature vector, including:
[0079] (1) Extract multiple feature parameters based on the corrected reflectivity value, target geometric features, and target motion features;
[0080] (2) Normalize the multiple parameters respectively, and then concatenate the normalized feature parameters in a fixed order to obtain the target comprehensive feature vector.
[0081] See Figure 2 This is a flowchart of the process for processing the target comprehensive feature vector.
[0082] For example, reflectivity features: taking the corrected true reflectivity Rcorrected (single value, range 0~100%), and the rate of change of reflectivity over 3 consecutive frames, expressed by the formula: This reflects the stability of the target material.
[0083] The geometric features of the target include: target size and shape features.
[0084] Target size: Spot area S (number of pixels), equivalent diameter The diameter of the light spot is approximated as a circle; shape characteristics: circularity. ;in, The circumference of the light spot is represented by a value between 0 and 1, with values closer to 1 indicating a circular shape; the aspect ratio of the circumscribed rectangle is also considered. Width / height, distinguishing between elongated and square targets.
[0085] Motion characteristics: instantaneous velocity, acceleration, and trajectory direction angle. Instantaneous velocity is expressed as:
[0086] ;
[0087] in, The coordinates of the target center;
[0088] Acceleration is expressed as: This reflects whether the motion is uniform;
[0089] The trajectory direction angle is expressed as: , and the angle between it and the horizontal direction.
[0090] The above 10 characteristic parameters are normalized to eliminate dimensional differences:
[0091] ;
[0092] in, Represents the original eigenvalues; This represents the standardized value.
[0093] The 10 standardized features are concatenated in a fixed order to form a 10-dimensional target comprehensive feature vector, which is represented as follows:
[0094] ;
[0095] S105 inputs the target's comprehensive feature vector into the classifier and outputs the recognition result of the micro-drone.
[0096] Specifically, in step S105, the classifier is a machine learning model-based classifier, including at least one of support vector machines and convolutional neural networks.
[0097] For example, a dataset can be built based on machine learning models such as Vector Machine (SVM) and Convolutional Neural Network (CNN) to distinguish target types. Introducing target reflectivity parameters corresponding to the laser band into traditional algorithms based on target motion trajectory and geometry can significantly improve the differentiation of different types of targets, such as stealth devices, birds, and drones, in environments with weak target recognition capabilities.
[0098] Secondly, the present invention provides a micro-UAV detection system based on target reflectivity, see [link to relevant documentation]. Figure 3 The system includes:
[0099] The laser emitting module is used to emit pulsed lasers towards the target area, which are then expanded and collimated to cover the detection area.
[0100] The dual-detector module is used to synchronously acquire target signals based on pulsed lasers through a common optical path design of two detectors, thereby obtaining the target's geometric features and reflectivity intensity signals.
[0101] The signal processing module is used to calculate the original reflectivity based on the reflectivity intensity signal, obtain the target distance information, and correct the original reflectivity based on the target distance information to obtain the corrected reflectivity value.
[0102] The feature fusion module is used to generate target motion features through multi-frame temporal analysis, and based on the fusion strategy, fuses the corrected reflectivity value, target geometric features and target motion features to construct a comprehensive target feature vector;
[0103] The classification and recognition module is used to input the comprehensive feature vector of the target into the classifier and output the recognition result of the micro-drone.
[0104] This invention combines the target's optical reflectivity with its geometric features, achieving deep fusion through correlation modeling of physical properties and morphological features. Specifically, this is reflected in the following two aspects:
[0105] 1. Feature Filtering Layer: Dynamically weighted target geometric features based on reflectivity.
[0106] For low reflectivity targets (R<5%), their geometric features (such as spot area and shape) are easily affected by noise. Therefore, the weight of the target geometric features is increased in the feature vector to compensate for the weakness of the reflectivity signal.
[0107] For high reflectivity targets (R>30%), reflectivity features are highly stable, so the weight of the target's geometric features should be reduced to avoid interference from redundant information.
[0108] In everyday life, the core distinction between high and low optical reflectivity targets lies in the object's ability to reflect visible light (or light of a specific wavelength). Reflectivity is usually expressed as a percentage; high reflectivity is generally >50%, and low reflectivity is generally <20%. The two differ significantly in terms of material, appearance, application scenarios, and optical properties.
[0109] The core characteristics of high optical reflectivity items are: visual characteristics: under natural light or artificial light, the surface of the object is bright and easily forms a "highlight area", and some specular reflections will clearly image the target (such as a mirror can reflect a person's image).
[0110] Optical properties: The reflectivity is usually between 50% and 100% (such as the reflectivity of mirrored glass, which is about 80% to 90%, and white latex paint, which is about 70% to 85%), and it absorbs and transmits very little light (except for transparent high-reflectivity materials such as some optical glass).
[0111] Common characteristics of materials: They are mostly smooth surfaces (specular reflection) or high whiteness materials (diffuse reflection). The materials themselves have a low absorption coefficient for visible light (e.g., free electrons in metals easily reflect light, while particles in white pigments easily scatter light).
[0112] The core characteristics of low optical reflectivity items are: visual characteristics: the surface color is mostly dark black, dark brown, or dark gray, and there is no obvious reflection after light shines on them. They appear "dark" and can even absorb stray light from the surrounding environment (such as light-absorbing cloth in a photography studio, which can eliminate background reflection).
[0113] Optical characteristics: The reflectivity is usually between 2% and 20% (e.g., pure black paint has a reflectivity of about 2% to 5%, and frosted black plastic has a reflectivity of about 10% to 15%). Most of the light is absorbed by the object (converted into heat or other energy), and the transmittance is extremely low (except for black transparent materials such as sunglasses lenses).
[0114] Common material characteristics: They are mostly dark pigment coatings, rough surface materials or highly absorbent materials (such as graphite, whose carbon atom structure easily absorbs visible light). The surface is not smooth and mirror-like, and light is either weakly diffused or directly absorbed after being irradiated.
[0115] 2. Feature Interaction Layer: Constructing a reflectance-target geometric feature correlation factor. Two new cross features are added to quantify the intrinsic correlation between reflectance and target geometric features: a reflectance-size correlation factor (for the same size, drones typically have higher reflectance than birds (drones are mostly made of metal / plastic, while birds are made of feathers), and this factor can amplify this difference); and a reflectance-shape correlation factor (drones have more regular shapes (larger C, smaller 1-C), while birds have irregular shapes due to wing flapping (smaller C, larger 1-C), and combining reflectance can distinguish between the two).
[0116] In this way, reflectivity (reflecting the material nature of the target) and geometric features (reflecting the shape of the target) complement each other: reflectivity can distinguish between targets with the same shape but different materials (such as plastic drones and bionic birds), while geometric features can distinguish between targets with the same material but different shapes (such as multi-rotor drones and fixed-wing drones). The combination of the two improves the recognition accuracy of micro drones (compared to a single feature).
[0117] In a specific embodiment provided by the present invention, 1. Hardware deployment: Laser (wavelength 1550nm, pulse energy 10mJ, pulse width 10ns). Dual detector common optical path system (imaging detector: 1920×1080 CMOS; non-imaging detector: APD avalanche photodiode array).
[0118] 2. Software Processing: After sampling by the ADC, the echo signal undergoes noise reduction through adaptive filtering (Wiener filtering). A target classifier trained based on the YOLOv9 model achieves a recognition accuracy of over 95%. Dynamic Adjustment: When the target reflectivity is below 1%, it automatically switches to a long pulse mode (50ns pulse width) to improve the signal-to-noise ratio.
[0119] Reflectivity + Feature Fusion Mechanism: For the first time, target reflectivity is combined with target geometric features to improve the accuracy of identifying micro-UAVs. Adaptive Laser Parameter Adjustment: Detection parameters are dynamically optimized based on reflectivity and the environment, expanding the system's applicability. Dual-Detector Collaborative Processing: Imaging and non-imaging detectors complement each other, balancing resolution and real-time performance.
[0120] The CMOS detector (1024×1024 resolution) directly outputs image data containing the spatial morphology of the target by acquiring imaging information of the laser reflected from the target. Geometric features of the target can be extracted from this image, such as the area of the light spot (target size), circularity (target shape regularity), and aspect ratio (target contour features), providing morphological basis for feature fusion.
[0121] As a non-imaging detector, the APD array focuses on capturing the peak signal of the echo light intensity. Combined with the intensity of laser emission The original reflectance is calculated using a formula.
[0122] Its high sensitivity (avalanche gain effect) can amplify weak echo signals, making it especially suitable for the accurate measurement of reflectivity of low reflectivity targets (such as small UAVs with reflectivity <1%), providing reliable raw data for subsequent reflectivity correction.
[0123] The dual detectors adopt a common optical path design to ensure that the detection field of view coincides and the time synchronization error is ≤1μs, realizing the spatiotemporal matching of "target geometric features-reflectivity". This provides hardware-level collaborative support for the innovation of "reflectivity + target geometric feature fusion", taking into account both imaging resolution and reflectivity measurement accuracy.
[0124] The various embodiments described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. All or part of this invention can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.
[0125] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present invention.
Claims
1. A method for detecting micro-UAVs based on target reflectivity, characterized in that, include: A pulsed laser is emitted toward the target area, and the beam is expanded and collimated to cover the detection area. Based on the pulsed laser, the target signal is synchronously acquired by dual detectors with a common optical path design to obtain the target's geometric features and reflectivity intensity signal; The original reflectivity is calculated based on the reflectivity intensity signal, and the target distance information is obtained. The original reflectivity is then corrected based on the target distance information to obtain a corrected reflectivity value. Target motion features are generated through multi-frame temporal analysis, and based on a fusion strategy, the corrected reflectivity value, the target geometric features, and the target motion features are fused to construct a comprehensive target feature vector; The target integrated feature vector is input into the classifier, and the recognition result of the micro-drone is output.
2. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The method, based on the pulsed laser, synchronously acquires target signals through dual detectors designed with a common optical path to obtain target geometric features and reflectivity intensity signals, including: The dual detectors include an imaging detector and a non-imaging detector; The imaging detector is a CMOS detector, used to acquire the geometric features of the target; The non-imaging detector is an APD avalanche photodiode array, used to acquire the reflectivity intensity signal.
3. The method for detecting micro-UAVs based on target reflectivity according to claim 2, characterized in that, The field-of-view overlap error between the imaging detector and the non-imaging detector is less than or equal to 0.1°, and the time synchronization error is less than or equal to 1μs.
4. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The formula for calculating the original reflectivity is: ； in, This represents the peak value of the echo intensity in the reflectivity intensity signal. This represents the peak value of the emitted light intensity in the reflectivity intensity signal; This represents the original reflectance.
5. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The step of acquiring target distance information and correcting the original reflectivity based on the target distance information to obtain a corrected reflectivity value includes: The distance information of the target is obtained using Time-of-Flight (TOF) ranging technology. ; An atmospheric attenuation model is used, based on the distance information. The calculated correction value is used to correct the original reflectance to eliminate the influence of atmospheric attenuation, resulting in a corrected reflectance value.
6. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The generation of target motion features through multi-frame temporal analysis includes: Based on continuous multi-frame data from the imaging detector, the position coordinates of the target at adjacent time points are obtained; Calculate the instantaneous velocity and acceleration based on the time difference of the position coordinates; Calculate the trajectory direction angle based on the spatial relationship of the position coordinates; The position coordinates, instantaneous velocity, acceleration, and trajectory direction angle at adjacent moments are used as the motion features of the target.
7. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The fusion strategy includes: The weights of the target geometric features in the fusion process are dynamically adjusted based on the magnitude of the corrected reflectance value. This dynamic adjustment includes: increasing the weights of the target geometric features when the corrected reflectance value is below 5%; and decreasing the weights of the target geometric features when the corrected reflectance value is above 30%. Construct a reflectance-target geometric feature correlation factor; wherein, the reflectance-target geometric feature correlation factor includes: a reflectance-size correlation factor and a reflectance-shape correlation factor; the reflectance-size correlation factor is expressed as: ； The reflectivity-shape correlation factor is expressed as: ； in, Indicates the area of the light spot; Indicates roundness; This indicates the corrected reflectance value.
8. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The process of fusing the corrected reflectivity value, the target geometric features, and the target motion features to construct a comprehensive target feature vector includes: Multiple feature parameters are extracted based on the corrected reflectivity value, the target geometric features, and the target motion features; Multiple parameters are normalized separately, and the normalized feature parameters are concatenated in a fixed order to obtain the target comprehensive feature vector.
9. The method for detecting micro-UAVs based on target reflectivity according to claim 1, characterized in that, The classifier is a machine learning model-based classifier, including at least one of support vector machines and convolutional neural networks.
10. A micro-UAV detection system based on target reflectivity, characterized in that, include: The laser emitting module is used to emit pulsed lasers towards the target area, which are then expanded and collimated to cover the detection area. The dual-detector module is used to synchronously acquire target signals based on the pulsed laser through dual detectors designed with a common optical path, thereby obtaining target geometric features and reflectivity intensity signals. The signal processing module is used to calculate the original reflectivity based on the reflectivity intensity signal, obtain target distance information, and correct the original reflectivity based on the target distance information to obtain a corrected reflectivity value. The feature fusion module is used to generate target motion features through multi-frame temporal analysis, and based on the fusion strategy, fuse the corrected reflectivity value, the target geometric features and the target motion features to construct a comprehensive target feature vector; The classification and recognition module is used to input the comprehensive feature vector of the target into the classifier and output the recognition result of the micro-drone.