Low-altitude echo target grading method and system

Abstract

The embodiment of the application discloses a low-altitude echo target grading method and system, and relates to the technical field of radar signal processing and target pattern recognition. The method comprises the following steps: obtaining echo data of a target in a low-altitude detection area in a continuous detection period; extracting a plurality of scattering centers from the echo data corresponding to each detection period; establishing a time sequence correlation between the scattering centers according to the scattering centers in adjacent detection periods; determining the appearance, disappearance, enhancement and weakening states of the scattering centers in the attitude change process of the target, and identifying a replacement scattering center for the disappeared scattering center; constructing a scattering center replacement chain according to the state change time of each scattering center and the bearing relationship; extracting the periodicity, spatial symmetry and state continuity of the replacement order from the scattering center replacement chain to generate a scattering structure evolution feature; and generating a low-altitude echo target grading result according to the scattering structure evolution feature. The application improves the stability, continuity and accuracy of low-altitude echo target grading.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing and target pattern recognition technology, specifically to a method and system for classifying low-altitude echo targets. Background Technology

[0002] Low-altitude targets are typically characterized by their small size, low flight altitude, frequent changes in motion, and susceptibility to interference from ground clutter and building reflections. When radar detects low-altitude targets, the target's fuselage, wings, rotor, wingtips, or other structural components create multiple scattering centers. As the target's attitude, heading, and radar observation angle change, the echo intensity and spatial location of each scattering center also change accordingly.

[0003] Existing low-altitude target identification methods typically extract features such as radar cross-section, velocity, micro-Doppler frequency, echo intensity, or time-frequency images from echo data in a single detection cycle, and then use classification models to identify the target type. However, when the target's attitude changes, the original scattering center may gradually weaken or disappear, while scattering centers corresponding to other scattering parts may appear or strengthen. In this case, the single-frame echo features are prone to significant jumps, and continuous tracking of the same scattering center is also prone to correlation interruptions.

[0004] Existing methods typically treat the appearance and disappearance of scattering centers as detection errors or the re-establishment of new scattering center trajectories, failing to further analyze the structural continuity between disappearing and newly appearing scattering centers. Consequently, they struggle to accurately represent the continuous evolution of the target structure as its attitude changes. Under conditions of low signal-to-noise ratio, rapid attitude changes, or multiple scattering centers being close to each other, problems such as incorrect association of scattering centers and unstable target classification results can easily arise. Summary of the Invention

[0005] This application provides a method and system for classifying low-altitude echo targets, which improves the stability, continuity and accuracy of low-altitude echo target classification.

[0006] This application provides the following solution:

[0007] According to the first aspect, a method for classifying low-altitude echo targets is provided, comprising: acquiring echo data of targets within a low-altitude detection area during continuous detection cycles; extracting multiple scattering centers from the echo data corresponding to each detection cycle; establishing a temporal correlation between scattering centers based on the positional changes, intensity changes, and phase continuity relationships of each scattering center in adjacent detection cycles; determining the appearance, disappearance, enhancement, and weakening states of each scattering center during target attitude changes based on the temporal correlation, and identifying replacement scattering centers used to take over the disappearing scattering centers; constructing a scattering center replacement chain according to the state change times and takeover relationships of each scattering center; extracting the periodicity, spatial symmetry, and state continuity of the replacement sequence from the scattering center replacement chain to generate scattering structure evolution characteristics; and generating low-altitude echo target classification results based on the scattering structure evolution characteristics.

[0008] According to the second aspect, a low-altitude echo target classification system is provided, comprising: a data acquisition module for acquiring echo data of targets within a low-altitude detection area during continuous detection cycles; a scattering center extraction module for extracting multiple scattering centers from the echo data corresponding to each detection cycle; a temporal correlation module for establishing a temporal correlation based on the positional changes, intensity changes, and phase continuity relationships of each scattering center; a replacement identification module for determining the appearance, disappearance, enhancement, and weakening states of each scattering center during target attitude changes based on the temporal correlation, and identifying replacement scattering centers used to take over the disappeared scattering centers; a replacement chain construction module for constructing a scattering center replacement chain according to the state change times and succession relationships of each scattering center; a feature generation module for extracting the periodicity, spatial symmetry, and state continuity of the replacement sequence from the scattering center replacement chain to generate scattering structure evolution features; and a target classification module for generating low-altitude echo target classification results based on the scattering structure evolution features.

[0009] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0010] This application establishes temporal correlations between scattering centers by jointly analyzing the positional changes, intensity changes, and phase continuity relationships of multiple scattering centers within a continuous detection period. Furthermore, it identifies the connection between disappearing and replacing scattering centers, thus organizing the discrete echo changes during target attitude changes into a continuous chain of scattering center replacements. Based on this replacement chain, the periodicity, spatial symmetry, and state continuity of the replacement sequence are extracted, enabling a more accurate characterization of the dynamic evolution of the target scattering structure. This reduces feature jumps and erroneous correlations caused by attitude changes, local obstruction, echo intensity fluctuations, or short-term disappearance of scattering centers, improving the stability, continuity, and accuracy of low-altitude echo target classification.

[0011] Of course, any product implementing this application does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart of a low-altitude echo target classification method provided in an embodiment of this application;

[0014] Figure 2 This is a schematic diagram of the replacement scattering center identification process provided in the embodiments of this application;

[0015] Figure 3 This is a schematic diagram of the scattering center replacement chain construction provided in an embodiment of this application;

[0016] Figure 4 This is a structural block diagram of a low-altitude echo target classification system provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0018] Figure 1 A flowchart illustrating the low-altitude echo target classification method provided in this application embodiment. Figure 1 As shown, the method may include the following steps:

[0019] Step 101: Acquire echo data of targets within the low-altitude detection area during continuous detection cycles.

[0020] Step 102: Extract multiple scattering centers from the echo data corresponding to each detection period.

[0021] Step 103: Establish the temporal correlation between scattering centers based on the position changes, intensity changes, and phase continuity of each scattering center in adjacent detection periods.

[0022] Step 104: Based on the temporal correlation, determine the appearance, disappearance, enhancement and weakening states of each scattering center during the target attitude change process, and identify the replacement scattering center used to receive the disappeared scattering center.

[0023] Step 105: Construct a scattering center replacement chain according to the state change time and succession relationship of each scattering center.

[0024] Step 106: Extract the periodicity, spatial symmetry and state continuity of the replacement sequence from the replacement chain of the scattering center to generate the evolution characteristics of the scattering structure.

[0025] Step 107: Generate low-altitude echo target classification results based on the scattering structure evolution characteristics.

[0026] As can be seen from the above process, this application establishes temporal correlations between scattering centers by jointly analyzing the position changes, intensity changes, and phase continuity relationships of multiple scattering centers within a continuous detection period. Furthermore, it identifies the connection between disappearing scattering centers and their replacements, thereby organizing the discrete echo changes during target attitude changes into a continuous chain of scattering center replacements. Based on this replacement chain, the periodicity, spatial symmetry, and state continuity of the replacement sequence can be extracted, enabling a more accurate characterization of the dynamic evolution of the target scattering structure. This reduces feature jumps and erroneous correlations caused by attitude changes, local obstruction, echo intensity fluctuations, or short-term disappearance of scattering centers, thus improving the stability, continuity, and accuracy of low-altitude echo target classification.

[0027] The following describes in detail each step of the above process and the effects that can be further produced, with reference to the embodiments.

[0028] Step 101 specifically involves acquiring echo data of targets within the low-altitude detection area during continuous detection cycles.

[0029] Acquiring echo data of targets within a low-altitude detection area during a continuous detection cycle refers to using radar equipment located within or around the low-altitude detection area to continuously scan targets entering the detection range and receive the echo signals formed after the radar detection signals are reflected by different structural parts of the targets. The radar equipment can be a phased array radar, millimeter-wave radar, multiple-input multiple-output radar, or other detection equipment capable of acquiring target distance and angle information. The low-altitude detection area can be the area surrounding airports, urban low-altitude areas, industrial parks, border areas, areas surrounding important facilities, or other areas requiring low-altitude target monitoring.

[0030] A continuous detection cycle refers to multiple temporally adjacent detection cycles formed by radar equipment continuously detecting the same target at a preset update frequency. Each detection cycle can correspond to a complete scan, a beam dwell process, or a target echo update process. By continuously detecting the same target, the echo changes of the target at different times and under different observation attitudes can be obtained, providing a time basis for subsequent analysis of the appearance, disappearance, enhancement, and weakening of the scattering center. The number and duration of continuous detection cycles can be set according to the target's movement speed, radar scanning frequency, and target classification accuracy requirements.

[0031] The echo data in each detection cycle includes at least the range, angle, intensity, and phase data of the target echo. The range data characterizes the distance differences between different scattering parts of the target and the radar equipment; the angle data characterizes the azimuth or elevation angle of each scattering part; the echo intensity data characterizes the strength of reflection of the radar detection signal by different scattering parts; and the phase data characterizes the phase change of the echo signal within consecutive detection cycles. In practice, the echo data may also include the Doppler frequency, timestamp, radar channel number, and target detection number.

[0032] After acquiring the raw echo signal, it can be divided according to the detection period, and the data corresponding to each detection period can be processed by range compression, beamforming, clutter suppression, noise filtering, and phase correction. Through the above processing, the influence of reflections from fixed objects and buildings on the target echo and system noise is reduced, forming continuous echo data with unified time reference and spatial coordinates that can be used for scattering center extraction.

[0033] When multiple targets exist simultaneously within a low-altitude detection area, the target echo data corresponding to each target can be separated from the received echo data based on the distance range, angle range, and motion continuity relationship of each target, and target numbers can be assigned to different targets. For the same target, its echo data in multiple consecutive detection cycles are stored in chronological order, so that the scattering centers in each detection cycle can be compared in subsequent processing based on changes in position, intensity, and phase continuity.

[0034] Step 102 specifically involves extracting multiple scattering centers from the echo data corresponding to each detection period.

[0035] Extracting multiple scattering centers from the echo data corresponding to each detection cycle refers to decomposing the target echo in spatial dimensions within each detection cycle, so that the reflection areas corresponding to different structural parts of the target can be distinguished from the overall echo. Since low-altitude targets are usually composed of fuselages, wings, rotors, wingtips, supports, or other local structures, the reflection intensity, position, and phase of different structural parts to radar signals are not completely identical. Therefore, these local echo areas with relatively independent reflection characteristics can be extracted as scattering centers.

[0036] In practice, the echo data corresponding to each detection cycle are first subjected to joint imaging in both range and angle dimensions. The range dimension is used to distinguish the distance differences between different reflective parts of the target and the radar, while the angle dimension is used to distinguish the spatial differences between different reflective parts of the target in the azimuth or elevation directions. Through the joint processing of the range and angle dimensions, a range-angle echo map corresponding to the target can be formed, transforming the originally mixed echo signals into a spatially clear echo intensity distribution result.

[0037] After generating the range-angle echo map, local echo peak regions can be searched within the map. A local echo peak region refers to an echo area where the echo intensity is higher than the surrounding area and meets a preset detection threshold. This region typically corresponds to a structural part of the target with strong reflectivity, such as the fuselage edge, rotor tip, wing edge, or metal connection. By identifying multiple local echo peak regions, several candidate scattering regions of the target in the current detection cycle can be preliminarily obtained.

[0038] Since the same scattering region may form multiple adjacent peaks in the distance or angle dimensions, or the same reflection region may be split into multiple peaks due to noise and imaging sidelobes, each peak cannot be simply taken as an independent scattering center. This application further combines and judges based on the peak intensity, phase convergence, and spatial spacing of each local echo peak region. Peak intensity is used to determine whether the region has a stable reflection contribution, phase convergence is used to determine whether adjacent peaks have the same source reflection characteristics, and spatial spacing is used to determine whether multiple peaks belong to the local unfolding result of the same structural region.

[0039] When the spatial interval between multiple local echo peak regions is small, the peak intensity variation is within the allowable range, and the phase state has high consistency, these peaks can be considered to originate from the same scattering location and merged into a single scattering center. After merging, the intensity-weighted center or peak location of the merged region can be taken as the location of the scattering center, the maximum or average intensity of the merged region can be taken as the echo intensity of the scattering center, and the representative value of the phase within the merged region can be taken as the phase state of the scattering center.

[0040] Through the above processing, a set of scattering centers can be obtained for each detection cycle. This set of scattering centers includes not only their spatial location but also their echo intensity and phase state, providing a basis for subsequent comparisons of position changes, intensity changes, and phase continuity between adjacent detection cycles. This method can reduce the impact of noise peaks, sidelobe peaks, or repeated counting of the same scattering location on subsequent time-series correlations, improving the stability and traceability of the scattering center extraction results.

[0041] Step 103 specifically involves establishing a temporal correlation between scattering centers based on the positional changes, intensity changes, and phase continuity of each scattering center in adjacent detection periods.

[0042] Establishing a temporal correlation between scattering centers based on their positional changes, intensity variations, and phase continuity relationships across adjacent detection cycles involves matching scattering centers extracted in the previous cycle with those extracted in the next cycle during continuous detection. This allows for the determination of which scattering centers represent the continuous behavior of the same target structural component at different times. Since low-altitude targets undergo displacement, turning, pitching, or attitude swaying during flight, the position of the same scattering component may change between adjacent detection cycles. Therefore, it is insufficient to simply require two scattering centers to be completely identical in position; a comprehensive judgment must consider the range of positional changes, intensity trends, and phase continuity.

[0043] In practice, each scattering center in the current detection cycle can be considered as a scattering center to be associated, and multiple scattering centers in adjacent detection cycles can be considered as objects to be matched. First, the positional changes between the scattering centers to be associated and each object to be matched are calculated. These positional changes can be determined based on changes in distance, azimuth, elevation, or three-dimensional coordinates. If the positional change of an object to be matched relative to the scattering center exceeds a preset association range, it indicates that the spatial jump between the two in adjacent detection cycles is too large, and they usually do not belong to continuous echoes from the same scattering location; therefore, they can be excluded.

[0044] For scattering centers whose positional changes meet the preset correlation range, their intensity change direction and phase continuity are further compared. The intensity change direction reflects whether the scattering center strengthens, weakens, or remains stable within adjacent detection periods. If the positional changes between two scattering centers are reasonable, and the intensity change direction matches the local echo change trend under target attitude changes, they can be considered as candidate correlation objects. The phase continuity reflects whether the echo signals corresponding to the two scattering centers have continuous propagation characteristics between adjacent detection periods. If the phase abrupt change is too large, it may indicate that the scattering center is not a continuous echo generated from the same structural part.

[0045] When a scattering center in a current detection cycle corresponds to multiple candidate associated scattering centers, it indicates that multiple local echo peaks with similar locations or intensity changes exist in the next detection cycle. In this case, phase continuity is prioritized for screening, and the candidate associated scattering center with the smoothest phase change, smallest phase difference, or closest phase trend is identified as the target associated scattering center. This reduces erroneous associations caused by adjacent scattering points being too close, echo intensities being similar, or imaging sidelobes, thus improving the stability of scattering center tracking.

[0046] After identifying the target-associated scattering centers, a correlation is established between the scattering centers in the current detection cycle and their corresponding target-associated scattering centers. The positional changes, intensity changes, and phase continuity corresponding to this correlation are recorded. As multiple consecutive detection cycles are processed sequentially, a continuous correlation result of the scattering centers in the time direction can be formed, i.e., a temporal correlation between the scattering centers. This temporal correlation provides the basis for subsequent judgments on the appearance, disappearance, enhancement, and weakening of scattering centers, as well as the identification of replacement scattering centers, enabling subsequent processing to distinguish between normal continuous changes in scattering centers and scattering center replacement phenomena caused by attitude changes.

[0047] Step 104 specifically involves: determining the appearance, disappearance, enhancement, and weakening states of each scattering center during the target attitude change process based on the aforementioned temporal correlation, and identifying replacement scattering centers used to receive the disappeared scattering centers.

[0048] Determining the appearance, disappearance, enhancement, and weakening states of each scattering center during target attitude changes, based on the established temporal correlation of scattering centers, involves observing the presence and echo intensity trends of each scattering center in consecutive detection cycles. If a scattering center newly appears in the current detection cycle and has no corresponding preceding associated scattering center in the previous detection cycle, it can be identified as appearing. If a scattering center fails to find a subsequent associated scattering center within several consecutive detection cycles, or its echo intensity drops below a preset retention threshold, it can be identified as disappearing. If the echo intensity of a scattering center gradually increases in consecutive detection cycles, it can be identified as enhancing; if its echo intensity gradually decreases, it can be identified as weakening.

[0049] When the attitude of a low-altitude target changes, some structural components may gradually move away from the radar line of sight, causing their corresponding scattering centers to weaken and eventually disappear; simultaneously, other structural components may turn towards the radar line of sight, causing their corresponding scattering centers to appear or strengthen. In this case, the newly appearing or strengthened scattering centers are not necessarily independent new structural features, but may be a continuation of the structural response represented by the disappearing scattering centers. Therefore, this application does not simply delete the disappearing scattering centers or treat the newly appearing scattering centers as new, unrelated targets, but rather further identifies replacement scattering centers to take over from the disappearing scattering centers.

[0050] In practice, the disappearance time and location corresponding to the disappearing scattering center are first determined. The disappearance time can be the detection period in which the scattering center last met the detection conditions, or the detection period in which its intensity decreased to a preset disappearance threshold. The disappearance location can be the location of the scattering center in the last effective detection period before disappearance, or the location predicted by the positional change trend of the scattering center over multiple detection periods before disappearance. By recording the disappearance time and location, the time range and spatial reference location for subsequent locating and replacing the scattering center can be determined.

[0051] After the disappearance time, the system selects newly emerging or intensified scattering centers from subsequent detection cycles as candidate replacement scattering centers. Newly emerging scattering centers are those detected for the first time after the disappearance time and have no corresponding temporal association in previous detection cycles. Intensified scattering centers are those whose echo intensity increases significantly after the disappearance time and may assume the scattered energy represented by the original disappeared scattering center. Using these two types of scattering centers as candidates can cover situations where the scattering response shifts from one structural part to another during target attitude changes.

[0052] Subsequently, the true replacement scattering center is determined based on the migration direction, intensity complementarity, and phase connection relationship between the candidate replacement scattering center and the disappearing scattering center. The migration direction is used to determine whether the location of the candidate replacement scattering center matches the positional change trend of the disappearing scattering center before it disappears. For example, if the disappearing scattering center shows a trend of moving to the right of the target before disappearing, then candidate replacement scattering centers that appear or intensify in this trend direction are preferentially selected.

[0053] The intensity complementarity relationship is used to determine whether the enhancement process of a candidate replacement scattering center corresponds to the weakening process of a disappearing scattering center. If the echo intensity of the disappearing scattering center gradually decreases, while the echo intensity of a candidate replacement scattering center or its neighboring region gradually increases within the corresponding time range, and the intensity changes of the two can form a good continuity, it indicates that the local scattering contribution of the target may have shifted from the disappearing scattering center to the candidate replacement scattering center.

[0054] Phase continuity is used to further determine whether the candidate replacement scattering center and the vanishing scattering center have a continuous electromagnetic scattering response. Based on the phase change trend of the vanishing scattering center before its disappearance, its phase state near the moment the candidate replacement scattering center appears can be predicted. Then, the actual phase of the candidate replacement scattering center is compared with the predicted phase. If the phase difference is small or the phase change trend is continuous, it indicates that the candidate replacement scattering center is more likely to have inherited the structural response of the vanishing scattering center.

[0055] By simultaneously utilizing migration direction, intensity complementarity, and phase continuity, we can avoid mistaking accidentally enhanced noise peaks, background clutter, or other irrelevant scattering components as replacement scattering centers. The ultimately determined replacement scattering centers represent the continuity of scattering responses during target attitude changes and provide a foundation for constructing a replacement scattering center chain.

[0056] As one feasible approach, determining the replacement scattering center based on the migration direction, intensity complementarity, and phase connection relationship between the candidate replacement scattering center and the vanished scattering center includes:

[0057] The predicted migration direction is determined based on the position change of the disappearing scattering center before it disappears, and a first candidate scattering center whose appearance direction is consistent with the predicted migration direction is selected from the candidate replacement scattering centers.

[0058] The intensity decay sequence of the vanishing scattering center is compared with the intensity growth sequence of each first candidate scattering center to determine the second candidate scattering center with an intensity complementary relationship.

[0059] Based on the phase change trend of the disappearing scattering center before its disappearance, predict its phase at the moment when the candidate replacement scattering center appears, and determine the second candidate scattering center with the highest degree of connection between the actual phase and the predicted phase as the replacement scattering center.

[0060] Figure 2 This is a schematic diagram of the replacement scattering center identification process provided in an embodiment of this application. In specific implementation, the predicted migration direction is first determined based on the positional changes of the disappearing scattering center over multiple consecutive detection cycles before its disappearance. This predicted migration direction can be obtained from the continuous positional changes of the disappearing scattering center in the range dimension, angle dimension, or a unified spatial coordinate system, representing the movement trend of the scattering center before its disappearance. Subsequently, the appearance positions of each candidate replacement scattering center are compared with the disappearance position of the disappearing scattering center to determine the appearance direction of the candidate replacement scattering center relative to the disappearing scattering center. If the appearance direction is consistent with the predicted migration direction, it indicates that the candidate replacement scattering center spatially conforms to the scattering response transfer trend caused by the target attitude change, and can be used as the first candidate scattering center.

[0061] After identifying the first candidate scattering center, the intensity decay sequence of the disappearing scattering center is further compared with the intensity growth sequence of each first candidate scattering center. The intensity decay sequence of the disappearing scattering center can be formed by the gradual decrease in echo intensity during several detection periods before its disappearance, while the intensity growth sequence can be formed by the gradual increase in echo intensity during the appearance or enhancement of the first candidate scattering center. If the time period of weakening of the disappearing scattering center corresponds to the time period of enhancement of the first candidate scattering center, and the attenuation and growth satisfy a preset complementary relationship, it indicates that the local scattering energy of the target may have transferred from the disappearing scattering center to the first candidate scattering center, thus it can be identified as the second candidate scattering center.

[0062] Preferably, the step of comparing the intensity decay sequence of the vanishing scattering center with the intensity growth sequence of each first candidate scattering center to determine the second candidate scattering center with an intensity complementary relationship includes: determining the state transition time window in which the intensity decay of the vanishing scattering center occurs, and extracting the intensity changes of each first candidate scattering center and its adjacent scattering centers within the state transition time window; determining the vanishing scattering intensity loss from the intensity decay sequence, and determining the corresponding local scattering intensity carrying capacity based on the intensity growth of each first candidate scattering center and its adjacent scattering centers; determining the intensity transfer residual corresponding to each first candidate scattering center based on the difference between the vanishing scattering intensity loss and the local scattering intensity carrying capacity; and determining the first candidate scattering center whose intensity growth start time is later than the intensity decay start time, whose intensity transfer residual satisfies a preset closure condition, and whose local scattering intensity carrying capacity is mainly concentrated in the corresponding first candidate scattering center as the second candidate scattering center.

[0063] In practice, the first step is to determine the state transition time window for intensity attenuation at the disappearing scattering center. This time window can start from the detection period when the echo intensity of the disappearing scattering center begins to decrease continuously, and end when the disappearing scattering center falls below the retention threshold or is completely undetectable. Alternatively, it can be extended to the detection period when the first candidate scattering center completes its enhancement. By setting a state transition time window, the intensity complementarity judgment can be limited to the critical period when the scattering response shifts due to changes in target attitude, reducing the impact of echo fluctuations in irrelevant time periods on the judgment result.

[0064] Within the state transition time window, the intensity attenuation sequence of the disappearing scattering center is extracted, and simultaneously, the intensity changes of each first candidate scattering center and its adjacent scattering centers are extracted. Adjacent scattering centers are introduced here because when the target attitude changes, the energy of the original scattering center does not necessarily all concentrate and transfer to a single new scattering center; it may also form local diffusion or multiple adjacent peaks around the new scattering center. Therefore, treating the first candidate scattering center and its adjacent scattering centers as a local receiving region better reflects the actual redistribution of scattered energy in radar echoes.

[0065] Subsequently, the intensity loss of vanishing scattering centers is determined from the intensity decay sequence of the vanishing scattering centers. This intensity loss can be accumulated based on the intensity decrease of the scattering center within the state transition time window, representing the scattering contribution lost by the original scattering location during attitude changes. For each first candidate scattering center, the intensity increase of itself and its adjacent scattering centers within the same time window is statistically analyzed, and the corresponding local scattering intensity absorption capacity is determined accordingly. This local scattering intensity absorption capacity indicates the degree to which the candidate region can absorb the original scattering intensity loss.

[0066] Then, based on the difference between the intensity loss of vanishing scattering and the intensity absorbed by the local scattering, the intensity transfer residual corresponding to each first candidate scattering center is determined. If the intensity absorbed by the local scattering corresponding to a certain first candidate scattering center is close to the intensity loss of vanishing scattering, it indicates that the candidate region has a good ability to absorb the intensity change of the original vanishing scattering center, and the intensity transfer residual is small. If the intensity absorbed by the local scattering is significantly too large or too small, it indicates that its enhancement may originate from background clutter, overall gain changes, reflection changes from other structural parts, or noise peaks, and should not be directly identified as absorption of the vanishing scattering center.

[0067] When determining the second candidate scattering center, the temporal sequence condition that the intensity increase begins later than the intensity decay must also be met. This condition ensures that the intensity enhancement is a subsequent response to the weakening process of the disappearing scattering center, rather than an independent enhancement process that existed before the weakening of the disappearing scattering center. Simultaneously, the intensity transfer residual must satisfy a preset closure condition, meaning that a relatively closed transfer relationship can be formed between the loss of the original scattering intensity and the inheritance of the local scattering intensity in the candidate region.

[0068] Furthermore, it is necessary to determine whether the local scattering intensity is mainly concentrated at the corresponding first candidate scattering center. If the local scattering intensity is mainly contributed by the first candidate scattering center, it indicates that it is the main enhanced scattering point in the local receiving region and can be considered as the second candidate scattering center. If the intensity increase is mainly dispersed among multiple adjacent scattering centers, or mainly contributed by other scattering centers in the neighboring region, it indicates that the first candidate scattering center is not the main receiving point and should not be determined as the second candidate scattering center.

[0069] After obtaining the second candidate scattering center, further confirmation is needed using phase continuity relationships. Specifically, based on the phase change trend of the vanishing scattering center before its disappearance, the phase state it should have near the time of appearance of the candidate replacement scattering center can be predicted. This prediction can be based on the phase change direction, phase change rate, or phase difference sequence of the vanishing scattering center in continuous detection cycles. Then, the actual phase of each second candidate scattering center at the time of appearance is compared with the predicted phase to determine the degree of continuity between the two.

[0070] When the actual phase of a second candidate scattering center differs little from its predicted phase, or when its phase change trend smoothly connects with the phase change trend of the disappearing scattering center before its disappearance, the second candidate scattering center can be considered to have a high degree of phase continuity with the disappearing scattering center. Ultimately, the second candidate scattering center with the highest degree of phase continuity is determined as the replacement scattering center. This process allows for the selection of the replacement object that best matches the continuous electromagnetic scattering response from multiple candidate objects that are spatially close and have similar intensity changes.

[0071] Step 105 specifically involves constructing a scattering center replacement chain based on the state change times and succession relationships of each scattering center.

[0072] Constructing a scattering center replacement chain based on the time of state changes and their successive relationships involves organizing the replacement process between scattering centers into a chain structure with temporal and directional relationships, after identifying the disappearing scattering centers and their corresponding replacement centers. This chain structure is used to represent how the original scattering centers gradually disappear, how new scattering centers appear or strengthen during target attitude changes, and whether there is a continuous structural succession relationship between them. By constructing a scattering center replacement chain, the information fragmentation caused by only recording the appearance and disappearance of individual scattering centers can be avoided, enabling subsequent analysis of the evolution of the target's scattering structure as a whole.

[0073] Figure 3This is a schematic diagram illustrating the construction of a scattering center replacement chain according to an embodiment of this application. In specific implementation, each scattering center can be initially defined as a node. The node can record the scattering center's number, appearance time, disappearance time, spatial location, intensity change state, and phase state. If a scattering center exists in multiple consecutive detection cycles, it can be considered a continuation of the same node in the time direction; if the scattering center disappears, its disappearance time is used as the time reference for establishing subsequent succession relationships.

[0074] For each disappearing scattering center, a corresponding replacement scattering center is identified after its disappearance time, and the appearance time or enhancement start time of the replacement scattering center is determined. When the disappearing scattering center and the replacement scattering center satisfy the aforementioned migration direction, intensity complementarity, and phase connection relationship, a directed connection is established from the node corresponding to the disappearing scattering center to the node corresponding to the replacement scattering center. This directed connection represents a transition of the target scattering response from the original scattering center to the replacement scattering center, rather than representing a normal spatial adjacency relationship.

[0075] After establishing multiple directed connections, they are arranged according to the order of their disappearance and appearance times. That is, the earlier disappearance and replacement relationships of scattering centers are arranged first, and the later disappearance and replacement relationships of scattering centers are arranged later. If a previous replacement scattering center weakens or disappears in a subsequent detection cycle and is taken over by a new scattering center, a new replacement node is connected after the replacement scattering center node, thus forming a continuous replacement path.

[0076] When the same scattering center reappears in subsequent detection cycles, it is not simply treated as a new, unrelated node. Instead, the node corresponding to the reappearing scattering center is associated with its original node. This can represent the repeated appearance of a structural part at different observation angles during target attitude changes. For example, the same wing edge or rotor tip may disappear during target turning and then re-enter a strong reflection state as attitude changes. By associating the reappearing node with the original node, a replacement chain with cyclic characteristics can be formed.

[0077] The scattering center replacement chain obtained through the above processing can simultaneously reflect the temporal sequence, directional relationship, and recurrence relationship among scattering centers. This replacement chain provides a data foundation for the subsequent extraction of the periodicity, spatial symmetry, and state continuity of the replacement sequence, enabling target classification to no longer rely on the static scattering characteristics of a single detection cycle, but rather on the dynamic evolution law of the target scattering structure during continuous attitude changes.

[0078] Step 106 specifically involves extracting the periodicity, spatial symmetry, and state continuity of the replacement sequence from the scattering center replacement chain to generate scattering structure evolution characteristics.

[0079] Extracting the periodicity, spatial symmetry, and state continuity of the replacement sequence from the scattering center replacement chain to generate scattering structure evolution characteristics means, based on the already constructed scattering center replacement chain, no longer focusing solely on whether a single scattering center appears or disappears, but further analyzing whether the replacement process between multiple scattering centers exhibits stable temporal patterns, spatial distribution patterns, and continuity. Since the scattering centers corresponding to different structural parts of low-altitude targets often enhance, weaken, disappear, and reappear in a certain sequence during turning, swaying, pitching, or periodic attitude changes, the replacement chain can reflect the dynamic evolution characteristics of the target's scattering structure.

[0080] In practice, the recurring node sequence can be extracted from the scattering center replacement chain. A node sequence refers to the combination of scattering center nodes that appear sequentially in the replacement chain. For example, if scattering center A disappears, scattering center B takes over, and scattering center C takes over in a certain time period, and the same or similar node arrangement from A to B to C reappears in subsequent time periods, then the node sequence can be considered repetitive. The recurring node sequence reflects the stable scattering replacement pattern formed by the target under periodic attitude changes or periodic component movements.

[0081] After identifying repeating node sequences, the time interval between two adjacent identical node sequences is calculated. If the time intervals between multiple repeating node sequences are relatively close, and the node order is basically consistent, it indicates that the scattering center replacement process has strong periodicity. If the time intervals between identical node sequences vary greatly, or the node order changes frequently, it indicates that the periodicity of the replacement order is weak. The periodicity of the replacement order can be quantitatively characterized by the dispersion of the time intervals and the consistency of the node arrangement.

[0082] Spatial symmetry can be extracted based on the position of each scattering center relative to the target echo center. The target echo center can be determined based on the overall echo energy distribution of the target, the weighted result of the positions of multiple scattering centers, or the center of the target detection frame. After determining the target echo center, it is determined whether there are paired scattering centers located on both sides, vertically, or in opposite spatial directions of the target echo center in the replacement chain. If the paired scattering centers exhibit an alternating characteristic of one side increasing while the other side decreases within a continuous detection period, followed by an inverse relationship, it indicates that the target scattering structure has a certain degree of spatial symmetry.

[0083] For example, for targets with left and right wings, symmetrical rotors, or symmetrical arms, when the target wobbles or turns, the left and right scattering centers may alternately increase or decrease in opposite directions. By determining whether the paired scattering centers alternately increase or decrease in opposite spatial directions, the symmetrical scattering response of the target structure in space can be identified, thus distinguishing it from the random changes in scattering points exhibited by flexible floating objects, clutter peaks, or targets with irregular reflections.

[0084] State continuity reflects the integrity of the succession relationships in the replacement chain of scattering centers. Specifically, it checks whether there are any unconnected disappearing states between adjacent nodes, or whether there are any appearing states without a preceding source. If a scattering center disappears and a corresponding replacement scattering center can be found within a preset time range, and the newly appearing scattering center can be traced back to a previously disappeared scattering center, then the state continuity between scattering centers is relatively continuous. If a large number of scattering centers disappear directly without corresponding replacement nodes, or a large number of scattering centers suddenly appear without a preceding source, then the evolution process of the scattering structure has a break, and the state continuity is low.

[0085] After obtaining the periodicity, spatial symmetry, and state continuity of the replacement sequence, these three factors can be combined to form a scattering structure evolution characteristic. This scattering structure evolution characteristic can be represented by eigenvectors, level identifiers, or a set of structured parameters, and at least includes the strength of periodicity, spatial symmetry, and state continuity. Using this characteristic, subsequent target classification steps can be based on the overall evolution law of the target scattering structure, rather than solely on the echo intensity or number of scattering centers at a single moment, thereby improving the stability and reliability of low-altitude echo target classification results.

[0086] Step 107 specifically involves generating a low-altitude echo target classification result based on the scattering structure evolution characteristics.

[0087] Generating low-altitude echo target classification results based on the aforementioned scattering structure evolution characteristics refers to, after obtaining the periodicity of the replacement sequence, spatial symmetry, and state continuity, further determining the stability of these evolution characteristics during continuous observation, and classifying the echo structure state of low-altitude targets accordingly. This step does not rely solely on the number or intensity of scattering centers at a single moment, but rather observes whether the target scattering structure maintains a similar evolutionary pattern across multiple consecutive time windows, thereby improving the stability of the target classification results.

[0088] In practice, the scattering center replacement chain can be segmented according to continuous time windows. Each continuous time window contains several adjacent detection periods, and the window length can be set according to the radar scanning frequency, target movement speed, and classification accuracy requirements. For each segment, the periodicity, spatial symmetry, and state continuity of the replacement sequence within that segment are extracted, and these three are combined to form the corresponding segmented evolution characteristics. This allows the scattering structure change process over a longer period to be divided into multiple local evolution segments, making it easier to determine whether the target scattering structure remains consistent across different time periods.

[0089] After obtaining multiple segmented evolutionary characteristics, the degree of change between adjacent segments can be calculated. This degree of change can be determined comprehensively based on the periodicity, spatial symmetry, and state continuity. If the changes in the evolutionary characteristics of two adjacent segments are small, it indicates that the target scattering structure remains relatively stable within that time range, and the corresponding segment can be identified as a stable evolutionary segment. If the changes in the evolutionary characteristics between two adjacent segments are significant, such as the sudden disappearance of periodicity, the disruption of spatial symmetry, or the appearance of a large number of unconnected scattering center states, the corresponding segment can be identified as a structural abrupt change segment.

[0090] After determining the stable evolution segments, the recurring scattering structure evolution patterns within each segment can be statistically analyzed. These patterns can be represented by the replacement sequence type within the segment, the alternating relationship of paired scattering centers, and the degree of state continuity. If a particular evolution pattern recurs in multiple stable evolution segments for a relatively long period, it can be identified as the dominant evolutionary pattern. The dominant evolutionary pattern is used to characterize the typical scattering structure changes exhibited by the target during the main observation period.

[0091] Subsequently, based on the persistence ratio of the dominant evolutionary pattern and the number and distribution of structural mutation segments, the classification result corresponding to the target is determined. A higher persistence ratio of the dominant evolutionary pattern indicates a more stable evolution of the target's scattering structure; a greater number or more dispersed distribution of structural mutation segments indicates more irregular changes in the target's scattering structure. If the persistence ratio of the dominant evolutionary pattern is high and the number of structural mutation segments is small, the target can be identified as having a stable evolutionary level. If the dominant evolutionary pattern still exists, but structural mutation segments are concentrated in certain time windows, the target can be identified as having a local mutation level. If there is no persistent dominant evolutionary pattern, or if structural mutation segments are scattered across multiple time windows, the target can be identified as having a disordered evolutionary level.

[0092] Finally, based on the determined stable evolution level, local mutation level, or disordered evolution level, a classification result for low-altitude echo targets is generated. This classification result may include the target number, classification level, dominant evolutionary mode, structural mutation segment location, and corresponding classification confidence level. Through this method, this application can utilize the temporal stability and mutation characteristics of scattering structure evolution features to perform more reliable classification of low-altitude targets, reducing the impact of short-term noise, occasional obstruction, or anomalous changes in individual scattering centers on the final classification result.

[0093] In one specific implementation, a millimeter-wave array radar is deployed in the low-altitude detection area. The radar continuously scans low-altitude targets entering the detection area at a refresh rate of 20 Hz, obtaining range, angle, intensity, and phase echo data of the targets in continuous detection cycles. The system first performs joint imaging of the echo data in the range and angle dimensions of each detection cycle, extracts local echo peak regions, and merges them into multiple scattering centers based on peak intensity, phase concentration, and spatial interval. Subsequently, for scattering centers in adjacent detection cycles, a temporal correlation is established based on position changes, intensity changes, and phase continuity to determine the appearance, disappearance, enhancement, and weakening states of each scattering center. For scattering centers that disappear, the system determines the predicted migration direction based on the position changes before disappearance, selects candidate replacement scattering centers from subsequently appearing or intensifying scattering centers, and determines the replacement scattering center by combining intensity complementarity and phase continuity relationships. Then, each scattering center is used as a node, and the connection between the disappearing scattering center and the replacement scattering center is used as a directed connection to construct the scattering center replacement chain. The periodicity, spatial symmetry and state continuity of the replacement order are extracted from it to generate the scattering structure evolution characteristics. Finally, the stable evolution level, local mutation level or disordered evolution level is output.

[0094] To verify the effectiveness of the proposed scheme, 160 sets of low-altitude echo sequences from multi-rotor UAVs, fixed-wing small targets, simulated bird targets, and flexible floating targets were tested, with each sequence lasting from 6 to 12 seconds. Test results showed that when using traditional single-frame echo intensity and the number of scattering centers for classification, the overall accuracy was 78.6%; when using only the continuity of scattering center trajectories, the overall accuracy was 84.3%; after adopting the proposed scheme, the overall classification accuracy reached 92.5%, with a stable evolution level identification accuracy of 94.1%, a local mutation level identification accuracy of 90.8%, and a disordered evolution level identification accuracy of 91.6%. Even with a signal-to-noise ratio reduced to 8 dB, the proposed scheme still maintained an overall accuracy of 86.9%, indicating that it can reduce the impact of short-term noise, attitude changes, and the transient disappearance of scattering centers on the target classification results.

[0095] According to another embodiment, a low-altitude echo target classification system is provided. Figure 4 A schematic block diagram of a low-altitude echo target classification system according to one embodiment is shown. Figure 4 As shown, the system includes:

[0096] The data acquisition module 401 is used to acquire echo data of targets in the low-altitude detection area during a continuous detection cycle.

[0097] The scattering center extraction module 402 is used to extract multiple scattering centers from the echo data corresponding to each detection period.

[0098] The temporal correlation module 403 is used to establish temporal correlation based on the position changes, intensity changes and phase continuity of each scattering center.

[0099] The replacement identification module 404 is used to determine the appearance, disappearance, enhancement and weakening states of each scattering center during the target attitude change process according to the time sequence correlation, and to identify the replacement scattering center used to take over the disappeared scattering center.

[0100] Replacement chain construction module 405 is used to construct a replacement chain for scattering centers according to the state change time and succession relationship of each scattering center.

[0101] The feature generation module 406 is used to extract the periodicity, spatial symmetry and state continuity of the replacement sequence from the scattering center replacement chain to generate scattering structure evolution features.

[0102] The target classification module 407 is used to generate low-altitude echo target classification results based on the scattering structure evolution characteristics.

[0103] As an implementable method, when the scattering center extraction module 402 extracts multiple scattering centers from the echo data corresponding to each detection period, it includes: performing joint imaging of the echo data of each detection period in the distance and angle dimensions to determine the local echo peak region; and merging the echo peaks belonging to the same scattering part according to the peak intensity, phase aggregation degree and spatial interval of each local echo peak region to obtain multiple scattering centers.

[0104] As an implementable approach, when the temporal correlation module 403 establishes temporal correlations between scattering centers based on the position changes, intensity changes, and phase continuity relationships of each scattering center in adjacent detection cycles, the following steps are taken: for each scattering center in the current detection cycle, determine its position changes relative to each scattering center in adjacent detection cycles; from scattering centers whose position changes satisfy a preset correlation range, determine candidate correlated scattering centers based on the direction of intensity change and the degree of phase continuity; when multiple candidate correlated scattering centers exist, determine the candidate correlated scattering center with the highest degree of phase continuity as the target correlated scattering center; and correlate the scattering centers in the current detection cycle with the corresponding target correlated scattering centers to obtain the temporal correlations between the scattering centers.

[0105] As one feasible approach, when identifying a replacement scattering center to take over a vanishing scattering center, the replacement identification module 404 includes: determining the vanishing time and vanishing position corresponding to the vanishing scattering center; determining candidate replacement scattering centers from scattering centers that newly appear or increase in intensity after the vanishing time; and determining the replacement scattering center based on the migration direction, intensity complementarity, and phase connection relationship between the candidate replacement scattering center and the vanishing scattering center.

[0106] As an implementable approach, when the replacement identification module 404 determines the replacement scattering center based on the migration direction, intensity complementarity, and phase connection relationship between the candidate replacement scattering center and the disappearing scattering center, the following steps are included: determining the predicted migration direction based on the position change of the disappearing scattering center before it disappears, and selecting a first candidate scattering center from the candidate replacement scattering centers whose appearance direction is consistent with the predicted migration direction; comparing the intensity decay sequence of the disappearing scattering center with the intensity growth sequence of each first candidate scattering center to determine a second candidate scattering center with an intensity complementarity relationship; predicting the phase of the disappearing scattering center at the moment of appearance of the candidate replacement scattering center based on the phase change trend of the disappearing scattering center before it disappears, and determining the second candidate scattering center with the highest degree of connection between the actual phase and the predicted phase as the replacement scattering center.

[0107] As an implementable approach, when the replacement identification module 404 compares the intensity decay sequence of the vanishing scattering center with the intensity growth sequence of each first candidate scattering center to determine the second candidate scattering center with an intensity complementary relationship, the following steps are taken: determining the state transition time window in which the intensity decay of the vanishing scattering center occurs, and extracting the intensity changes of each first candidate scattering center and its adjacent scattering centers within the state transition time window; determining the vanishing scattering intensity loss from the intensity decay sequence, and determining the corresponding local scattering intensity carrying capacity based on the intensity growth of each first candidate scattering center and its adjacent scattering centers; determining the intensity transfer residual corresponding to each first candidate scattering center based on the difference between the vanishing scattering intensity loss and the local scattering intensity carrying capacity; and determining the first candidate scattering center whose intensity growth start time is later than the intensity decay start time, whose intensity transfer residual satisfies a preset closure condition, and whose local scattering intensity carrying capacity is mainly concentrated in the corresponding first candidate scattering center, as the second candidate scattering center.

[0108] As an implementable approach, the replacement chain construction module 405, when constructing the scattering center replacement chain according to the state change times and succession relationships of each scattering center, includes: determining each scattering center as a corresponding node; for each disappearing scattering center, determining the appearance time of the corresponding replacement scattering center after its disappearance time, and establishing a directed connection from the node corresponding to the disappearing scattering center to the node corresponding to the replacement scattering center; arranging multiple directed connections consecutively according to the order of the disappearance time and appearance time corresponding to each directed connection; when the same scattering center reappears in a subsequent detection cycle, associating the node corresponding to the reappearing scattering center with its original node to obtain the scattering center replacement chain.

[0109] As an implementable approach, the feature generation module 406, when extracting the periodicity, spatial symmetry, and state continuity of the replacement sequence from the scattering center replacement chain, includes: extracting a repeating node arrangement sequence from the scattering center replacement chain; calculating the time interval between two adjacent identical node arrangement sequences; determining the periodicity of the replacement sequence based on the discreteness of the time interval and the consistency of the node arrangement; determining the paired scattering centers in the replacement chain based on the position of each scattering center relative to the target echo center; determining the spatial symmetry by judging whether the paired scattering centers alternately enhance or weaken in opposite spatial directions; and determining the state continuity based on whether there are unaccompanied disappearing states or appearing states without a preceding source between adjacent nodes.

[0110] As an implementable approach, the target classification module 407, when generating low-altitude echo target classification results based on the scattering structure evolution characteristics, includes: segmenting the scattering center replacement chain according to a continuous time window, and generating segmented evolution characteristics based on the periodicity, spatial symmetry, and state continuity of the replacement order corresponding to each segment; calculating the degree of change between adjacent segment evolution characteristics to determine stable evolution segments and structural mutation segments; determining the dominant evolutionary mode based on the recurring scattering structure evolution patterns in the stable evolution segments; determining the stable evolution level, local mutation level, or disordered evolution level corresponding to the target based on the persistence ratio of the dominant evolutionary mode and the number and distribution state of the structural mutation segments; and generating the low-altitude echo target classification results based on the determined levels.

[0111] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for classifying low-altitude echo targets, characterized in that, include: Acquire echo data of targets within the low-altitude detection area during continuous detection cycles; Multiple scattering centers are extracted from the echo data corresponding to each detection period; Based on the positional changes, intensity changes, and phase continuity of each scattering center in adjacent detection periods, a temporal correlation between scattering centers is established. Based on the temporal correlation, the appearance, disappearance, enhancement and weakening states of each scattering center during the target attitude change process are determined, and replacement scattering centers used to take over the disappearing scattering centers are identified. Construct a scattering center replacement chain based on the state change time and succession relationship of each scattering center; The periodicity, spatial symmetry, and state continuity of the replacement sequence are extracted from the replacement chain of the scattering center to generate the evolution characteristics of the scattering structure. The low-altitude echo target classification results are generated based on the scattering structure evolution characteristics.

2. The method according to claim 1, characterized in that, The extraction of multiple scattering centers from the echo data corresponding to each detection period includes: Joint imaging of the echo data in the range and angle dimensions is performed on the echo data of each detection cycle to determine the local echo peak region; Based on the peak intensity, phase concentration, and spatial interval of each local echo peak region, echo peaks belonging to the same scattering location are merged to obtain multiple scattering centers.

3. The method according to claim 1, characterized in that, The step of establishing a temporal correlation between scattering centers based on the positional changes, intensity changes, and phase continuity of each scattering center in adjacent detection periods includes: For each scattering center in the current detection cycle, determine its positional changes relative to each scattering center in adjacent detection cycles; From the scattering centers whose position changes satisfy the preset correlation range, candidate correlated scattering centers are determined based on the direction of intensity change and the degree of phase continuity. When there are multiple candidate associated scattering centers, the candidate associated scattering center with the highest phase continuity is determined as the target associated scattering center; By associating the scattering centers in the current detection cycle with the corresponding target-associated scattering centers, the temporal correlation between the scattering centers is obtained.

4. The method according to claim 1, characterized in that, The identification of replacement scattering centers for receiving disappearing scattering centers includes: Determine the vanishing time and location corresponding to the vanishing scattering center; Candidate replacement scattering centers are determined from the scattering centers that newly appear or increase in intensity after the disappearance time; The replacement scattering center is determined based on the migration direction, intensity complementarity, and phase connection relationship between the candidate replacement scattering center and the disappearing scattering center.

5. The method according to claim 4, characterized in that, The step of determining the replacement scattering center based on the migration direction, intensity complementarity, and phase connection relationship between the candidate replacement scattering center and the vanished scattering center includes: The predicted migration direction is determined based on the position change of the disappearing scattering center before it disappears, and a first candidate scattering center whose appearance direction is consistent with the predicted migration direction is selected from the candidate replacement scattering centers. The intensity decay sequence of the vanishing scattering center is compared with the intensity growth sequence of each first candidate scattering center to determine the second candidate scattering center with an intensity complementary relationship. Based on the phase change trend of the disappearing scattering center before its disappearance, predict its phase at the moment when the candidate replacement scattering center appears, and determine the second candidate scattering center with the highest degree of connection between the actual phase and the predicted phase as the replacement scattering center.

6. The method according to claim 5, characterized in that, The step of comparing the intensity decay sequence of the vanishing scattering center with the intensity growth sequence of each first candidate scattering center to determine the second candidate scattering center with an intensity complementary relationship includes: Determine the state transition time window in which the intensity of the disappearing scattering center decreases, and extract the intensity changes of each first candidate scattering center and its adjacent scattering centers within the state transition time window; The amount of scattering intensity loss is determined from the intensity decay sequence, and the corresponding local scattering intensity carrying capacity is determined based on the intensity increase of each first candidate scattering center and its adjacent scattering centers. Based on the difference between the loss of intensity of vanishing scattering and the amount of intensity absorbed by local scattering, the intensity transfer residual corresponding to each first candidate scattering center is determined; The first candidate scattering center is defined as the first candidate scattering center whose intensity growth start time is later than the intensity decay start time, whose intensity transfer residual satisfies the preset closure condition, and whose local scattering intensity is mainly concentrated in the corresponding first candidate scattering center.

7. The method according to claim 1, characterized in that, The construction of the scattering center replacement chain according to the state change time and succession relationship of each scattering center includes: Each scattering center is identified as a corresponding node; For each disappearing scattering center, determine the emergence time of the corresponding replacement scattering center after its disappearance time, and establish a directed connection from the node corresponding to the disappearing scattering center to the node corresponding to the replacement scattering center; Arrange multiple directed connections consecutively according to the order of their disappearance and appearance times. When the same scattering center reappears in a subsequent detection cycle, the node corresponding to the reappearing scattering center is associated with its original node to obtain the scattering center replacement chain.

8. The method according to claim 1, characterized in that, The extraction of the periodicity, spatial symmetry, and state continuity of the replacement sequence from the scattering center replacement chain includes: Extract the repeating node permutation sequence from the scattering center replacement chain; Calculate the time interval between two consecutive permutations of the same nodes; The periodicity of the replacement order is determined based on the discreteness of the time intervals and the consistency of the node arrangement; Based on the position of each scattering center relative to the target echo center, determine the scattering centers that are distributed in pairs in the replacement chain; Determine whether paired scattering centers alternately increase or decrease in opposite spatial directions to establish spatial symmetry; State continuity is determined by whether there are any unaccompanied disappearing states or appearing states without a preceding source between adjacent nodes.

9. The method according to claim 1, characterized in that, The generation of low-altitude echo target classification results based on the scattering structure evolution characteristics includes: The scattering center replacement chain is segmented according to a continuous time window, and segmented evolution characteristics are generated based on the periodicity, spatial symmetry, and state continuity of the replacement order corresponding to each segment. Calculate the degree of change between the evolutionary characteristics of adjacent segments to determine stable evolutionary segments and structurally abrupt segments; The dominant evolutionary mode is determined based on the repeating scattering structure evolution mode in the stable evolution segment; Based on the persistence ratio of the dominant evolutionary mode and the number and distribution of the structural mutation segments, determine the stable evolution level, local mutation level, or disordered evolution level corresponding to the target. The low-altitude echo target classification results are generated based on the determined levels.

10. A low-altitude echo target classification system, characterized in that, include: The data acquisition module is used to acquire echo data of targets within the low-altitude detection area during continuous detection cycles; The scattering center extraction module is used to extract multiple scattering centers from the echo data corresponding to each detection period; The temporal correlation module is used to establish temporal correlations based on the positional changes, intensity changes, and phase continuity relationships of each scattering center. The replacement identification module is used to determine the appearance, disappearance, enhancement and weakening states of each scattering center during the target attitude change process based on the temporal correlation, and to identify the replacement scattering center used to take over the disappeared scattering center; The replacement chain construction module is used to construct a replacement chain for scattering centers according to the state change time and succession relationship of each scattering center. The feature generation module is used to extract the periodicity, spatial symmetry and state continuity of the replacement sequence from the scattering center replacement chain to generate scattering structure evolution features. The target classification module is used to generate low-altitude echo target classification results based on the evolution characteristics of the scattering structure.

Images