False alarm rate dynamic adjustment method for maritime target feature detection

By using convex hull detectors and SVM detectors in the feature detection of maritime targets, combined with the initial virtual alarm rate control method, dynamically adjusting the judgment area, solving the problem of false alarm rate adjustment caused by changes in sea clutter characteristics, and achieving small The effective false alarm rate control at the cost of time significantly improves the false alarm control performance of the detector.

CN119064884AActive Publication Date: 2024-12-03NAVAL AVIATION UNIV
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411241582.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-05
Publication Date
2024-12-03
Estimated Expiration
2044-09-05

AI Technical Summary

Technical Problem

In the detection of maritime target feature, false alarm rate control is difficult to adapt to the space-time non-Gaussian and non-stationary characteristics of sea clutter, resulting in high cost of adjusting false alarm rate and is not conducive to real-time performance in the detection process.

Method used

Using convex hull detector and SVM detector, by extracting features from the time and frequency domains of radar echo signals, combined with the initial false alarm rate control method, the judgment area is dynamically adjusted to reduce the preset false alarm rate and actual false alarm rate. The difference between the two realizes online adjustment of false alarm rate.

Benefits of technology

It realizes effective control of false alarm rate at the cost of small time, improves the detector's false alarm control ability, is suitable for a variety of feature detection methods, and significantly improves the false alarm control performance of the detector through HH and VV polarization verification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119064884A_ABST
    Figure CN119064884A_ABST
Patent Text Reader

Abstract

The invention relates to a false alarm rate dynamic adjustment method for maritime target feature detection, and belongs to the technical field of radar target detection. Comprising the steps of 1) extracting features from a time domain and a frequency domain of a radar echo signal; 2) training a convex hull detector and an SVM (Support Vector Machine) detector according to the extracted clutter features and target echo features, and preliminarily controlling a false alarm rate; 3) performing detection work according to the input echo data, counting the statistical alarm rate of a period of time every time, taking the statistical alarm rate as the estimation of an actual false alarm rate, determining the adjustment amplitude of a judgment area according to the actual false alarm rate and a preset false alarm rate, and performing re-adjustment of the false alarm rate; the method is applied to a convex hull detector and an SVM (Support Vector Machine) detector to form a target detection method, so that the false alarm control capability of the detector is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for dynamically adjusting false alarm rate of marine target feature detection, belonging to the technical field of radar target detection. Background Art

[0002] False alarm rate control has always been an important issue in the field of radar detection. Feature detection has better detection performance than traditional detection methods, but features are different from amplitudes, and it is impossible to give the optimal detection solution from the perspective of probability density. The false alarm rate control method for feature detectors also remains at the control of the initial false alarm rate. In feature detection, due to the quantitative differences between clutter data and target data, detection problems are often classified as anomaly detection, and the false alarm rate control link of the detector can also be regarded as a part of the recognition of sea clutter characteristics. As we all know, the space-time non-Gaussian and non-stationary characteristics of sea clutter are the fundamental reason for the difficulty in recognizing sea clutter characteristics, which means that to achieve true constant false alarm detection, it is necessary to perceive the sea clutter characteristics that change with space and time in large scenes in real time.

[0003] The current feature detection false alarm rate control method focuses on keeping the false alarm rate of the detector at a preset false alarm rate level within the range of training clutter data. The spatial probability distribution of sea clutter features is the same as the spatial probability of sea clutter features used when training the detector, which is the basic condition for this type of false alarm rate control method to still take effect when receiving clutter data other than training clutter data. However, since sea clutter has spatial and temporal non-stationarity and non-uniformity, and the instability of non-stationarity and non-uniformity in the spatial dimension is often greater than that in the temporal dimension, even if enough clutter data is collected from enough distance resolution units, it is impossible to accurately estimate the spatial probability density of the current clutter with time-varying non-stationarity, and it can only make the clutter training sample set contain as many feature distribution changes caused by clutter non-stationarity as possible. Therefore, increasing the number of training samples of the feature detector not only increases the time required for a single detection, but is also not conducive to false alarm rate control during the detection process.

[0004] Therefore, there is an urgent need for an online adjustment method of false alarm rate with low time cost, which can re-plan the judgment area according to the real-time sea clutter characteristics to cope with the sea clutter characteristics that change with time and space, which is very important for the control of false alarm rate. Summary of the invention

[0005] In order to solve the deficiencies of the above-mentioned prior art, a method for dynamically adjusting the false alarm rate of marine target feature detection is provided. The method is applied to a convex hull detector and a SVM detector to form a target detection method, thereby improving the false alarm control capability of the detector.

[0006] The method for dynamically adjusting the false alarm rate of marine target feature detection of the present invention is characterized in that it comprises the following steps:

[0007] Step 1: Extract features from the time domain and frequency domain of radar echo signals;

[0008] Step 2: Train the convex hull detector and the SVM detector according to the extracted clutter features and target echo features, and use the initial false alarm rate control method to preliminarily control the false alarm rate according to the preset false alarm rate value;

[0009] Step 3: Perform detection according to the input echo data, and count the total alarm rate of each period of time as an estimate of the actual false alarm rate. Determine the adjustment range of the judgment area according to the actual false alarm rate and the preset false alarm rate, and readjust the false alarm rate to achieve the purpose of reducing the difference with the preset false alarm rate;

[0010] Step 4: Return to step 3 and continue the test until the test is completed;

[0011] Step 5: Use measured data to verify the proposed method. Take the channel buoy as the small sea surface target to be detected, and verify the performance of the proposed method under two polarization modes: HH and VV (HH polarization means that the electromagnetic waves transmitted and received by the radar are both horizontally polarized waves, and VV polarization means that the electromagnetic waves transmitted and received by the radar are both vertically polarized waves).

[0012] Preferably, the features extracted in step 1 are from the time domain and the frequency domain. The following are specific extraction methods of three typical features:

[0013] 1) Relative average amplitude

[0014] The relative average amplitude is defined as the ratio of the time domain average amplitude of the unit to be detected within the number of coherent pulses to the time domain average amplitude of the reference unit, which is used to reflect the difference between the sea clutter echo and the target echo. The relative average amplitude of the sea clutter unit is 1. Since the amplitude of the target echo is larger than that of the sea clutter echo, the relative average amplitude of the target unit is often greater than that of the sea clutter unit.

[0015]

[0016] In the formula, represents the echo vector of the unit to be detected with a length of N, Represents an echo sequence with a length of N to be detected. Refers to the echo sequence subscript, Refers to the average amplitude of the echo of the unit to be detected. represents an echo sequence with a reference unit length of N, Refers to the average amplitude of the echo of the reference unit, P is the number of reference units, Represents the relative average amplitude characteristics of the unit to be detected;

[0017] (2) Relative Doppler peak height

[0018] The Doppler spectrum is calculated as follows:

[0019]

[0020] In the formula, refers to the Doppler frequency, refers to the pulse repetition period, Refers to the echo sequence subscript, Refers to the echo sequence length, Represents an echo sequence with a length of N to be detected. Refers to the Doppler spectrum of the echo sequence, based on which the Doppler peak height and peak height location are extracted;

[0021] use , The reference Doppler unit is defined as Refers to the width of the Doppler main peak, Refers to the Doppler band unit for reference;

[0022] D = [ − d 1 , − d 2 ] ∪ [ d 2 , d 1 ] RDPH ( x ) = max ( X ( f d )) 1 # D ∑ f d ∈ f d max ( x ) + D X ( f d )

[0023] refers to the Doppler frequency, refer to The frequency corresponding to the maximum value is, Refers to the Doppler frequency range available for reference, Fingers fall into The number of Doppler bins in the frequency bin, , Refers to the frequency range for estimating the relative Doppler peak height, represents the echo vector of the unit to be detected with a length of N, Refers to the echo vector of the unit to be detected After the reference unit is introduced, the relative Doppler peak height characteristic value of the unit to be detected is Calculated as follows:

[0024]

[0025] Represents the relative Doppler peak height characteristics of the unit to be detected, represents the echo vector of the unit to be detected with a length of N, represents the echo vector of the unit to be detected with a length of N, P refers to the number of reference units, Refers to the echo vector of the unit to be detected The Doppler peak height is Refers to the reference unit echo vector The Doppler peak height.

[0026] (3) Relative Doppler entropy

[0027] The difference in randomness is measured by entropy. The greater the randomness, the greater the amount of information and the greater the entropy.

[0028]

[0029] refers to the Doppler frequency, refers to the Doppler spectrum of the echo sequence, refers to the normalized Doppler spectrum of the echo sequence, represents the echo vector of the unit to be detected with a length of N, Refers to the entropy of the echo vector with a length of N for the unit to be detected.

[0030] After the reference unit is introduced, the relative Doppler entropy eigenvalue of the unit to be detected is Calculated as follows:

[0031] .

[0032] P refers to the number of reference cells, represents the echo vector of the unit to be detected with a length of N, represents the echo vector of the unit to be detected with a length of N, Refers to the entropy of the echo vector of the length N of the unit to be detected, Refers to the entropy of the echo vector of the length N of the unit to be detected, Represents the relative Doppler entropy characteristics of the unit to be detected

[0033] Preferably, the specific steps of step 2 are as follows:

[0034] (1) Training of convex hull classifier and false alarm control

[0035] According to the collected clutter three-feature samples, the clutter feature sample set is converted into a convex area surrounded by a triangular section as the original decision area. Once the feature sample point to be detected falls into the decision area, the feature sample point to be detected is judged as clutter; after the original decision area is formed, the sample points in the decision area are removed to reduce the decision area to control the false alarm rate; according to the regional distribution of the clutter three features and the target three features, sample points whose spatial distribution is more similar to the target three features are selected from the clutter three feature sample points used for training, and the following conditions are met:

[0036]

[0037] In the formula The set of training sample points to be deleted, represents the mean point in the original training feature point set, represents the original training feature point set, Refers to the feature vector samples in the feature point set, It means that the sample set is a clutter sample set;

[0038] From the training sample point set to be deleted The sample points as vertices are selected as the sample point set to be eliminated, and the point with the farthest Euclidean distance from the center point of the original sample point set is found as the sample point to be eliminated in the current round. After eliminating it, the decision area is re-formed until the number of remaining sample points and the total number of original sample points meet the following relationship, indicating that the initial control of the false alarm rate is completed;

[0039] 1 − P F ≈ #{ i : x i ∈ Oh } I or #{ i : x i ∈ Oh } = [ I × ( 1 − P F )] (8)

[0040] Refers to the preset false alarm rate, Refers to the decision area formed by clutter feature samples after false alarm rate control. Refers to the total number of clutter feature samples before false alarm rate control. refer to The number of feature sample points in the decision area.

[0041] The decision criteria of the convex hull detector can be summarized as follows:

[0042] ς ≡ max{det([ v q ( 1 ) − x , v q ( 2 ) − x , v q ( 3 ) − x ]), q = 1 , 2 ,..., Q } ≤ 0

[0043] In the formula represents the current feature vector to be detected, [ v q ( 1 ) , v q ( 2 ) , v q ( 3 ) ] Represents the coordinates of the three vertices of the triangle that makes up the convex hull surface, arranged clockwise, with subscript Represents the sequence number of the triangles that make up the convex hull surface, and its maximum value is The convex hull surface consists of Different triangle vertices are connected to each other. is the judgment result. If the value is less than zero, it means that the feature vector to be detected is In the convex hull judgment area, the feature vector to be detected is judged as a clutter vector, otherwise the feature vector to be detected is judged as a target vector;

[0044] (2) SVM classifier training and false alarm control

[0045] According to the collected clutter three-feature samples and target three-feature samples, each feature sample point is marked with a corresponding label , where the target feature sample label is +1, indicating the target class, which is represented as a positive sample in the SVM classifier, and the clutter feature sample point is marked as -1, indicating the clutter class, which is represented as a negative sample in the SVM classifier. The corresponding hyperplane is then found using the soft margin SVM constraint criterion;

[0046] minutes oh , b , x 1 2 ‖ oh ‖ 2 + ∑ i = 1 M ( 1 − y i 2 β 0 + 1 + y i 2 β 1 ) c i s . t . C 1 : y i [ k ( oh , F i ) − b ] ≥ 1 − x i , i = 1 , 2 ,..., M C 2 : c i ≥ 0 , i = 1 , 2 ,..., M

[0047] In the formula, , is the hyperplane weight coefficient, is the i-th feature sample, is the kernel transformation function, is the penalty factor, is the penalty parameter of clutter samples, is the penalty parameter of the target sample;

[0048] To initially control the false alarm rate, the penalty parameter of the clutter sample Will be bounded on the penalty parameter and the penalty parameter lower bound After repeated adjustments, when the hyperplane is formed, the clutter data used in training SVM is input into the classifier, and the number of clutter feature points misclassified as targets is counted to obtain the false alarm rate within the training sample range;

[0049] When the false alarm rate is high, the penalty parameter and the upper and lower bounds of the penalty parameter are updated using the following formula;

[0050]

[0051] is the penalty parameter of clutter samples, is the upper bound of the penalty parameter for clutter samples, is the lower bound of the penalty parameter for clutter samples.

[0052] When the false alarm rate is low, use the following formula to update the penalty parameter and the upper and lower bounds of the long hair parameter;

[0053]

[0054] is the penalty parameter of clutter samples, is the upper bound of the penalty parameter for clutter samples, is the lower bound of the penalty parameter for clutter samples.

[0055] This updating step is repeated until the false alarm rate obtained from the training clutter feature samples is With the preset false alarm rate The difference between them is within a preset range, and the control of false alarm rate is completed at this time;

[0056] (13)

[0057] Refers to the false alarm rate obtained from the training clutter feature samples, Refers to the preset false alarm rate, Refers to the difference between the two, usually taken as .

[0058] The decision criteria of the SVM detector can be summarized as follows:

[0059]

[0060] In the formula is the feature vector to be detected, is the corresponding test result, , is the hyperplane weight coefficient

[0061] Preferably, the specific steps of step 3 are:

[0062] After the initial false alarm control of the detector is completed, the actual detection work begins; the false alarm rate obtained by detecting unknown data is used as a means of perceiving the characteristics of the sea surface. After each pre-set time period, the false alarm rate obtained by the detector detecting the clutter unit during the time period is counted as an estimate of the actual false alarm rate. In order to ensure that the detector can respond correctly to the characteristics of the sea surface, the preset false alarm rate is calculated based on a large number of experimental observations. The actual false alarm rate The relationship is established as a linear function with a slope of 1 and a change in intercept:

[0063] (15)

[0064] In the actual detection process, due to the intercept is unknown and needs to be estimated. The estimation of is a reflection of the detector to the clutter environment, which is obtained by the difference between the actual false alarm rate and the preset false alarm rate. The difference between the actual false alarm rate and the preset false alarm rate is obtained by statistics on the intercept Update, use the updated relationship to guide the update of the detector's decision area, and in terms of behavior, change the value of the false alarm rate that needs to be maintained in the detector, and control the size of the decision area by constantly changing the value, and apply the adjusted detector to the detection work in the next time period;

[0065] (16)

[0066] Refers to the false alarm rate obtained during the actual detection process. Refers to the preset false alarm rate

[0067] Estimate the intercept Then, use the intercept Calculate the false alarm rate value that the detector should be adjusted to at this time ;

[0068] (17)

[0069] Refers to the false alarm rate of the current detector setting, Refers to the false alarm rate that the current detector should be adjusted to. Refers to the previously estimated intercept.

[0070] The method for dynamically adjusting the false alarm rate of marine target feature detection of the present invention has the following beneficial effects:

[0071] 1. It can reuse training data without the need to re-collect data within a certain period of time, and maintain the false alarm rate near a given false alarm rate under the condition that the fluctuation range of the false alarm rate remains roughly unchanged;

[0072] 2. It can control the false alarm rate at a sufficiently low time cost, thus ensuring the real-time nature of detection to a certain extent;

[0073] 3. This method can be applied to a variety of feature detection methods, which is helpful for evaluating the performance of the detector. After verification by HH and VV measured data, it can effectively improve the false alarm control ability of the detector. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a three-level sea condition relationship diagram between the preset false alarm rate and the actual false alarm rate when the detector receives clutter data other than the training data;

[0075] Figure 2 It is a five-level sea condition relationship diagram between the preset false alarm rate and the actual false alarm rate when the detector receives clutter data other than the training data;

[0076] Figure 3 This is a graph showing the change in false alarm rate before and after the use of the method for level 3 sea conditions;

[0077] Figure 4 This is a graph showing the change in false alarm rate before and after using the method in level 5 sea conditions. DETAILED DESCRIPTION

[0078] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and technical solutions.

[0079] A method for dynamically adjusting the false alarm rate of marine target feature detection in this embodiment includes the following steps:

[0080] Step 1: Extract features from the time domain and frequency domain of radar echo signals;

[0081] Step 2: Train the convex hull detector and the SVM detector according to the extracted clutter features and target echo features, and use the initial false alarm rate control method to preliminarily control the false alarm rate according to the preset false alarm rate value;

[0082] Step 3: Perform detection according to the input echo data, and count the total alarm rate of each period of time as an estimate of the actual false alarm rate. Determine the adjustment range of the judgment area according to the actual false alarm rate and the preset false alarm rate, and readjust the false alarm rate to achieve the purpose of reducing the difference with the preset false alarm rate;

[0083] Step 4: Return to step 3 and continue the test until the test is completed;

[0084] Step 5: Use measured data to verify the proposed method. Take the channel buoy as the small sea surface target to be detected, and verify the performance of the proposed method under HH and VV polarization modes.

[0085] The echo feature extraction involved in step 1 comes from the time domain and frequency domain. Three typical feature extraction methods are introduced below.

[0086] (1) Relative Average Amplitude (RAA)

[0087] This feature is defined as the ratio of the time domain average amplitude of the unit to be detected within the number of coherent pulses to the time domain average amplitude of the reference unit, which can be used to reflect the difference between the sea clutter echo and the target echo. The relative average amplitude value of the sea clutter unit will be close to 1. Since the amplitude of the target echo is often larger than that of the sea clutter echo, the relative average amplitude of the target unit is often greater than the relative average amplitude of the sea clutter unit.

[0088] (1)

[0089] In the formula, represents the echo vector of the unit to be detected with a length of N, Represents an echo sequence with a length of N to be detected. Refers to the echo sequence subscript, Refers to the average amplitude of the echo of the unit to be detected. represents an echo sequence with a reference unit length of N, Refers to the average amplitude of the echo of the reference unit, P is the number of reference units, Represents the relative average amplitude characteristics of the unit to be detected;

[0090] (2) Relative Doppler Peak Height (RDPH)

[0091] In the Doppler domain, the energy of the target echo Doppler spectrum is more concentrated than that of the clutter spectrum, which means that there is a large difference between the peak value of the target echo spectrum and the spectrum value outside the main lobe. If the peak value is compared with the spectrum value outside the main lobe, a larger value will be obtained. Compared with the target echo Doppler spectrum, the energy distribution of the clutter Doppler spectrum is more even, and the difference between the peak value and the spectrum value outside the main lobe is not as good as that of the target echo Doppler spectrum. The Doppler spectrum can be calculated according to the following formula

[0092] (2)

[0093] In the formula, refers to the Doppler frequency, refers to the pulse repetition period, Refers to the echo sequence subscript, Refers to the echo sequence length, Represents an echo sequence with a length of N to be detected. Refers to the Doppler spectrum of the echo sequence, based on which the Doppler peak height and peak height location are extracted;

[0094] D = [ − d 1 , − d 2 ] ∪ [ d 2 , d 1 ] RDPH ( x ) = peak ( x ) 1 # D ∑ f d ∈ f d max ( x ) + D X ( f d ) (3)

[0095] refers to the Doppler frequency, refer to The frequency corresponding to the maximum value is, Refers to the Doppler frequency range available for reference, Fingers fall into The number of Doppler bins in the frequency bin, , Refers to the frequency range for estimating the relative Doppler peak height, represents the echo vector of the unit to be detected with a length of N, Refers to the echo vector of the unit to be detected After the reference unit is introduced, the relative Doppler peak height characteristic value of the unit to be detected is Calculated as follows:

[0096] (4)

[0097] Represents the relative Doppler peak height characteristics of the unit to be detected, represents the echo vector of the unit to be detected with a length of N, represents the echo vector of the unit to be detected with a length of N, P refers to the number of reference units, Refers to the echo vector of the unit to be detected The Doppler peak height is Refers to the reference unit echo vector The Doppler peak height.

[0098] (3) Relative Doppler entropy (RVE)

[0099] The target echo Doppler spectrum and the clutter Doppler spectrum differ not only in the degree of energy concentration, but also in the degree of chaos. The clutter Doppler spectrum has greater randomness than the target echo Doppler spectrum. The difference in randomness can be measured by entropy. The greater the randomness, the greater the amount of information and the greater the entropy.

[0100] (5)

[0101] refers to the Doppler frequency, refers to the Doppler spectrum of the echo sequence, Refers to the normalized Doppler spectrum of the echo sequence, indicating the echo vector with a unit length of N to be detected. Refers to the entropy of the echo vector with a length of N for the unit to be detected.

[0102] After the reference unit is introduced, the relative Doppler entropy eigenvalue of the unit to be detected is Calculate as follows.

[0103] (6)

[0104] P refers to the number of reference cells, represents the echo vector of the unit to be detected with a length of N, represents the echo vector of the unit to be detected with a length of N, Refers to the entropy of the echo vector of the length N of the unit to be detected, Refers to the entropy of the echo vector of the length N of the unit to be detected, Represents the relative Doppler entropy characteristics of the unit to be detected

[0105] The specific steps of step 2 are as follows:

[0106] (1) Training of convex hull classifier and false alarm control

[0107] According to the collected clutter three-feature samples, the clutter feature sample set is converted into a convex area surrounded by a triangular section as the original decision area. Once the feature sample point to be detected falls into the decision area, the feature sample point to be detected is judged as clutter; after the original decision area is formed, the sample points in the decision area are removed to reduce the decision area to control the false alarm rate; according to the regional distribution of the clutter three features and the target three features, sample points whose spatial distribution is more similar to the target three features are selected from the clutter three feature sample points used for training, and the following conditions are met:

[0108] (7)

[0109] In the formula The set of training sample points to be deleted, represents the mean point in the original training feature point set, represents the original training feature point set, Refers to the feature vector samples in the feature point set, It means that the sample set is a clutter sample set;

[0110] From the training sample point set to be deleted The sample points as vertices are selected as the sample point set to be eliminated, and the point with the farthest Euclidean distance from the center point of the original sample point set is found as the sample point to be eliminated in the current round. After eliminating it, the decision area is re-formed until the number of remaining sample points and the total number of original sample points meet the following relationship, indicating that the initial control of the false alarm rate is completed;

[0111] 1 − P F ≈ #{ i : x i ∈ Oh } I or #{ i : x i ∈ Oh } = [ I × ( 1 − P F )] (8)

[0112] Refers to the preset false alarm rate, Refers to the decision area formed by clutter feature samples after false alarm rate control. Refers to the total number of clutter feature samples before false alarm rate control. refer to The number of feature sample points in the decision area.

[0113] The decision criteria of the convex hull detector can be summarized as follows:

[0114] ς ≡ max{det([ v q ( 1 ) − x , v q ( 2 ) − x , v q ( 3 ) − x ]), q = 1 , 2 ,..., Q } ≤ 0 (9)

[0115] In the formula represents the current feature vector to be detected, [ v q ( 1 ) , v q ( 2 ) , v q ( 3 ) ] Represents the coordinates of the three vertices of the triangle that makes up the convex hull surface, arranged clockwise, with subscript Represents the sequence number of the triangles that make up the convex hull surface, and its maximum value is The convex hull surface consists of Different triangle vertices are connected to each other. is the judgment result. If the value is less than zero, it means that the feature vector to be detected is In the convex hull judgment area, the feature vector to be detected is judged as a clutter vector, otherwise the feature vector to be detected is judged as a target vector;

[0116] (2) SVM classifier training and false alarm control

[0117] According to the collected clutter three-feature samples and target three-feature samples, each feature sample point is marked with a corresponding label , where the target feature sample label is +1, indicating the target class, which is represented as a positive sample in the SVM classifier, and the clutter feature sample point is marked as -1, indicating the clutter class, which is represented as a negative sample in the SVM classifier. The corresponding hyperplane is then found using the soft margin SVM constraint criterion;

[0118] minutes oh , b , x 1 2 ‖ oh ‖ 2 + ∑ i = 1 M ( 1 − y i 2 β 0 + 1 + y i 2 β 1 ) x i s . t . C 1 : y i [ k ( oh , F i ) − b ] ≥ 1 − x i , i = 1 , 2 ,..., M C 2 : x i ≥ 0 , i = 1 , 2 ,..., M (10)

[0119] In the formula, , is the hyperplane weight coefficient, is the i-th feature sample, is the kernel transformation function, is the penalty factor, is the penalty parameter of clutter samples, is the penalty parameter of the target sample;

[0120] To initially control the false alarm rate, the penalty parameter of the clutter sample Will be bounded on the penalty parameter and the penalty parameter lower bound After repeated adjustments, when the hyperplane is formed, the clutter data used in training SVM is input into the classifier, and the number of clutter feature points misclassified as targets is counted to obtain the false alarm rate within the training sample range;

[0121] To initially control the false alarm rate, the penalty parameter of the clutter sample Will be bounded on the penalty parameter and the penalty parameter lower bound After repeated adjustments, when the hyperplane is formed, the clutter data used in training SVM is input into the classifier, and the number of clutter feature points misclassified as targets is counted to obtain the false alarm rate within the training sample range;

[0122] When the false alarm rate is high, use the following formula to update the penalty parameter and the upper and lower bounds of the penalty parameter

[0123] (11)

[0124] is the penalty parameter of clutter samples, is the upper bound of the penalty parameter for clutter samples, is the lower bound of the penalty parameter for clutter samples.

[0125] When the false alarm rate is low, use the following formula to update the penalty parameter and the upper and lower bounds of the long hair parameter.

[0126] (12)

[0127] is the penalty parameter of clutter samples, is the upper bound of the penalty parameter for clutter samples, is the lower bound of the penalty parameter for clutter samples.

[0128] This updating step is repeated until the false alarm rate obtained from the training clutter feature samples is With the preset false alarm rate When the difference between them is within a preset range, the control of false alarm rate is completed.

[0129] (13)

[0130] Refers to the false alarm rate obtained from the training clutter feature samples, Refers to the preset false alarm rate, Refers to the difference between the two, usually taken as .

[0131] The decision criterion of the SVM detector can be summarized as follows:

[0132] (14)

[0133] In the formula is the feature vector to be detected, is the corresponding test result, , is the hyperplane weight coefficient.

[0134] The specific steps of step 3 are:

[0135] When the initial false alarm control of the detector is completed, the actual detection work can begin. The non-stationary and non-uniform characteristics of the sea surface mean that in order to maintain the false alarm rate during the detection process, the characteristics of the sea surface must be perceived in real time. The false alarm rate obtained by detecting unknown data can be used as a means of perceiving the characteristics of the sea surface. After each pre-set time period, the false alarm rate obtained by the detector detecting the clutter unit during the time period is counted as an estimate of the actual false alarm rate. In order to ensure that the detector can respond correctly to the characteristics of the sea surface, based on a large number of experimental observations, the preset false alarm rate can be The actual false alarm rate The relationship is established as a linear function with a slope of 1 and a change in intercept, as shown in the attached Figure 1-2 shown.

[0136] (15)

[0137] In the actual detection process, due to the intercept is unknown and needs to be estimated. The estimation of is a reflection of the detector to the clutter environment, which is obtained by the difference between the actual false alarm rate and the preset false alarm rate. The difference between the actual false alarm rate and the preset false alarm rate is obtained by statistics on the intercept Update, use the updated relationship to guide the update of the detector's decision area, and in terms of behavior, change the value of the false alarm rate that needs to be maintained in the detector, and control the size of the decision area by constantly changing the value, and apply the adjusted detector to the detection work in the next time period;

[0138] (16)

[0139] Refers to the false alarm rate obtained during the actual detection process. Refers to the preset false alarm rate

[0140] Estimate the intercept Then, use the intercept Calculate the false alarm rate value that the detector should be adjusted to at this time ;

[0141] (17)

[0142] Refers to the false alarm rate of the current detector setting, Refers to the false alarm rate that the current detector should be adjusted to. Refers to the previously estimated intercept.

[0143] Specific implementation steps of step 5: Test in HH polarization and VV polarization X-band sea detection data set with channel buoy as the target to be detected. First, extract a part of the clutter data as training data, and train the classifier according to the preset false alarm rate. Then, at the beginning of detection, collect the false alarm rate of the clutter area, correct the preset false alarm rate according to formula (17), and re-form the decision area according to the corrected false alarm rate. Use the updated decision area to continue target detection. The false alarm rate control situation is shown in Table 1.

[0144] Table 1: Comparison of false alarm rate control performance under different data sets

[0145]

[0146] Under the third level sea condition, under the condition of the preset false alarm rate of 1%, the average false alarm rate difference can be reduced to below 0.1% or even lower after using the method. The curve of the actual false alarm rate changing with time before and after using the method is attached. Figure 3 , Attachment Figure 4 From the attached Figure 3 , 4 It can be seen that the mean of the actual false alarm rate after using the method is closer to the preset false alarm rate.

[0147] Under level 5 sea conditions, the actual false alarm rate fluctuated more dramatically over time before the method was used. Although the degree of fluctuation was not reduced after the method was used, the actual false alarm rate had fluctuated around the preset false alarm rate, and the difference between its mean and the preset false alarm rate could still be maintained within 0.1%.

[0148] Regardless of HH, VV polarization or low and high sea conditions, the proposed method can improve the false alarm rate control ability of the detector and control the false alarm rate within an ideal range, which is not only conducive to the detection work, but also convenient for more effective evaluation of the performance of the detector.

Claims

1. A method for dynamically adjusting the false alarm rate of marine target feature detection, characterized in that The following steps are involved: Step 1) extract features from the time domain and frequency domain of the radar echo signal; Step 2) training the convex hull detector and the SVM detector according to the extracted clutter features and target echo features, using the initial false alarm rate control method to preliminarily control the false alarm rate according to the preset false alarm rate value; Step 3) Perform detection work according to the input echo data, and count the total alarm rate of each period of time as an estimate of the actual false alarm rate. Determine the adjustment range of the judgment area according to the actual false alarm rate and the preset false alarm rate, and readjust the false alarm rate to achieve the purpose of reducing the difference with the preset false alarm rate; Step 4) Return to step 3) and continue the testing work until the testing work is completed; Step 5) Use measured data to verify the proposed method. Take the channel buoy as the small sea surface target to be detected, and verify the performance of the proposed method under two polarization modes: HH polarization means that the electromagnetic waves transmitted and received by the radar are both horizontally polarized waves, and VV polarization means that the electromagnetic waves transmitted and received by the radar are both vertically polarized waves.

2. A method for dynamically adjusting false alarm rate of marine target feature detection according to claim 1, characterized in that The features extracted in step 1) come from the time domain and frequency domain. The following are the specific extraction methods of three typical features: (1) Relative average amplitude The relative average amplitude is defined as the ratio of the time domain average amplitude of the unit to be detected within the number of coherent pulses to the time domain average amplitude of the reference unit, which is used to reflect the difference between the sea clutter echo and the target echo. The relative average amplitude of the sea clutter unit is 1. Since the amplitude of the target echo is larger than that of the sea clutter echo, the relative average amplitude of the target unit is often greater than the relative average amplitude of the sea clutter unit. ; In the formula, represents the echo vector of the unit to be detected with a length of N, Represents an echo sequence with a length of N to be detected. Refers to the echo sequence subscript, Refers to the average amplitude of the echo of the unit to be detected. represents an echo sequence with a reference unit length of N, Refers to the average amplitude of the echo of the reference unit, P is the number of reference units, Represents the relative average amplitude characteristics of the unit to be detected; (2) Relative Doppler peak height The Doppler spectrum is calculated as follows: ; In the formula, refers to the Doppler frequency, refers to the pulse repetition period, Refers to the echo sequence subscript, Refers to the echo sequence length, Represents an echo sequence with a length of N to be detected. Refers to the Doppler spectrum of the echo sequence, based on which the Doppler peak height and peak height location are extracted; use , The reference Doppler unit is defined as Refers to the width of the Doppler main peak, Refers to the Doppler band unit for reference; ; refers to the Doppler frequency, refer to The frequency corresponding to the maximum value is, Refers to the Doppler frequency range available for reference, Fingers fall into The number of Doppler bins in the frequency bin, , Refers to the frequency range for estimating the relative Doppler peak height, represents the echo vector of the unit to be detected with a length of N, Refers to the echo vector of the unit to be detected The Doppler peak height of the unit to be detected is the relative Doppler peak height characteristic value after the reference unit is introduced. Calculated as follows: ; Represents the relative Doppler peak height characteristics of the unit to be detected, represents the echo vector of the unit to be detected with a length of N, represents the echo vector of the unit to be detected with a length of N, P refers to the number of reference units, Refers to the echo vector of the unit to be detected The Doppler peak height is Refers to the reference unit echo vector The Doppler peak height; (3) Relative Doppler entropy The difference in randomness is measured by entropy. The greater the randomness, the greater the amount of information and the greater the entropy. ; refers to the Doppler frequency, refers to the Doppler spectrum of the echo sequence, refers to the normalized Doppler spectrum of the echo sequence, represents the echo vector of the unit to be detected with a length of N, Refers to the entropy of the echo vector of the unit length N to be detected; After the reference unit is introduced, the relative Doppler entropy eigenvalue of the unit to be detected is Calculated as follows: ; P refers to the number of reference cells, represents the echo vector of the unit to be detected with a length of N, represents the echo vector of the unit to be detected with a length of N, Refers to the entropy of the echo vector of the length N of the unit to be detected, Refers to the entropy of the echo vector of the length N of the unit to be detected, Represents the relative Doppler entropy characteristics of the unit to be detected.

3. A method for dynamically adjusting false alarm rate of marine target feature detection according to claim 1, characterized in that The specific steps of step 2) are as follows: (1) Training of convex hull classifier and false alarm control According to the collected clutter three-feature samples, the clutter feature sample set is converted into a convex area surrounded by a triangular section as the original decision area. Once the feature sample point to be detected falls into the decision area, the feature sample point to be detected is judged as clutter; after the original decision area is formed, the sample points in the decision area are removed to reduce the decision area to control the false alarm rate; according to the regional distribution of the clutter three features and the target three features, sample points whose spatial distribution is more similar to the target three features are selected from the clutter three feature sample points used for training, and the following conditions are met: ; In the formula The set of training sample points to be deleted, represents the mean point in the original training feature point set, represents the original training feature point set, Refers to the feature vector samples in the feature point set, It means that the sample set is a clutter sample set; From the training sample point set to be deleted The sample points as vertices are selected as the sample point set to be eliminated, and the point with the farthest Euclidean distance from the center point of the original sample point set is found as the sample point to be eliminated in the current round. After eliminating it, the decision area is re-formed until the number of remaining sample points and the total number of original sample points meet the following relationship, indicating that the initial control of the false alarm rate is completed; (8); Refers to the preset false alarm rate, Refers to the decision area formed by clutter feature samples after false alarm rate control. Refers to the total number of clutter feature samples before false alarm rate control, refer to The number of characteristic sample points in the decision area; The decision criteria of the convex hull detector can be summarized as follows: ; In the formula represents the current feature vector to be detected, Represents the coordinates of the three vertices of the triangle that makes up the convex hull surface, arranged clockwise, with subscript Represents the sequence number of the triangles that make up the convex hull surface, and its maximum value is The convex hull surface consists of Different triangle vertices are connected to each other. is the judgment result. If the value is less than zero, it means that the feature vector to be detected is In the convex hull judgment area, the feature vector to be detected is judged as a clutter vector, otherwise the feature vector to be detected is judged as a target vector; (2) SVM classifier training and false alarm control According to the collected clutter three-feature samples and target three-feature samples, each feature sample point is marked with a corresponding label , where the target feature sample label is +1, indicating the target class, which is represented as a positive sample in the SVM classifier, and the clutter feature sample point is marked as -1, indicating the clutter class, which is represented as a negative sample in the SVM classifier. The corresponding hyperplane is then found using the soft margin SVM constraint criterion; ; In the formula, , is the hyperplane weight coefficient, is the i-th feature sample, is the kernel transformation function, is the penalty factor, is the penalty parameter of clutter samples, is the penalty parameter of the target sample; To initially control the false alarm rate, the penalty parameter of the clutter sample Will be bounded on the penalty parameter and the penalty parameter lower bound After repeated adjustments, when the hyperplane is formed, the clutter data used in training SVM is input into the classifier, and the number of clutter feature points misclassified as targets is counted to obtain the false alarm rate within the training sample range; When the false alarm rate is high, the penalty parameter and the upper and lower bounds of the penalty parameter are updated using the following formula; ; is the penalty parameter of clutter samples, is the upper bound of the penalty parameter for clutter samples, is the lower bound of the penalty parameter for clutter samples; When the false alarm rate is low, use the following formula to update the penalty parameter and the upper and lower bounds of the long hair parameter; ; is the penalty parameter of clutter samples, is the upper bound of the penalty parameter for clutter samples, is the lower bound of the penalty parameter for clutter samples; This updating step is repeated until the false alarm rate obtained from the training clutter feature samples is With the preset false alarm rate The difference between them is within a preset range, and the control of false alarm rate is completed at this time; (13); Refers to the false alarm rate obtained from the training clutter feature samples, Refers to the preset false alarm rate, Refers to the difference between the two, usually taken as ; The decision criteria of the SVM detector can be summarized as follows: ; In the formula is the feature vector to be detected, is the corresponding test result, , is the hyperplane weight coefficient.

4. A method for dynamically adjusting false alarm rate of marine target feature detection according to claim 1, characterized in that The specific steps of step 3) are: After the initial false alarm control of the detector is completed, the actual detection work begins; the false alarm rate obtained by detecting unknown data is used as a means of perceiving the characteristics of the sea surface. After each pre-set time period, the false alarm rate obtained by the detector detecting the clutter unit during the time period is counted as an estimate of the actual false alarm rate. In order to ensure that the detector can respond correctly to the characteristics of the sea surface, the preset false alarm rate is calculated based on a large number of experimental observations. The actual false alarm rate The relationship is established as a linear function with a slope of 1 and a change in intercept: (15); In the actual detection process, due to the intercept is unknown and needs to be estimated. The estimation of is a reflection of the detector to the clutter environment, which is obtained by the difference between the actual false alarm rate and the preset false alarm rate. The difference between the actual false alarm rate and the preset false alarm rate is obtained by statistics on the intercept Update, use the updated relationship to guide the update of the detector's decision area, and in terms of behavior, change the value of the false alarm rate that needs to be maintained in the detector, and control the size of the decision area by constantly changing the value, and apply the adjusted detector to the detection work in the next time period; (16); Refers to the false alarm rate obtained during the actual detection process. Refers to the preset false alarm rate; Estimate the intercept Then, use the intercept Calculate the false alarm rate value that the detector should be adjusted to at this time ; (17); Refers to the false alarm rate of the current detector setting, Refers to the false alarm rate that the current detector should be adjusted to. refers to the previously estimated intercept; During the entire detection process, the decision area of ​​the detector is constantly changing according to the changes in the clutter environment, coping with the sea clutter environment with space-time non-stationary characteristics and completing real-time and precise control of the false alarm rate.

Citation Information

Patent Citations

  • Sea surface weak target detection method based on multi-domain features

    CN117129958A

  • Sea surface small target detection method and system

    CN117518115A

  • Deep neural network (DNN)-based multi-target constant false alarm rate (CFAR) detection methods

    US20240004032A1