Multi-channel ATI-SAR moving target detection method based on pseudo signal assistance

Abstract

The invention provides a multi-channel ATI-SAR moving target detection method based on pseudo signal assistance. The method comprises the following steps: firstly, synchronously injecting power-controlled known reference signals into two residual images subjected to adaptive clutter suppression, then constructing an interferogram and carrying out complex multi-view interference processing to obtain amplitude and phase detection quantities, then respectively solving amplitude and phase detection thresholds based on a constant false alarm criterion, and finally, carrying out adaptive clutter suppression on the two residual images according to the amplitude and phase detection thresholds. And the pixels exceeding the double thresholds at the same time are judged as moving targets through a two-step detector. According to the method, the detection threshold is deduced in a non-uniform clutter environment, the working mode is selected according to requirements, and the optimal detection performance can be realized under a constant false alarm probability. Due to the fact that the spatial degree of freedom is fully utilized and pseudo signals are introduced for assistance, the detection capacity of the low-speed weak moving target with the low signal-to-noise ratio under the non-uniform strong clutter background is remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of radar technology, and more specifically relates to a moving target detection method based on pseudo-signal-assisted multi-channel along-track interferometry-synthetic aperture radar (ATI-SAR) in the field of radar target detection technology. This invention can be used in ATI-SAR systems to detect low signal-to-noise ratio, slow-moving, small targets against non-uniform, strong clutter backgrounds. Background Technology

[0002] ATI-SAR systems, by configuring two or more SAR imaging channels along the platform's motion track, enable multi-frame SAR image observation of the same scene, representing a special form of multi-channel SAR system. Compared to single-channel systems, multi-channel systems increase the spatial degrees of freedom of the signal, enabling more effective clutter suppression and thus improving the detection performance of moving targets under low signal-to-clutter ratios. However, in non-uniform clutter backgrounds, even after adaptive clutter suppression, ATI-SAR systems may still experience clutter residuals coexisting with noise, severely impacting target detection. Current technologies for multi-channel ATI-SAR systems still fall short in utilizing spatial degrees of freedom, especially in non-uniform, strong clutter backgrounds, where detecting slow-moving targets under low signal-to-clutter ratios remains a significant challenge.

[0003] CH Gierull, I. Sikaneta et al., in their published paper "Two-step detector for RADARSAT-2's experimental GMTI mode" (IEEE Transactions on Geoscience and RemoteSensing, vol. 51, no. 1, A two-step detection method is proposed in pp. 436–454 (2013). The method involves: first, using signal amplitude information after multi-channel adaptive clutter suppression for preliminary detection of the SAR image; then, using two-channel track-interferometric ATI phase data for a second-step detection; and finally, performing a logical AND operation on the two-step detection results. However, this method still has shortcomings. The ATI phase detection statistic in the second step only utilizes echo data from two channels, resulting in a waste of spatial degrees of freedom for multi-channel radar systems. Furthermore, the estimation performance of this ATI phase statistic deteriorates significantly under low signal-to-noise ratio conditions, leading to instability in detection performance against strong clutter backgrounds.

[0004] Xi'an University of Electronic Science and Technology, in its patent application "A Moving Target Detection Method Based on Optimal Fusion of Multi-channel ATI-SAR under Strong Clutter Background" (Application No.: 2024119027902, Authorization Announcement No.: CN 119846626 B), proposed a detection method based on optimal fusion for airborne multi-channel ATI-SAR systems. This method first uses a set of multi-baseline de-cluttering ATI phase tests and AMF amplitude tests for local detection; then, it designs global detection optimization based on optimal fusion rules and performs global detection. However, this method still has shortcomings. Because it fails to maximize the utilization of the spatial degrees of freedom of the multi-channel synthetic aperture radar system, strong clutter residue may still remain after adaptive clutter suppression. When the target signal coexists with clutter, the ATI phase estimation becomes unreliable, leading to an increase in the minimum resolvable SCNR and a decrease in detection performance.

[0005] M. Tian and B. Liao, in their published paper "Interferometric Phase of Clutter-Suppression Residuals Aided" A moving target detection method utilizing additional spatial degrees of freedom (DoFs) was proposed in "Multichannel SAR-GMTI" (IEEE Transactions on Geoscience and Remote Sensing, pp. 1–16, 2025). This method uses the residual amplitude of the adaptive clutter suppression output and combines it with the interference phase calculated between the two clutter-suppressed residuals from the preceding and following channels, thus enhancing the contrast between the target and the background to some extent and improving detection performance. However, a drawback remains: since most residuals approximate noise signals in the absence of moving targets, and the interference phase of noise tends to be randomly distributed within [−π, π], the presence of this noise competes with the real target during phase-based detection, leading to a performance degradation. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the existing technologies by proposing a pseudo-signal-assisted multi-channel ATI-SAR moving target detection method. This method aims to solve the problems of insufficient utilization of spatial degrees of freedom and difficulty in detecting slow targets under low signal-to-clutter ratios in multi-channel ATI-SAR systems under non-uniform strong clutter backgrounds.

[0007] The technical solution to achieve the objective of this invention is as follows: This invention utilizes SAR images from N channels to obtain clutter suppression residual maps for the first N-1 channels and the last N-1 channels. Then, a pseudo-signal injection mechanism is established, whereby a known reference signal with controlled power (set to be higher than the noise floor but lower than the minimum target echo power) is simultaneously superimposed onto the two generated clutter suppression residual images. Next, the two residuals after pseudo-signal injection are used to construct an interferogram. Through complex multi-look interferometry, amplitude and phase detection values ​​are obtained respectively. Then, based on the constant false alarm rate (CFAR) principle, corresponding amplitude and phase detection thresholds are set. Finally, a two-step detector combining amplitude and interferometric phase is used to detect pixels that simultaneously exceed both amplitude and phase thresholds as moving targets. Because this invention fully utilizes the signal-to-noise ratio (SNR) improvement effect of spatial diversity and adaptive clutter suppression, and by introducing pseudo-signal assistance, forces the originally randomly distributed noise interference phase to be clamped near zero radians, noise interference is effectively suppressed. This provides an effective solution to the problem of difficult detection of slow, weak targets with low signal-to-noise ratio in non-uniform, strong clutter backgrounds.

[0008] The specific implementation steps of this invention include the following:

[0009] Step 1: Generate echo data for N SAR images, where N is equal to the total number of channels in the GMTI synthetic aperture radar system.

[0010] Step 2: Construct two sub-apertures and apply adaptive clutter suppression to generate two residual images;

[0011] Step 3: Inject a power-controlled pseudo-signal into the targetless, pure noise region of the residual image;

[0012] Step 4: Construct an L-Look complex interferogram and extract the detection quantities for amplitude and phase tests;

[0013] Step 5: Design a two-step moving target detector, determine the detection threshold based on the false alarm rate, solve for the detection threshold, and adjust the detection mode;

[0014] Step 6: Preprocess the SAR image to be detected using the same method as steps 1 to 4; perform the first detection on the preprocessed image to be detected to filter out pixels with larger amplitudes; perform the second detection on the filtered pixels and determine the pixels that pass the second detection as moving targets.

[0015] Furthermore, the two sub-apertures are respectively composed of the first N-1 and the last N-1 channels of the N-channel system; wherein, the first N-1 channels refer to channels 1 to N-1; and the last N-1 channels refer to channels 2 to N.

[0016] Furthermore, the application of adaptive clutter suppression to generate two residual images refers to using adaptive clutter suppression weights. Image data is processed separately for two sub-apertures, where... satisfy The superscript H indicates the conjugate transpose operation.

[0017] Furthermore, the expression for the power-controlled pseudo-signal is as follows:

[0018] ;

[0019] in, The first image represents the residual image. The pseudo-signal injected into each pixel Indicates a complex Gaussian distribution. This indicates the strength of the artificially injected spurious signal. This represents the noise power in the residual image. Set within the range of (1,10) to satisfy , This represents the target's remaining power, which is typically greater than the noise power. 10 times.

[0020] Furthermore, the L-Look complex interferogram is obtained by the following formula:

[0021] ;

[0022] in, The pseudo signal indicating controlled injection power is the first A complex interferogram with L-number of views per pixel, where L represents the number of views in the complex interferogram. This represents the k-th pixel in the first residual image. Individual sample, This represents the k-th pixel in the second residual image. Individual sample, The conjugate operator is represented by E[·], which represents the function that calculates the expected value.

[0023] Furthermore, the detection quantities for the amplitude test and phase test are obtained by the following formula:

[0024] ;

[0025] in, and They represent the first The number of pixels detected for amplitude and phase testing. and They represent the first Normalized interferometric and phase values ​​of the L-look number complex interferogram of pixels. The modulus of a complex number, This represents the phase angle when taken as a complex number.

[0026] Furthermore, the steps for designing the two-step moving target detector are as follows:

[0027] The first step is to construct the moving target detector using the obtained amplitude and phase test detection quantities according to the following formula. and The current pixel is determined to be a moving target if it satisfies the following conditions:

[0028] ;

[0029] in, and These represent the threshold values ​​for the detection quantities in amplitude testing and phase testing, respectively.

[0030] The second step is to obtain the hypothesis according to the following formula. Down and Correlation coefficient between :

[0031] ;

[0032] in, This indicates a situation where no target was detected.

[0033] The third step is to obtain the effective number of views according to the following formula. :

[0034] ;

[0035] in, This represents the operation of taking the real part of a complex number. Represented by natural constant An exponential function with base 0. Symbol for the imaginary unit;

[0036] Fourth, obtain the following estimate from the echo data sample according to the following formula:

[0037] ;

[0038] in, and Let represent the degrees of freedom and shape parameter of the inverse chi-square distribution, respectively. and They represent respectively to and The estimated value, , These represent the detection quantities t for amplitude testing and phase testing, respectively. Step moment, , , This represents the Beta function;

[0039] Fifth, under clutter background, generate hypotheses according to the following formula. Down and Joint probability density function:

[0040] ;

[0041] in, This represents the estimated average interference phase, typically . , Represents pi (π). Represents the gamma function. Represents the Gaussian hypergeometry function. Indicates the amount of detection in the amplitude test. of Power of 1 Represents the cosine function;

[0042] Step 6: Obtain the detection thresholds of the two-stage moving target detectors at the specified false alarm rate using the following formula: ;

[0043] in, Indicates the false alarm rate, within a specified range. Below, solve for two or more combinations of detection thresholds. .

[0044] Furthermore, the steps of solving the detection threshold and adjusting the detection mode are as follows:

[0045] The first step is to establish a preset constant false alarm rate. The index, using the derived interferogram statistical distribution model, establishes an index regarding the amplitude threshold. With phase threshold The constraint equations;

[0046] The second step is to solve the constraint equations to obtain multiple combinations of amplitude and phase thresholds that satisfy the same constant false alarm rate. ;

[0047] The third step is to select the detection mode based on the actual task requirements and the detection thresholds for different groups. Different detection modes are corresponding to this; when the detection task is low signal-to-clutter-to-noise ratio target detection, the combined midpoint amplitude threshold is selected. Smaller and phase threshold The larger set is used as the parameters for the amplitude-priority mode; when the detection task is slow target detection, the combined phase threshold is selected. Smaller and amplitude threshold The larger set is used as the parameters for amplitude-priority mode.

[0048] Furthermore, the first step of detection on the preprocessed image to be detected refers to comparing the amplitude value of each pixel in the preprocessed image to the threshold value of the amplitude test detection. Compare the values ​​and assign values ​​with amplitudes less than the threshold. All pixels are identified as background pixels.

[0049] Furthermore, the second step of detecting the selected pixels refers to comparing the phase value of the selected pixels with the threshold of the phase test detection quantity. Compare the values ​​and assign phase values ​​less than a threshold. All pixels are identified as pixels of the moving target.

[0050] Compared with the prior art, the present invention has the following advantages:

[0051] First, this invention employs two sub-apertures, consisting of the first N-1 channels and the last N−1 channels. This method, while preserving the interference phase baseline, maximizes the utilization of the system's spatial degrees of freedom for adaptive clutter suppression, overcoming the shortcomings of insufficient utilization of spatial degrees of freedom in existing technologies. This allows the invention to suppress clutter in non-uniform, strong clutter backgrounds, resulting in a significantly higher signal-to-clutter-to-noise ratio (SCNR) gain than existing technologies, providing a cleaner signal foundation for subsequent detection.

[0052] Secondly, this invention introduces a pseudo-signal injection mechanism, namely, injecting a power-controlled pseudo-signal into the residual obtained after adaptive clutter suppression. In a targetless, purely noisy region, the injected pseudo-signal can dominate the phase characteristics of the synthesized signal, forcibly clamping the interference phase of the background noise near zero radians. This overcomes the problem of unreliable phase detection at low signal-to-noise ratios in existing technologies, significantly enhancing the contrast between the interference phases of moving targets and stationary backgrounds, thereby effectively solving the problem of unreliable phase detection at low signal-to-noise ratios. Attached Figure Description

[0053] Figure 1 This is a flowchart of the method of the present invention;

[0054] Figure 2 This is an observational geometric diagram showing the relationship between the N-channel ATI-SAR system used in this invention and moving ground targets;

[0055] Figure 3These are simulation results that verify the performance of the method proposed in this invention. Detailed Implementation

[0056] The present invention will now be described in more detail with reference to the accompanying drawings and embodiments.

[0057] Reference Figure 1 The specific implementation steps of the embodiments of the present invention will be further described below.

[0058] Step 1: Construct a radar signal model and acquire image data.

[0059] Step 1.1: First, establish the radar signal model. Consider an airborne GMTI synthetic aperture radar equipped with a linear array of N antenna elements uniformly distributed along the platform's flight path, with an element spacing of d, operating in side-looking mode, as shown below. Figure 2 As shown.

[0060] Figure 2 In this context, Platform velocity refers to the platform speed, Radar array antenna refers to the radar array antenna, Incident Angle refers to the incident angle, Azimuth Angle refers to the azimuth angle, Radial velocity refers to the radial velocity, and Building refers to the building.

[0061] exist Figure 2 At the reference time, the origin O and the x-axis correspond to the aircraft's ground projection position and track (azimuth) direction, respectively. Without loss of generality, the z-axis and y-axis are defined as perpendicular to the ground and the xOz plane, respectively. Within one coherent processing interval (CPI), it is assumed that the aircraft moves at a constant speed along the x-axis. Flight. For moving ground targets, the azimuth and angle of incidence are respectively used as... and The radial velocity between the target and the radar is expressed as... express.

[0062] Step 1.2: Using an N-channel airborne GMTI synthetic aperture radar system, echo snapshot data of N synthetic aperture radar images equal to the number of channels are generated. Then, these images are corrected through image registration and calibration procedures.

[0063] Step 2: Define the binary hypothesis test and process the SAR image data.

[0064] Step 2.1, let This represents N×1 data vector snapshots from N processed SAR images, in pixels. ∈{1, 2, ..., K}. Where, Indicates from the first The first channel obtained Complex signals in amplitude synthetic aperture radar images, ∈{1, 2, ..., N};

[0065] The binary hypothesis test used for target detection is defined as follows:

[0066] ;

[0067] in, Indicates no objective assumption. This indicates that there is a target assumption; and These represent the clutter vector and the signal vector of the moving target, respectively; noise vector. It follows a zero-mean complex Gaussian distribution, denoted as , Indicates noise power. It is an identity matrix.

[0068] In the assumption Assuming the amplitude of the focused target signal is the same on all channels, The expression is:

[0069] ;

[0070] in, Indicates the magnitude of a complex target. This represents the target space direction vector.

[0071] ;

[0072] For radial velocity is Ground moving target, target Doppler , It is the radar wavelength.

[0073] Step 2.2, refer to Figure 2 As shown, after obtaining echo snapshot data of N synthetic aperture radar images, synthetic aperture radar images from the first N-1 and last N-1 channels are used to construct a given pixel. The two data vectors are:

[0074] ;

[0075] The first N-1 channels refer to channels 1 to N-1; the last N-1 channels refer to channels 2 to N.

[0076] Step 2.3, using a filter processor Adaptive clutter suppression is applied to the two sets of data from the first N-1 and the last N-1 SAR channels respectively, generating two residual images. and as follows:

[0077] ;

[0078] in, Indicates used to satisfy The normalized weight vector for adaptive clutter suppression. This indicates the conjugate transpose operation. .

[0079] Step 3: Introduce a pseudo-signal model, perform complex interferometry on the two residual images, and extract the amplitude and phase detection quantities.

[0080] Step 3.1, the pseudo-signal with power greater than the bottom of the noise but lower than the target power. Added to the two residuals, resulting in two modified residuals. and .

[0081] False signals Introduced into two residuals:

[0082] ;

[0083] in, The first image represents the residual image. The pseudo-signal injected into each pixel This invention proposes a constant to represent the intensity of a pseudo-signal. This represents the target's remaining power, which is typically greater than the noise power. Ten times that. Therefore, it can be Set within the range of (1,10) to satisfy .

[0084] Step 3.2, for and Constructing a complex L-Look interferogram involves fusing the two signals to extract physical features.

[0085] ;

[0086] in, The pseudo signal indicating controlled injection power is the first A complex interferogram with L-number of views per pixel, where L represents the number of views in the complex interferogram. This represents the k-th pixel in the first residual image. Individual sample, This represents the k-th pixel in the second residual image. Individual sample, E[·] represents the conjugate operation, and E[·] represents the function for finding the expected value.

[0087] Step 3.3: Extract two independent detection quantities from the complex L-Look interferogram. and :

[0088] ;

[0089] in, and They represent the first The number of pixels detected for amplitude and phase testing. and They represent the first Normalized interferometric and phase values ​​of the L-look number complex interferogram of pixels. The modulus of a complex number, This represents the phase angle when taken as a complex number.

[0090] Step 4: Construct a two-stage moving target detector.

[0091] Design a two-stage moving target detector that combines the interferometric amplitude and phase:

[0092] ;

[0093] in, and These are the amplitude and phase detection thresholds. According to the constant false alarm rate (CFAR) principle, the amplitude needs to be measured against a background of clutter and noise. and phase The joint detection statistical properties are characterized in order to determine the threshold. and To ensure a constant false alarm rate Unaffected by environmental changes.

[0094] Step 5: Modeling the non-uniform clutter residual model.

[0095] To solve for the detection threshold of the target detector in the previous step and First, the residuals need to be modeled to facilitate the subsequent derivation of the statistical properties of the interferogram. Considering the non-uniformity of ground clutter, the residual clutter after adaptive clutter suppression often exhibits non-Gaussian statistical characteristics. Multiplicative textures are used to simulate non-uniform clutter, thereby obtaining the residual model after clutter suppression.

[0096] use Representing the pixel index, the clutter residuals from the first N−1 SAR images are modeled as... ,in ∈[0, ∞) is a texture variable describing amplitude variation, while These are complex Gaussian speckle components. For the subsequent N−1 SAR images, an additional interferometric phase term is included: ,in To represent the clutter interference phase, assuming that in The value is approximately zero below. The two residuals can be reformulated as:

[0097] ;

[0098] Will Defined as a combination term describing clutter residuals and pseudo-signals, i.e.:

[0099] ;

[0100] in, It follows a complex Gaussian distribution with zero mean and variance. : .

[0101] Step 6: Derive the statistical properties of the interferogram and solve for the local detection threshold.

[0102] Step 6.1: Based on the residual model established in the previous step, this step derives the statistical properties of the residual interferogram to obtain the given false alarm rate. The required local detection threshold and .

[0103] According to the following formula, under the assumption Down and The joint probability density function (pdf) can be analytically expressed as:

[0104] ;

[0105] in, Indicates the number of valid views. and Let represent the degrees of freedom and shape parameter of the inverse chi-square distribution, respectively. This represents the estimated average interference phase, typically . , Represents pi (π). Represents the real part of a complex number. Represents the gamma function. This represents the Gaussian hypergeometric function.

[0106] use Indicates a hypothesis Down and The correlation coefficient between them can be empirically estimated as:

[0107] ;

[0108] in, The modulus of a complex number.

[0109] By using To calculate the interference phase on the test data The second moment, This can be obtained by solving the following equation:

[0110] ;

[0111] Estimate from data sample and for:

[0112] ;

[0113] in, and They represent respectively to and The estimated value, , These represent the detection quantities t for amplitude testing and phase testing, respectively. Step moment, , , This represents the Beta function.

[0114] According to the following formula, under the assumption Down, and The combined PDF is represented as:

[0115] ;

[0116] in, express The second-order modified Bessel function of the second kind, Represented by natural constant An exponential function with base , for the assumption Assuming the interference amplitude associated with a deterministic moving target Correlation coefficient and interference phase It is known or can be estimated from the remaining samples of the target.

[0117] Step 6.2, given and The joint probability density function pdf and the specified global false alarm rate The corresponding local detection threshold can be obtained by solving the following equation. and :

[0118] .

[0119] Among them, in the specified Below, there are two or more combinations of detection thresholds. This equation can be satisfied.

[0120] Step 7: Two-step cascaded detection to determine the final target.

[0121] The two-stage moving target detector proposed in this invention employs a two-step cascaded detection strategy.

[0122] The first step is to perform amplitude detection and judgment. Filter out points with larger amplitudes;

[0123] The second step is to perform phase detection, and further determine the phase based on the amplitude detection. ;

[0124] When amplitude test and phase test Pixels that exceed their corresponding detection thresholds are considered targets; otherwise, they are considered background.

[0125] Step 8: Threshold selection analysis and detection probability estimation.

[0126] Step 8.1, there may be multiple pairs of thresholds. and It can satisfy: In the formula (False alarm rate) requirement, and each pair of threshold combinations These factors can all lead to different target detection results.

[0127] The detection mode is switched based on the selection of the local detection threshold mentioned above, such as phase priority mode (prioritizing the detection of low-speed targets) and amplitude priority mode (prioritizing the detection of targets with low SCNR).

[0128] Step 8.2, for adaptive clutter suppression and minimum detectable velocity (denoted as...) ),about and The minimum resolvable output signal-to-noise ratio (denoted as ) (Unit: decibel) can be approximately expressed as:

[0129] ;

[0130] Given a local detection threshold, the corresponding detection probability It is represented as:

[0131] .

[0132] The effects of the present invention will be further explained below with reference to simulation experiments.

[0133] 1. Simulation experimental conditions:

[0134] The simulation environment for this invention is: MATLAB R2021a, Intel(R) Core(TM)2 Duo CPU 16GHz, Windows 10 Professional Edition.

[0135] 2. Simulation content and result analysis:

[0136] The simulation experiments of this invention used the proposed methods (Magnitude-Priority, Phase-Priority) and three existing methods (Mag+ATI, Optimal Fusion, InPhase Aided) to predict the receiver operating characteristic (ROC) curves under actual clutter backgrounds. The false alarm rate of each method was observed. Detection probability ( The curve showing the change in output signal-to-noise ratio (SCNR) as a function of output, such as... Figure 3 As shown, the detection performance of the method proposed in this invention is comprehensively evaluated.

[0137] The existing Mag+ATI method refers to the two-step detection method (MAG+ATI phase) combining amplitude and interference phase proposed by CH Gierull, I. Sikaneta, and D. Cerutti-Maori in their paper "Two-step detector for RADARSAT-2's experimental GMTImode" (IEEE Transactions on Geoscience and Remote Sensing, vol. 51, no. 1, pp.436–454, 2013).

[0138] The existing Optimal Fusion method refers to the optimal fusion method proposed by M. Tian and B. Liao in their paper "Optimal Fusion-Based Target Detection With Multichannel ATI SAR" (IEEE Transactions on Geoscience and Remote Sensing, vol. 62, pp. 1–15, 2024.).

[0139] The existing InPhase Aided method refers to the interferometric phase-aided detection method based on clutter suppression residuals proposed by M. Tian, ​​B. Liao, B. Yuan, and DHHu in their paper "Interferometric Phase of Clutter-Suppression Residuals Aided Multichannel SAR-GMTI" (IEEE Transactions on Geoscience and Remote Sensing, pp. 1–16, 2025).

[0140] To ensure a fair comparison, each method employed five-view averaging and estimated its respective statistical parameters from real clutter measurement data to calculate the false alarm rate. The detection threshold below.

[0141] In the Mag+ATI (two-step detection) method, the amplitude detection threshold is set to... The ATI phase detection threshold is set to To ensure the overall .

[0142] For the method proposed in this invention, two local detection thresholds (i.e. and ) is at a given false alarm rate The values ​​are estimated below. Here, the parameters of the Magnitude-Priority and Phase-Priority detectors proposed in this invention are... , The values ​​are (0.6998, 0.83 rad) and (2.2495, 0.425 rad) respectively.

[0143] Figure 3Indicates the false alarm rate Below, the detection probability ( The graph shows the curve (estimated ROC curve) of how the output signal-to-noise ratio (SCNR) changes, where... Figure 3 (a) and Figure 3 (b) Corresponding to the target radial velocity =1.0m / s and =2.0m / s. At the false alarm rate When, the detection probability of the comparison method ( The graph shows the variation of the output signal-to-clutter ratio (OutputSCNR), where Figure (a) corresponds to the radial velocity of the target. =1.0 m / s, radial velocity of the target corresponding to Figure (b) =2.0m / s.

[0144] exist Figure 3 In this context, Output SCNR is the output signal-to-clutter-noise ratio, Mag+ATI is amplitude + along-track interference, OptimalFusion is optimal fusion, InPhase Aided is in-phase aid, Magnitude-Priority is amplitude priority, and Phase-Priority is phase priority.

[0145] When using detection probability When used as a detection indicator, Table I records the minimum resolvable input SCNR value for each comparison method.

[0146] Table I. Quantitative Indicators for Target Detection

[0147]

[0148] As shown in Table 1, the Magnitude-Priority and Phase-Priority methods proposed in this invention both achieved lower minimum resolvable input SCNR values ​​than the three methods used for comparison (Mag+ATI, Optimal Fusion, and InPhaseAided), indicating that the method proposed in this invention achieved superior target detection performance.

[0149] When the target radial velocity At that time, the minimum resolvable input signal-to-clutter ratio (SNR) obtained by the Magnitude-Priority method was -8.5 dB, which is better than the result of the Phase-Priority method (-5.5 dB). This shows that the Magnitude-Priority method is more suitable for detecting targets with low SNR; when the target radial velocity... At that time, the minimum resolvable input signal-to-noise ratio obtained by the Phase-Priority method was 0dB, which is better than the result of the Magnitude-Priority method (4.5dB). It can be seen that the Phase-Priority method is more suitable for detecting low-speed targets.

[0150] The above description is merely a preferred embodiment of the present invention. For those skilled in the art, various improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A method for detecting moving targets using a multi-channel ATI-SAR based on pseudo-signal assistance, characterized in that, The detection method includes the following steps: Step 1: Generate echo data for N SAR images, where N is equal to the total number of channels in the GMTI synthetic aperture radar system. Step 2: Construct two sub-apertures and apply adaptive clutter suppression to generate two residual images; Step 3: Inject a power-controlled pseudo-signal into the targetless, pure noise region of the residual image; Step 4: Construct an L-Look complex interferogram and extract the detection quantities for amplitude and phase tests; Step 5: Design a two-step moving target detector, determine the detection threshold based on the false alarm rate, solve for the detection threshold, and adjust the detection mode; Step 6: Preprocess the SAR image to be detected using the same method as steps 1 to 4; perform the first detection on the preprocessed image to be detected to filter out pixels with larger amplitudes; perform the second detection on the filtered pixels and determine the pixels that pass the second detection as moving targets.

2. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 1, characterized in that, The two sub-apertures mentioned in step 2 are composed of the first N-1 and the last N-1 channels of the N-channel system, respectively; where the first N-1 channels refer to channel 1 to channel N-1; and the last N-1 channels refer to channel 2 to channel N.

3. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 1, characterized in that, The application of adaptive clutter suppression to generate two residual images in step 2 refers to using adaptive clutter suppression weights. Image data is processed separately for two sub-apertures, where... satisfy , This indicates the conjugate transpose operation.

4. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 1, characterized in that, The expression for the power-controlled pseudo-signal mentioned in step 3 is as follows: ； in, The first part represents the residual image. The pseudo-signal injected into each pixel Indicates a complex Gaussian distribution. This indicates the strength of the artificially injected spurious signal. This represents the noise power in the residual image. Set within the range of (1,10) to satisfy , This represents the target's remaining power, which is typically greater than the noise power. 10 times.

5. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 4, characterized in that, The L-Look complex interferogram mentioned in step 4 is obtained by the following formula: ； in, The pseudo-signal indicating controlled injection power is the first A complex interferogram with L-number of views per pixel, where L represents the number of views in the complex interferogram. This represents the k-th pixel in the first residual image. Individual sample, This represents the k-th pixel in the second residual image. Individual sample, E[·] represents the conjugate operation, and E[·] represents the function for finding the expected value.

6. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 5, characterized in that, The detection quantities for amplitude and phase testing described in step 4 are obtained by the following formula: ； in, and They represent the first The number of pixels detected for amplitude and phase testing. and They represent the first Normalized interferometric and phase values ​​of L-look complex interferograms of pixels. The modulus of a complex number, This represents the phase angle when taken as a complex number.

7. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 6, characterized in that, The steps for designing the two-step moving target detector described in step 5 are as follows: The first step is to construct the moving target detector using the obtained amplitude and phase test detection quantities according to the following formula. and The current pixel is determined to be a moving target if it satisfies the following conditions: ； in, and These represent the threshold values ​​for the detection quantities in amplitude testing and phase testing, respectively. The second step is to obtain the hypothesis according to the following formula. Down and Correlation coefficient between : ； in, This indicates a situation where no target was detected. The third step is to obtain the effective number of views according to the following formula. : ； in, This represents the operation of taking the real part of a complex number. Represented by natural constant An exponential function with base 0. The symbol representing the imaginary unit; Fourth, obtain the following estimate from the echo data sample according to the following formula: ； in, and Let represent the degrees of freedom and shape parameter of the inverse chi-square distribution, respectively. and They represent respectively to and The estimated value, , These represent the detection quantities t for amplitude testing and phase testing, respectively. Step moment, , , This represents the Beta function; Fifth, under clutter background, generate hypotheses according to the following formula. Down and Joint probability density function: ； in, This represents the estimated average interference phase, typically . , Represents pi (π). Represents the gamma function. Represents the Gaussian hypergeometry function. Indicates the amount of detection in the amplitude test. of Power of 1 Represents the cosine function; Step 6: Obtain the detection thresholds of the two-stage moving target detectors at the specified false alarm rate using the following formula: ; in, Indicates the false alarm rate, within a specified range. Below, solve for two or more combinations of detection thresholds. .

8. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 7, characterized in that, The steps for solving the detection threshold and adjusting the detection mode in step 5 are as follows: The first step is to establish a preset constant false alarm rate. The index, using the derived interferogram statistical distribution model, establishes an index regarding the amplitude threshold. With phase threshold The constraint equations; The second step is to solve the constraint equations to obtain multiple combinations of amplitude and phase thresholds that satisfy the same constant false alarm rate. ; The third step is to select the detection mode based on the actual task requirements and the detection thresholds for different groups. Different detection modes are corresponding to this; when the detection task is low signal-to-clutter-to-noise ratio target detection, the combined midpoint amplitude threshold is selected. Smaller and phase threshold The larger set is used as the parameters for amplitude-priority mode; When the detection task is slow target detection, the combined centering phase threshold is selected. Smaller and amplitude threshold The larger set is used as the parameters for amplitude-priority mode.

9. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 7, characterized in that, Step 6, which describes the first step of detection on the preprocessed image to be detected, refers to comparing the amplitude value of each pixel in the preprocessed image to the threshold value for amplitude testing. Compare the values ​​and assign values ​​with amplitudes less than the threshold. All pixels are identified as background pixels.

10. The multi-channel ATI-SAR moving target detection method based on pseudo-signal assistance according to claim 8, characterized in that, The second step of detection for the selected pixels mentioned in step 6 refers to comparing the phase value of the selected pixels with the threshold of the phase test detection quantity. Compare the values ​​and assign phase values ​​less than a threshold. All pixels are identified as pixels of the moving target.

Images