A method and system for extracting slow-moving small targets on the sea surface

Abstract

This invention discloses a method and system for extracting slow-moving small targets on the sea surface, relating to the field of radar target detection. The method includes: acquiring pre-processed data from multiple scanning cycles and performing initial threshold detection to obtain threshold-crossing echoes for each scanning cycle; clustering multiple threshold-crossing echoes to form multiple threshold-crossing echo clusters; recording the number, position, and amplitude of threshold-crossing echoes within each cluster; performing correlation accumulation processing on each threshold-crossing echo cluster; performing multi-cycle joint detection and decision-making on each threshold-crossing echo cluster after correlation accumulation processing; obtaining threshold-crossing echoes from the clusters that pass the decision-making process; obtaining target trace information based on the threshold-crossing echoes from the clusters that pass the decision-making process; and outputting the target trace based on the target trace information to complete target extraction. This method effectively extracts slow-moving small targets on the sea surface and significantly improves the detection capability of slow-moving small targets against sea clutter backgrounds.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of radar target detection, and in particular to a sea-surface slow-speed small target extraction method and system. BACKGROUND

[0002] In the target detection process of a marine radar, sea clutter inevitably affects the detection of sea-surface slow-speed targets such as kayaks, periscopes, and floating mines. Because the scattering cross-sectional area of the sea-surface slow-speed targets is small, the radar echo is weak, and in the time domain, the echo intensity is comparable to that of sea clutter, which cannot be effectively detected under the false alarm condition required by the system. In the frequency domain, because the sea clutter spectrum is wide, and the sea clutter spectrum has a heavy tail phenomenon, the slow-speed targets cannot be separated from the sea clutter spectrum, which brings great challenges to target detection and extraction. SUMMARY

[0003] The present application relates to the field of radar target detection, and in particular to a sea-surface slow-speed small target extraction method and system.

[0004] The technical problems solved by the present application are as follows:

[0005] A sea-surface slow-speed small target extraction method, comprising:

[0006] S1, obtaining pre-processed multi-scan period data;

[0007] S2, performing initial threshold detection on the target data of each scan period to obtain over-threshold echoes of each scan period;

[0008] S3, clustering the over-threshold echoes to form a plurality of over-threshold echo clusters, and recording the number, position, and amplitude of the over-threshold echoes in the over-threshold echo clusters;

[0009] S4, performing correlation accumulation processing on each over-threshold echo cluster;

[0010] S5, performing multi-period joint detection and decision on each over-threshold echo cluster after the correlation accumulation processing to obtain over-threshold echoes in the over-threshold echo clusters that pass the decision;

[0011] S6, obtaining target track information according to the over-threshold echoes in the over-threshold echo clusters that pass the decision;

[0012] S7, outputting a target track according to the target track information to complete target extraction;

[0013] The target track information obtained according to the over-threshold echoes in the over-threshold echo clusters that pass the decision specifically includes:

[0014] When a ratio of a number of over-threshold echoes in a through-judgment over-threshold echo cluster to a total number of over-threshold echoes in the through-judgment over-threshold echo cluster satisfies a preset ratio,

[0015] The over-threshold echoes in the through-judgment over-threshold echo cluster are weighted and condensed by using an amplitude-weighted centroid method, a position identification value and an amplitude value of the condensed over-threshold echoes are calculated, and target track information is formed according to the position identification value and the amplitude value.

[0016] The over-threshold echoes in the through-judgment over-threshold echo cluster specifically include the following steps:

[0017] It is judged whether the accumulated amplitude value of each threshold echo cluster is greater than a first preset value;

[0018] And whether the position accumulated identification of the threshold echo cluster is greater than a second preset value, if both are satisfied, the over-threshold echoes in the through-judgment over-threshold echo cluster are obtained.

[0019] The associated accumulation processing on each over-threshold echo cluster specifically includes the following steps:

[0020] Each over-threshold echo cluster is processed by a preset associated accumulation method to obtain an accumulated amplitude value and a position accumulated identification of each over-threshold echo cluster.

[0021] The beneficial effects of the present application are that: the present scheme firstly performs initial low threshold detection on each scanning period data, ensures target detection probability while allowing high false alarm, suppresses over-threshold sea clutter through multi-period joint detection judgment, finally retains slow moving targets, realizes effectiveness of the sea surface slow small target extraction method, and effectively improves slow small target detection capability in sea clutter background.

[0022] Further, it further includes the following steps:

[0023] A threshold echo cluster echo number threshold value is set;

[0024] Threshold echo clusters with an echo number less than the echo number threshold value are deleted to obtain a plurality of screened over-threshold echo clusters;

[0025] The S4 specifically includes the following steps:

[0026] Each screened over-threshold echo cluster is subjected to associated accumulation processing.

[0027] The beneficial effects of the above further scheme are that: by setting the threshold echo cluster echo number threshold value, interference data that does not meet the requirements is effectively eliminated.

[0028] The advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0029] Figure 1 A flowchart illustrating a method for extracting slow-moving small targets on the sea surface, provided as an embodiment of the present invention;

[0030] Figure 2 The initial low-threshold detection processing result of a scan cycle measured data is provided for other embodiments of the present invention;

[0031] Figure 3 A schematic diagram of threshold echo clustering in a local region of measured data during a scanning cycle, provided for other embodiments of the present invention;

[0032] Figure 4 The amplitude correlation accumulation processing result between scans of measured data from multiple consecutive scan cycles is provided for other embodiments of the present invention;

[0033] Figure 5 This provides the result of position accumulation and association accumulation processing of measured data across multiple consecutive scanning cycles, as shown in other embodiments of the present invention.

[0034] Figure 6 The measured data of the Kth scan cycle provided for other embodiments of the present invention are processed by multi-cycle joint detection and decision-making.

[0035] Figure 7 The target extraction processing result of the measured data in the Kth scan cycle provided for other embodiments of the present invention. Detailed Implementation

[0036] The principles and features of the present invention are described below with reference to the accompanying drawings. The embodiments described are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0037] like Figure 1 As shown, an embodiment of the present invention provides a method for extracting slow-moving small targets on the sea surface, comprising:

[0038] S1, acquire the preprocessed data from multiple scanning cycles; in one embodiment, the target data may include: data extracted from the actual data of the sea radar with a size of 200 range units × 200 azimuth units, and one target in the data is a small fishing boat.

[0039] In one embodiment, S1 may include: sequentially acquiring modulus data Scan after preprocessing from multiple consecutive scan cycles. k k = 1, ..., K, where K is the number of joint processing cycles.

[0040] In this embodiment of the invention, the measured data obtained is preprocessed data after K=6 consecutive scan cycles, with a data size of 200 range units × 200 azimuth units. The preprocessing includes echo signal pulse compression and non-coherent accumulation between adjacent pulses. S2, initial threshold detection is performed on the target data for each scan cycle to obtain the threshold-crossing echo for each scan cycle;

[0041] In one embodiment, S2 may include: modulo data Scan for each scan cycle. k Initial threshold detection is performed sequentially along the distance dimension using low-threshold CA-CFAR to obtain over-threshold echo data (DataCFAR). k , k=1,…,K.

[0042] The initial low-threshold detection in this embodiment uses the Cell Average Constant False Alarm Rate (CA-CFAR) detection algorithm, a mature algorithm in the field of radar target detection. In this embodiment, the low-threshold detection coefficient is set to a false alarm rate of P. f =10 -3 The corresponding threshold coefficient The protection unit is set to 3, and the reference unit is set to 16. The threshold echo after the data from the first scan cycle passes through the low-threshold detection is shown below. Figure 2 As shown in the figure, one target in the data is an experimental small fishing boat, and the other threshold echoes are all sea clutter. It can be seen from the figure that the amplitude of the sea clutter is close to the amplitude of the target. If high-threshold detection is used to suppress the sea clutter, the target will also be suppressed simultaneously. S3, multiple threshold echoes are clustered to form multiple threshold echo clusters, and the number, position, and amplitude of the threshold echoes within each threshold echo cluster are recorded.

[0043] In one embodiment, S3 may include: for each scan cycle, the threshold echo data DataCFAR k The threshold echo units are clustered according to their nearest neighbor connectivity to form threshold echo clusters. Each threshold echo cluster records the number, location, and amplitude of threshold echoes within it, where the location is the distance-azimuth unit number of the threshold echo. In another embodiment, the method further includes setting a threshold for detecting the number of echoes in a threshold echo cluster as Thr. EPNum The number of deleted echoes is less than Thr EPNum The threshold echo clusters are obtained by collecting the threshold echo cluster data EPGroup for each scan cycle. k , k=1,…,K.

[0044] A schematic diagram of threshold echo clustering is shown below. Figure 3As shown, adjacent threshold echoes are grouped into threshold echo clusters, each cluster being identified by an elliptical line. In this embodiment, the threshold for detecting the number of threshold echoes is set to Thr. EPNum =10, echoes that do not meet the threshold are marked with dashed lines, and echoes that meet the threshold are marked with solid lines.

[0045] S4, perform correlation accumulation processing on each threshold echo cluster;

[0046] In one embodiment, the association accumulation process may include: according to a set maximum speed V max Calculate the maximum displacement r of the target motion between scans. max =V max T scan T scan For each threshold echo cluster EPGroup in the k-th period, the scan period is defined as follows: k (j), j = 1, ..., EPGroupNum k EPGroupNum k The number of threshold echo clusters in the k-th period, denoted by EPGroup. k (j) The origin is the center of the threshold echo position, and r max The search radius is used to check if a threshold echo cluster exists in the (k-1)th scan cycle. If it does, then EPGroup is added. k Accumulate the amplitude at the corresponding position in (j) and increment the accumulation flag at that position by 1. If there is no accumulation flag, increment it by 2r. max The search radius is used to check if a threshold echo cluster exists in the (k-2)th scan period. If it does, the EPGroup is added. k (j) Accumulate the amplitude at the corresponding position and increment the accumulation flag at that position by 1. If there is no accumulation flag, then the threshold echo cluster EPGroup is entered. k (j) The position accumulation flag is set to 1, and the next threshold echo cluster EPGroup is processed. k (j+1).

[0047] S5, perform multi-cycle joint detection and decision on each threshold echo cluster after correlation accumulation processing, and obtain the threshold echoes in the threshold echo clusters that pass the decision.

[0048] In one embodiment, the multi-cycle joint detection and decision process may include: setting an amplitude accumulation detection threshold Thr. amp and location accumulation identifier detection threshold Thr N After processing K consecutive cycles of data according to the above steps, for each threshold echo cluster EPGroup in the Kth cycle...K (i) Perform joint detection and decision based on the accumulated results over multiple cycles. If the accumulation magnitude is greater than Thr... amp And the location accumulation identifier is greater than Thr N If the threshold echo accumulation amplitude of the range-azimuth cell is reached, the range-azimuth cell will output 0; otherwise, the range-azimuth cell will output 0, where i = 1, ..., EPGroupNum. K EPGroupNum K This represents the number of threshold echo clusters in the Kth period.

[0049] S6, Obtain target point information based on the threshold echoes in the threshold echo cluster that have passed the judgment;

[0050] S7 outputs the target trace based on the target trace information, completing target extraction. Specifically, target extraction involves extracting the target small fishing boat from the marine radar's measured data.

[0051] In one embodiment, the target extraction process may include: counting the number of threshold echoes that meet the joint detection decision in each threshold echo cluster; when the ratio of the number of threshold echoes after joint detection decision of the echo cluster to the total number of threshold echoes in the echo cluster exceeds a set ratio, using the amplitude-weighted centroid method to perform weighted aggregation on the threshold echoes that meet the joint detection decision, calculating the position and amplitude after aggregation, forming point trace information, and outputting the target point trace; if the set ratio is not met, the echo cluster is directly deleted.

[0052] For each threshold echo cluster EPGroup K (j), the number of threshold echoes that meet the joint detection decision conditions in this data cluster is EP. num_detect The ratio of the number of threshold echoes to the total number of threshold echoes in the data cluster is calculated as r. EP =EP num_detect / EP num Set the proportional detection threshold as like Then, based on the threshold echoes that meet the joint detection decision conditions, the centroid (R) of the data cluster is calculated. j A j The calculation method is the same as above. The target location information is obtained and output. If the proportional detection threshold is not met, the echo cluster is directly deleted.

[0053] In this embodiment of the invention, the detection threshold for the proportion of echoes exceeding the threshold that meet the joint detection decision conditions is set as follows: The target extraction result for the K=6th scan cycle is as follows: Figure 7 As shown, the measured data processing results verify the effectiveness of the method for extracting slow-moving small targets on the sea surface using the inter-scan joint detection method of the present invention.

[0054] This scheme first performs initial low-threshold detection on the data of each scanning cycle to ensure the target detection probability while allowing for high false alarms. Through multi-cycle joint detection and decision, it suppresses sea clutter that exceeds the threshold and ultimately retains slow-moving targets, thus realizing the effectiveness of the method for extracting slow-moving small targets on the sea surface and effectively improving the detection capability of slow-moving small targets in the background of sea clutter.

[0055] In another embodiment, actual measurement data from a certain type of shore-based maritime radar was used for verification. The data was extracted in a size of 200 range cells × 200 azimuth cells. One target in the data was a small fishing boat. In one scan cycle, the measured data was detected with a low threshold, and the result was as follows. Figure 2 As shown, the joint detection and decision results of measured data from multiple scanning cycles are as follows: Figure 6 As shown, comparison Figure 2 and Figure 6 It can be seen that using low-threshold detection can ensure that the target is detected. However, there are a large number of sea clutter exceeding the threshold. After using the method of this invention for multi-cycle joint detection and decision-making, the sea clutter exceeding the threshold is effectively suppressed, and slow-moving targets are preserved. The experimental data processing results verify the effectiveness of the inter-scan joint detection and decision-making method for extracting slow-moving small targets on the sea surface in this invention.

[0056] The specific steps of obtaining target point information based on threshold echoes in the threshold echo cluster determined by the criteria include:

[0057] When the ratio of the number of threshold-crossing echoes in the threshold-crossing echo cluster that passed the judgment to the total number of threshold-crossing echoes in the threshold-crossing echo cluster meets a preset ratio,

[0058] The amplitude-weighted centroid method is used to perform weighted aggregation of threshold echoes in the threshold echo cluster that has passed the decision. The position identifier value and amplitude value of the aggregated threshold echo are calculated, and target point trace information is formed based on the position identifier value and amplitude value.

[0059] The step of performing multi-cycle joint detection and decision-making on each threshold-crossing echo cluster after correlation accumulation processing to obtain the threshold-crossing echoes in the threshold-crossing echo clusters that pass the decision specifically includes:

[0060] Determine whether the accumulated amplitude value of each threshold echo cluster is greater than the first preset value;

[0061] And whether the position accumulation flag of the threshold echo cluster is greater than the second preset value. If both are satisfied, then the threshold echo in the threshold echo cluster that passed the decision is obtained.

[0062] The specific steps for performing correlation and accumulation processing on each threshold echo cluster include:

[0063] Each threshold echo cluster is processed using a pre-defined correlation accumulation method to obtain the accumulated amplitude value and position accumulation identifier of each threshold echo cluster.

[0064] Preferably, in any of the above embodiments, it further includes:

[0065] Set the threshold value for the number of echoes in the threshold echo cluster;

[0066] Delete threshold echo clusters whose number of echoes is less than the echo count threshold, and obtain multiple threshold echo clusters after filtering;

[0067] In one embodiment, the threshold for detecting the number of echo cluster echoes is set to Thr. EPNum The number of deleted echoes is less than Thr EPNum The threshold echo clusters are obtained by collecting the threshold echo cluster data EPGroup for each scan cycle. k , ...

[0068] S4 specifically includes:

[0069] Each filtered threshold echo cluster is subjected to correlation and accumulation processing.

[0070] By setting a threshold value for the number of echoes in a threshold echo cluster, interference data that does not meet the requirements can be effectively eliminated.

[0071] Preferably, in any of the above embodiments, obtaining the target point information based on the threshold echoes in the threshold echo cluster that have passed the decision specifically includes:

[0072] When the ratio of the number of threshold-crossing echoes in the threshold-crossing echo cluster that passed the judgment to the total number of threshold-crossing echoes in the threshold-crossing echo cluster meets a preset ratio,

[0073] The amplitude-weighted centroid method is used to perform weighted aggregation of threshold echoes in the threshold echo cluster that has passed the decision. The position identifier value and amplitude value of the aggregated threshold echo are calculated, and the target point trace information is formed based on the position identifier value and amplitude value.

[0074] Preferably, in any of the above embodiments, the step of performing multi-cycle joint detection and decision on each threshold-crossing echo cluster after correlation accumulation processing to obtain the threshold-crossing echoes in the threshold-crossing echo clusters that have passed the decision specifically includes:

[0075] Determine whether the accumulated amplitude value of each threshold echo cluster is greater than the first preset value;

[0076] Furthermore, if the position accumulation flag of the threshold echo cluster is greater than the second preset value, then the threshold echo in the threshold echo cluster that has passed the decision is obtained.

[0077] This scheme effectively suppresses sea clutter beyond the threshold by using multi-cycle joint detection and decision-making, thus separating slow-moving targets from the sea clutter spectrum and preserving slow-moving targets.

[0078] Preferably, in any of the above embodiments, the correlation accumulation processing for each threshold echo cluster specifically includes:

[0079] Each threshold echo cluster is processed using a pre-defined correlation accumulation method to obtain the accumulated amplitude value and position accumulation identifier of each threshold echo cluster.

[0080] In one embodiment, the method for extracting slow-moving small targets on the sea surface through joint inter-scan detection includes:

[0081] The embodiments use measured data from a certain type of shore-based sea radar for verification. The implementation of the method of the present invention is specifically described in conjunction with flowcharts and embodiments. The processing procedure includes the following steps:

[0082] Step 11: Data acquisition across multiple scan cycles;

[0083] The preprocessed and modulated data from multiple consecutive radar scan cycles are acquired sequentially and recorded as DataScan. k DataScan k For N range ×N azimuth A data matrix, where N range N represents the number of distance cells. azimuth The number of azimuth units is k = 1, ..., K, where K is the number of joint detection and processing cycles.

[0084] The measured data obtained in this embodiment of the invention are preprocessed modulus data after K=6 consecutive scanning cycles, with a data size of 200 distance units × 200 azimuth units. The preprocessing includes echo signal pulse compression and non-coherent accumulation between adjacent pulses.

[0085] Step 12: Initial low threshold detection;

[0086] DataScan is used to extract modulo data for each scan cycle. k The cell-average constant false alarm rate (CFAR) detection algorithm is used to scan the distance dimension data in each azimuth cell. k (:,n) Perform initial low-threshold detection sequentially along the distance dimension, n=1,…,N azimuth After the range dimension data on all azimuth units has been detected, the threshold echo for that scan period is obtained and recorded as DataCFAR. k k = 1, ..., K, DataCFAR k For N range ×N azimuthThe data matrix;

[0087] The initial low-threshold detection in this embodiment uses the Cell Average Constant False Alarm Rate (CA-CFAR) detection algorithm, a mature algorithm in the field of radar target detection. In this embodiment, the low-threshold detection coefficient is set to a false alarm rate of P. f =10 -3 The corresponding threshold coefficient The protection unit is set to 3, and the reference unit is set to 16. The threshold echo after the data from the first scan cycle passes through the low-threshold detection is shown below. Figure 2 As shown in the figure, one target in the data is the experimental small fishing boat, and the other threshold echoes are all sea clutter. It can be seen from the figure that the amplitude of the sea clutter is close to the amplitude of the target. If a high threshold detection is used to suppress the sea clutter, the target will also be suppressed at the same time.

[0088] Step 13: Threshold echo clustering;

[0089] For each scan cycle, the threshold echo is a DataCFAR. k Based on the nearest neighbor connectivity of the distance-azimuth unit, threshold echoes are clustered to form threshold echo clusters. Each threshold echo cluster records the number of threshold echoes within the cluster, their location in the distance-azimuth unit, and their amplitude. The threshold for detecting the number of threshold echoes is set to Thr. EPNum Delete threshold echoes; the number of threshold echoes is less than Thr. EPNum The threshold echo clusters are obtained for each scan cycle, resulting in the threshold echo cluster EPGroup. k k = 1, ..., K;

[0090] A schematic diagram of threshold echo clustering is shown below. Figure 3 As shown in the diagram (a clustering diagram of threshold echoes in a local area of ​​measured data for one scan cycle), threshold echoes that are adjacent in position are formed into threshold echo clusters. Each threshold echo cluster is identified by an elliptical line. In this embodiment of the invention, the threshold for detecting the number of threshold echoes is set to Thr. EPNum =10, echoes that do not meet the threshold are marked with dashed lines, and echoes that meet the threshold are marked with solid lines.

[0091] Step 14: Accumulate correlations between scans;

[0092] For each threshold echo cluster (EPGroup) in the k-th periodk (j), j = 1, ..., EPGroupNum k EPGroupNum k For the threshold echo cluster in the k-th period, calculate the centroid (R) of the data cluster. j A j )for

[0093]

[0094]

[0095] Among them, EP num For data clusters EPGroup k (j) is the number of threshold echoes, (R(i),A(i)) is the distance-azimuth cell where the threshold echo is located, and Amp(R(i),A(i)) is the amplitude of the threshold echo.

[0096] The maximum displacement of the target between two adjacent scans is set as (r max ,a max ), with (R j A j (R) is the origin of the position, and (R) is the origin of the position. j -r max :R j +r max A j -a max :A j +a max The search range is defined as follows: In the (k-1)th scan period, the search is performed to determine if a threshold echo or threshold echo cluster (EPGroup) exists. k-1 (l) If it is within the search range, then EPGroup k (j) and EPGroup k-1 (l) sums and accumulates the amplitudes of the same distance-azimuth units, and increments the accumulation flag Sign(R(i),A(i)) of the corresponding distance-azimuth unit (R(i),A(i)) by 1, i.e.

[0097] {Amp(R(i),A(i))∈EPGroup k (j)}={Amp(R(i),A(i))∈EPGroup k (j)}+{Amp(R(i),A(i))∈EPGroup k-1 (l)}

[0098] {Sign(R(i),A(i))∈EPGroup k(j)}={Sign(R(i),A(i))∈EPGroup k-1 (l)}+1

[0099] If no threshold echo cluster is found in the (k-1)th scan cycle, then 2(R) j -r max :R j +r max A j -a max :A j +a max The search range is defined as follows: During the (k-2)th scan period, the system searches for the existence of threshold echo clusters. If any are found, the EPGroup is selected. k (j) and EPGroup k-2 (l) The amplitudes of the same range-azimuth units are summed and accumulated, and the accumulation flag of the corresponding range-azimuth unit (R(i), A(i)) is incremented by 1. If there is no accumulation flag, the threshold echo cluster EPGroup is then saved. k (j) The position accumulation flag is set to 1, i.e., {Sign(R(i),A(i))∈EPGroup k (j)}=1, continue processing the next threshold echo cluster EPGroup k (j+1);

[0100] In this embodiment of the invention, the maximum displacement of the target between two adjacent scans is set to (r max ,a max The amplitude correlation accumulation processing result of measured data between scans over K=6 consecutive scan cycles is as follows: )=(2,1), Figure 4 As shown, the location accumulation identifier associated accumulation processing result is as follows: Figure 5 As shown;

[0101] Step 15: Multi-cycle joint detection and decision;

[0102] Set the cumulative amplitude detection threshold Thr amp and location accumulation identifier detection threshold Thr N After processing K consecutive cycles of data according to steps 11 to 14, for each threshold echo cluster (EPGroup) in the Kth cycle... K (j) Perform joint detection and decision-making based on multi-cycle accumulation results; the decision-making process is as follows: for threshold echo clusters (EPGroup)... K For each distance-azimuth cell (R(i), A(i)) in (j), if the accumulated amplitude value {Amp(R(i), A(i)) ∈ EPGroup} K (j)} is greater than Thr ampAnd the position accumulation identifier {Sign(R(i),A(i))∈EPGroup k (j)} is greater than Thr N If the threshold echo is reached, the accumulated amplitude value of the range-azimuth unit will be output; otherwise, the range-azimuth unit will output 0.

[0103] In this embodiment of the invention, the accumulated amplitude detection threshold is set as follows:

[0104] Thr amp =0.5*mean({Amp(R(i),A(i))∈EPGroup K (j)}),

[0105] mean(·) represents averaging the accumulated amplitude values ​​of all cells in the threshold echo cluster, with the position accumulation flag detection threshold set to Thr. N =K-2=4, the measured data of the K=6th scan cycle after multi-cycle joint detection and decision processing is as follows: Figure 6 As shown, comparison Figure 2 and Figure 6 It can be seen that after adopting multi-cycle joint detection and decision, the data of sea clutter across the threshold is significantly reduced, and slow-moving targets are preserved;

[0106] Step 16: Target extraction and output;

[0107] For each threshold echo cluster EPGroup K (j), the number of threshold echoes in the data cluster that meet the joint detection decision condition in step 15 is EP. num_detect The ratio of the number of threshold echoes to the total number of threshold echoes in the data cluster is calculated as r. EP =EP num_detect / EP num Set the proportional detection threshold as like Then, based on the threshold echoes that meet the joint detection decision conditions, the centroid (R) of the data cluster is calculated. j A j The calculation method is the same as in step 14. The target location information is obtained and output. If the proportional detection threshold is not met, the echo cluster is directly deleted.

[0108] In this embodiment of the invention, the detection threshold for the proportion of echoes exceeding the threshold that meet the joint detection decision conditions is set as follows: The target extraction result for the K=6th scan cycle is as follows: Figure 7 As shown, the measured data processing results verify the effectiveness of the method for extracting slow-moving small targets on the sea surface using the inter-scan joint detection method of the present invention.

[0109] In another embodiment, a method for extracting slow-moving small targets on the sea surface through joint detection between scans includes the following steps:

[0110] Step 111: Data acquisition across multiple scan cycles;

[0111] The preprocessed and modulated data from multiple consecutive radar scan cycles are acquired sequentially and recorded as DataScan. k DataScan k For N range ×N azimuth A data matrix, where N range N represents the number of distance cells. azimuth The number of azimuth units is k = 1, ..., K, where K is the number of joint detection and processing cycles.

[0112] Step 112: Initial low threshold detection;

[0113] DataScan is used to extract modulo data for each scan cycle. k The cell-average constant false alarm rate (CFAR) detection algorithm is used to scan the distance dimension data in each azimuth cell. k (:,n) Perform initial low-threshold detection sequentially along the distance dimension, n=1,…,N azimuth After the range dimension data on all azimuth units has been detected, the threshold echo for that scan period is obtained and recorded as DataCFAR. k k = 1, ..., K, DataCFAR k For N range ×N azimuth The data matrix;

[0114] Step 113: Threshold echo clustering;

[0115] For each scan cycle, the threshold echo is a DataCFAR. k Based on the nearest neighbor connectivity of the distance-azimuth unit, threshold echoes are clustered to form threshold echo clusters. Each threshold echo cluster records the number of threshold echoes within the cluster, their location in the distance-azimuth unit, and their amplitude. The threshold for detecting the number of threshold echoes is set to Thr. EPNum Delete threshold echoes; the number of threshold echoes is less than Thr. EPNum The threshold echo clusters are obtained for each scan cycle, resulting in the threshold echo cluster EPGroup. k k = 1, ..., K;

[0116] Step 114: Accumulate correlations between scans;

[0117] For each threshold echo cluster (EPGroup) in the k-th period k (j), j = 1, ..., EPGroupNum k EPGroupNum k For the threshold echo cluster in the k-th period, calculate the centroid (R) of the data cluster. j A j )for

[0118]

[0119]

[0120] Among them, EP num For data clusters EPGroup k (j) is the number of threshold echoes, (R(i),A(i)) is the distance-azimuth cell where the threshold echo is located, and Amp(R(i),A(i)) is the amplitude of the threshold echo;

[0121] The maximum displacement of the target between two adjacent scans is set as (r) max ,a max ), with (R j A j (R) is the origin of the position, and (R) is the origin of the position. j -r max :R j +r max A j -a max :A j +a max The search range is defined as follows: In the (k-1)th scan period, the search is performed to determine if a threshold echo cluster (EPGroup) exists. k-1 (l) If it is within the search range, then EPGroup k (j) and EPGroup k-1 (l) The amplitudes of the same distance-azimuth units are added together and accumulated, and the accumulation flag Sign(R(i),A(i)) of the corresponding distance-azimuth unit (R(i),A(i)) is incremented by 1, that is...

[0122] {Amp(R(i),A(i))∈EPGroup k (j)}={Amp(R(i),A(i))∈EPGroup k (j)}+{Amp(R(i),A(i))∈EPGroup k-1 (l)}

[0123] {Sign(R(i),A(i))∈EPGroup k (j)}={Sign(R(i),A(i))∈EPGroup k-1 (l)}+1

[0124] If no threshold echo cluster is found in the (k-1)th scan cycle, then 2(R j -r max :R j +r max A j -a max :A j +a max The search range is defined as follows: During the (k-2)th scan period, the system searches for the existence of a threshold echo cluster. If found, the EPGroup is selected. k (j) and EPGroup k-2 (l) The amplitudes of the same range-azimuth units are summed and accumulated, and the accumulation flag of the corresponding range-azimuth unit (R(i), A(i)) is incremented by 1. If there is no accumulation flag, the threshold echo cluster EPGroup is then saved. k (j) The position accumulation flag is set to 1, i.e., {Sign(R(i),A(i))∈EPGroup k (j)}=1, continue processing the next threshold echo cluster EPGroup k (j+1);

[0125] Step 115: Multi-cycle joint detection and decision;

[0126] Set the cumulative amplitude detection threshold Thr amp and location accumulation identifier detection threshold Thr N After K consecutive cycles of data have been processed according to steps 111 to 114, for each threshold echo cluster EPGroup in the Kth cycle... K (j) Perform joint detection and decision-making based on multi-cycle accumulation results; the decision-making process is as follows: for the threshold echo cluster EPGroup K For each distance-azimuth cell (R(i), A(i)) in (j), if the accumulated amplitude value {Amp(R(i), A(i)) ∈ EPGroup} K (j)} is greater than Thr amp And the position accumulation identifier {Sign(R(i),A(i))∈EPGroup k (j)} is greater than Thr N If the threshold echo accumulation amplitude of the range-azimuth unit is reached, the range-azimuth unit will output the value; otherwise, the range-azimuth unit will output 0.

[0127] Step 116: Target extraction and output;

[0128] For each threshold echo cluster EPGroup K (j), count the number of threshold echoes in the data cluster that meet the joint detection decision condition in step 115 as EP. num_detect The ratio of the number of threshold echoes to the total number of echoes in the data cluster is calculated as r. EP =EP num_detect / EP num Set the proportional detection threshold as like Then, the centroid (R) of the data cluster is calculated based on the threshold echo that meets the joint detection decision conditions. j A j The calculation method is the same as in step 114. The target location information is obtained and output. If the proportional detection threshold is not met, the echo cluster is directly deleted. Readers should understand that in the description of this specification, the reference to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., means that the specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0129] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the method embodiments described above are merely illustrative. For instance, the division of steps is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple steps may be combined or integrated into another step, or some features may be ignored or not executed.

[0130] If the above methods are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0131] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for extracting slow-moving small targets on the sea surface, characterized in that, include: S1, acquire the preprocessed data from multiple scan cycles; S2, perform initial threshold detection on the target data for each scan cycle to obtain the threshold echo for each scan cycle; S3, perform threshold echo clustering on multiple threshold echoes to form multiple threshold echo clusters, and record the number, position and amplitude of the threshold echoes obtained in the threshold echo clusters; S4, perform correlation accumulation processing on each threshold echo cluster; S5, perform multi-cycle joint detection and decision on each threshold echo cluster after correlation accumulation processing, and obtain the threshold echoes in the threshold echo clusters that pass the decision. S6, Obtain target point information based on the threshold echoes in the threshold echo cluster that have passed the judgment; S7, output the target point trace based on the target point trace information to complete the target extraction; The specific steps of obtaining target point information based on threshold echoes in the threshold echo cluster determined by the criteria include: When the ratio of the number of threshold-crossing echoes in the threshold-crossing echo cluster that passed the judgment to the total number of threshold-crossing echoes in the threshold-crossing echo cluster meets a preset ratio, The amplitude-weighted centroid method is used to perform weighted aggregation of threshold echoes in the threshold echo cluster that has passed the decision. The position identifier value and amplitude value of the aggregated threshold echo are calculated, and target point trace information is formed based on the position identifier value and amplitude value. The step of performing multi-cycle joint detection and decision-making on each threshold-crossing echo cluster after correlation accumulation processing to obtain the threshold-crossing echoes in the threshold-crossing echo clusters that pass the decision specifically includes: Determine whether the accumulated amplitude value of each threshold echo cluster is greater than the first preset value; And whether the position accumulation flag of the threshold echo cluster is greater than the second preset value. If both are satisfied, then the threshold echo in the threshold echo cluster that passed the decision is obtained. The specific steps for performing correlation and accumulation processing on each threshold echo cluster include: Each threshold echo cluster is processed using a pre-defined correlation accumulation method to obtain the accumulated amplitude value and position accumulation identifier of each threshold echo cluster.

2. The method for extracting slow-moving small targets on the sea surface according to claim 1, characterized in that, Also includes: Set the threshold value for the number of echoes in the threshold echo cluster; Delete threshold echo clusters whose number of echoes is less than the echo count threshold, and obtain multiple threshold echo clusters after filtering; S4 specifically includes: Each filtered threshold echo cluster is subjected to correlation and accumulation processing.

Images