Multi-channel sar moving target detection method and device for complex urban scene

Abstract

This invention discloses a multi-channel SAR moving target detection method and apparatus for complex urban scenarios, belonging to the field of synthetic aperture radar technology. The method addresses the problems of high false alarm rate and poor ATI phase reliability in traditional DPCA detection caused by strong scatterers in densely built-up urban areas. First, it performs registration and equalization preprocessing on the two-channel complex images. Then, it performs track-based interferometry and constructs a high-coherence mask using a coherence coefficient map to constrain the interferometric phase and eliminate low-reliability phases. Next, it constructs a weight function using the constrained interferometric phase to weight and fuse the DPCA differential amplitude results, enhancing the response of real moving targets and suppressing strong clutter residues. Finally, it performs constant false alarm rate detection on the weighted results, achieving robust detection of moving targets. This invention effectively solves the problem of high false alarm rate in moving target detection under complex urban backgrounds and significantly improves detection reliability.

Description

Technical Field

[0001] This invention belongs to the field of Synthetic Aperture Radar (SAR) technology, specifically relating to a multi-channel SAR moving target detection method and device for complex urban scenarios. It is suitable for robust detection of slow-moving, weakly scattering ground targets with low false alarm rate in observation scenarios with dense urban buildings and complex scattering environments. Background Technology

[0002] Synthetic Aperture Radar (SAR) has become an important tool for Earth observation due to its all-weather, all-day imaging capabilities. Combining Ground Moving Target Indication (GMTI) with SAR imaging can further acquire dynamic information in the scene, enabling the detection, localization, and parameter estimation of moving targets, which has significant application value in fields such as traffic monitoring.

[0003] Single-channel SAR systems suffer from limited detection performance when detecting slow-moving targets within the Doppler spectrum of ground clutter due to the lack of spatial degrees of freedom. To address this, multi-channel SAR-GMTI systems deployed along flight paths have emerged. By introducing spatial-dimensional observation information and combining it with two-dimensional spatiotemporal processing techniques, these systems can effectively suppress stationary clutter and improve the detection capability of moving targets.

[0004] Currently, the mainstream methods for multi-channel SAR-GMTI mainly include Displaced Phase Center Antenna (DPCA) technology, Along-track Interferometry (ATI) technology, and Space-time Adaptive Processing (STAP) technology. DPCA technology suppresses clutter by canceling images between channels. Its principle is simple and easy to implement in engineering, but its ideal cancellation requires complete coherence between channel signals. In densely built-up urban areas, there are many strong scatterers such as building facades and metal components. The strong backscattering signals and multipath effects generated by these objects easily lead to decorrelation between channel signals, leaving a large amount of strong clutter energy after DPCA cancellation, resulting in serious false alarms. This is particularly detrimental to the detection of slow-moving, weakly scattering targets. ATI technology uses inter-channel interferometric phase to distinguish moving targets from stationary scenes, but it itself does not have clutter suppression capabilities, and its detection performance is limited in low signal-to-clutter ratio environments. Especially in complex urban scattering environments, the spatial distribution of the interference phase is extremely unstable due to the mismatch between strong scatterers and channels. Low-coherence regions generate a large amount of random phase noise, making it difficult to set detection thresholds based on a unified statistical model and control false alarms. While STAP technology theoretically possesses optimal clutter suppression performance, its performance heavily relies on a large number of training samples that satisfy the independent and identically distributed condition to accurately estimate the clutter covariance matrix. Urban clutter is highly non-uniform and non-stationary, making it difficult to obtain satisfactory training samples, leading to a decline in STAP filter performance. Furthermore, STAP processing involves high-dimensional matrix operations, resulting in high computational complexity and significant engineering implementation challenges.

[0005] In summary, existing methods such as DPCA, ATI, and STAP all have significant limitations when dealing with typical complex scenarios like urban built-up areas, which exhibit strong scattering, significant multipath effects, and easily decorrelated channels, making it difficult to achieve robust moving target detection with low false alarm rates. Therefore, there is an urgent need for a novel SAR moving target detection method that is tailored to the characteristics of complex urban scenarios, can effectively integrate the advantages of multi-channel information, and is also engineering-feasible. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a multi-channel SAR moving target detection method and apparatus for complex urban scenarios. Starting from the practical problem of uneven spatial distribution of interferometric phase reliability in strong urban clutter backgrounds, it introduces a high coherence region constraint mechanism based on coherence coefficient, and weightedly fuses the constrained ATI results with the DPCA differential results to achieve robust detection of ground moving targets.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A multi-channel SAR moving target detection method for complex urban scenarios, the method comprising:

[0009] Step 1: Acquire multi-channel SAR data and preprocess it to obtain two single-view complex images after registration and channel equalization;

[0010] Step 2: Perform mean filtering and flat phase removal processing on the two single-view complex images after registration and channel equalization to obtain a pre-processed image pair;

[0011] Step 3: Perform trace interferometry on the pre-processed image pair to obtain an interferometric phase map and a coherence coefficient map. Then, constrain the interferometric phase map based on the coherence coefficient map, remove the interferometric phase corresponding to the low coherence region, and obtain a high-confidence interferometric phase.

[0012] Step 4: Perform offset phase center antenna processing on the pre-processed image pair to obtain differential amplitude results, and use the high-confidence interferometric phase to construct a weight function to perform weighted fusion on the differential amplitude results to obtain a weighted fusion result;

[0013] Step 5: Perform constant false alarm rate (CFAR) detection on the weighted fusion result and output the moving target detection result.

[0014] Furthermore, step 1 includes: using the main channel single-view complex image as a reference, performing sub-pixel level image registration on the auxiliary channel single-view complex image; performing two-dimensional frequency domain adaptive equalization on the registered auxiliary channel single-view complex image, and solving the equalization filter by minimizing the energy function of the difference between the two channels to make its spectral characteristics consistent with the main channel single-view complex image.

[0015] Furthermore, step 2 includes: applying the same mean filter operator to the two single-view complex images after registration and channel equalization to obtain a filtered image pair; multiplying the filtered image pair by their conjugates to generate an interferogram; performing a Fourier transform on the interferogram along the distance direction and estimating the main peak frequency offset; shifting the spectral peak to the zero-frequency position to eliminate the flat-ground phase effect; and adding the compensated flat-ground phase to the auxiliary channel single-view complex image.

[0016] Furthermore, in step 3, constraining the interference phase map based on the coherence coefficient map specifically involves: setting a coherence coefficient threshold; for pixels with a coherence coefficient less than or equal to the threshold, setting their interference phase to zero; and for pixels with a coherence coefficient greater than the threshold, retaining their original interference phase.

[0017] Furthermore, in step 4, the weight function is constructed as 1 minus the cosine value of the high-confidence interference phase. The weight function is then multiplied by the differential amplitude result obtained from the bias phase center antenna processing to obtain the weighted fusion result.

[0018] Furthermore, in step 5, the constant false alarm rate (CFAR) detection adopts the maximum selection CFAR detection method, which estimates the background power of the unit to be detected by selecting a larger reference window mean.

[0019] Furthermore, the multi-channel SAR data refers to the echo data acquired by a synthetic aperture radar system with two or more channels deployed along the flight path.

[0020] On the other hand, the present invention provides a multi-channel SAR moving target detection device for complex urban scenarios, comprising:

[0021] The image preprocessing module is used to acquire multi-channel SAR data and preprocess it to obtain two single-view complex images after registration and channel equalization.

[0022] The image optimization module is used to perform mean filtering and flat phase removal processing on the two single-view complex images after registration and channel equalization to obtain a pre-processed image pair.

[0023] The interferometric processing and phase constraint module is used to perform trace interferometric processing on the pre-processed image pair to obtain an interferometric phase map and a coherence coefficient map, and constrain the interferometric phase map based on the coherence coefficient map to remove the interferometric phase corresponding to the low coherence region to obtain a high-confidence interferometric phase.

[0024] The weighted fusion module is used to perform offset phase center antenna processing on the pre-processed image pair to obtain differential amplitude results, and to construct a weight function using the high-confidence interferometric phase to perform weighted fusion on the differential amplitude results to obtain a weighted fusion result.

[0025] The constant false alarm rate (CFAR) detection module is used to perform CFAR detection on the weighted fusion result and output the moving target detection result.

[0026] Thirdly, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned multi-channel SAR moving target detection method for complex urban scenarios.

[0027] Fourthly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned multi-channel SAR moving target detection method for complex urban scenarios.

[0028] The beneficial effects of this invention are as follows:

[0029] Highly targeted and with significant false alarm suppression: This invention specifically addresses the core challenges of densely built-up urban areas, including numerous strong scatterers, complex multipath effects, and easily decorrelated channels, proposing a systematic solution. By introducing a high-coherence region constraint mechanism based on the coherence coefficient, it effectively eliminates unstable random phases generated by low-coherence regions (such as shadows and strong scatterer sidelobes) in traditional ATI, suppressing invalid information interference at its source. Furthermore, the constrained reliable interference phase is used to weighted modulate the DPCA results, significantly enhancing the response of real moving targets while strongly suppressing clutter energy caused by residual strong scatterers such as buildings. Experimental results show that this method effectively solves the problem of high false alarm rates in traditional DPCA in urban areas, greatly improving detection reliability.

[0030] Integration and Innovation, Complementary Advantages: This invention is not a simple improvement on a single technology, but rather a creative integration of the phase discrimination information provided by ATI and the clutter cancellation capability of DPCA. Through an innovative "constraint-weighted" architecture, it overcomes the respective shortcomings of traditional ATI in its lack of clutter suppression capability and DPCA in its insufficient suppression of strong scatterers, achieving complementarity and synergy between the advantages of these two classic technologies, and achieving a detection effect of "1+1>2" in complex scenarios.

[0031] The process is clear and highly feasible in engineering: the entire processing flow of this invention is completed in the image domain or complex image domain, avoiding the complex high-dimensional matrix estimation and inversion operations of schemes such as STAP, as well as the huge computational burden brought by multi-parameter global search. The method is modular, the computational complexity is controllable, and it is easier to implement and apply in practical engineering systems. Attached Figure Description

[0032] Figure 1 This is a flowchart of the multi-channel SAR moving target detection method for complex urban scenarios according to the present invention;

[0033] Figure 2 The SAR imaging results of the reference channel in one embodiment of the present invention are shown;

[0034] Figure 3(a) shows the intermediate results of moving target detection obtained using the traditional ATI method;

[0035] Figure 3(b) shows the intermediate results of moving target detection obtained using the traditional DPCA method;

[0036] Figure 3(c) shows the intermediate results obtained after interferometric phase constraint processing using the method of the present invention;

[0037] Figure 3(d) shows the output result obtained after weighted fusion of DPCA results using the method of the present invention;

[0038] Figure 4This is the final moving target detection result after applying GO-CFAR detection to the weighted DPCA results of this invention. Detailed Implementation

[0039] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0040] This invention proposes a multi-channel SAR moving target detection method and device for urban built-up areas. The basic idea is as follows: First, preprocessing is performed on single-view complex images acquired along the flight path using two channels, including registration and channel equalization. Then, flat-ground phase surface fitting is performed to eliminate the influence of flat-ground phase. Based on this, track-side interferometry is performed to calculate the interferometric phase and coherence coefficient, and a high-coherence region mask is constructed based on the coherence coefficient threshold. This mask is used to constrain the ATI results and extract physically reliable phase information. Simultaneously, DPCA processing is performed on the multi-channel data, and the constrained ATI information is introduced to weight the DPCA output. Finally, the weighted DPCA results are subjected to Constant False Alarm Rate (CFAR) detection.

[0041] like Figure 1 As shown, the method specifically includes:

[0042] Step 1: Process the acquired multi-channel SAR data to obtain two registered and channel-equalized single-view complex images;

[0043] Single-view composite image of the main channel along the track (SLC1 in the figure) is used as a reference for the single-view complex image of the auxiliary channel along the track. (SLC2 shown in the figure) Subpixel-level image registration is performed. The registration process includes offset estimation, interpolation resampling, and other operations to obtain the registered auxiliary channel single-view complex image, denoted as... The image is in geometric coordinates with... Fully aligned, each pixel value remains a complex value. Single-view complex image of the main channel. No geometric transformations are performed during the registration process, therefore the registered master channel single-view complex image remains the same. .

[0044] Subsequently, in order to eliminate system errors introduced by factors such as differences in antenna patterns and inconsistencies in the amplitude and phase characteristics of the receiving channel, it is necessary to process the registered auxiliary channel single-look complex image. Perform two-dimensional frequency domain adaptive equalization to make its spectral characteristics match those of the main channel single-view complex image. Towards consensus.

[0045] This invention is based on the least squares criterion, solving for the equalization filter by minimizing the energy function of the difference between the two channels. Let... Let be the frequency response function of the two-dimensional equalized filter to be solved, where Indicates the range frequency. Representing the azimuth frequency, its objective function is simplified as follows:

[0046] ,

[0047] The integral region covers the effective support domain of the two-dimensional frequency domain. The two-dimensional spectrum of the main channel single-view complex image. This is the two-dimensional spectrum of the registered auxiliary channel single-view complex image. Let be the frequency response function of the two-dimensional equalized filter to be solved. For range frequency, This refers to the azimuth frequency.

[0048] The optimal equalizer frequency response function can be obtained by solving the above minimization problem using iterative or closed-form methods. Then, it is compared with the registered secondary channel spectrum. After performing complex multiplication and then undergoing a two-dimensional inverse Fourier transform, the equalized single-view complex image is obtained, denoted as... This step only processes the auxiliary channel data; therefore, the equalized main channel single-view complex image remains unchanged. To ensure consistency, record .

[0049] Step 2: Perform mean filtering and flat-ground phase removal processing on the two registered and channel-equalized single-view complex images to remove the phase error caused by the residual intersecting orbit baseline, and obtain the pre-processed image.

[0050] (1) Mean Filtering. If two registered and channel-equalized single-view complex images are directly multiplied by their conjugates, the interference phase will fluctuate significantly due to speckle noise, severely affecting the texture features of the interference image. Neighborhood mean filtering is typically used to reduce the impact of phase noise. The mean filtering operator is defined as a sliding window average, with the window size preset according to the image resolution and coherence requirements. and Applying the same filtering operator, mean-filtered single-view complex images of the main channel were obtained respectively. With auxiliary channel single-view complex image .

[0051] (2) Removal of Flat-Ground Phase. This step uses a surface fitting method based on the interferogram to estimate and remove the flat-ground phase. Its advantage lies in the fact that it does not require precisely known prior information such as platform motion parameters, baseline length, and slant range. Specifically, the method involves: performing a Fourier transform on the interferogram along the range direction to obtain its range spectrum; estimating the main peak frequency shift through spectral analysis; and shifting the spectral peak to the zero-frequency position, which is equivalent to multiplying the interferogram by a linear phase factor in the time domain, thereby eliminating the flat-ground phase effect. The input for this step is the interferogram generated by the conjugate multiplication of the main and auxiliary channel complex images after mean filtering. ,Right now: ;in To obtain the complex conjugate operator, and These are the main and auxiliary channel single-view complex images after mean filtering.

[0052] Estimating the main peak frequency shift in the frequency domain Then the interferogram after removing the flat phase The calculation is as follows:

[0053] ,

[0054] In the time domain, it is equivalent to:

[0055] ,

[0056] Where t represents the distance in the time domain, For range frequency, For distance to Fourier transform operator, For the distance-to-inverse Fourier transform operator, This is the interferogram generated by the conjugate multiplication of the main and auxiliary channel complex images after mean filtering. The peak offset is estimated in the frequency domain, where j represents the imaginary unit.

[0057] Further, the main channel single-view complex image after removing the flat phase is obtained. With auxiliary channel single-view complex image For simplicity, the compensated flat phase is added to the auxiliary channel single-view complex image. Above, that is:

[0058] ,

[0059] in, For phase operators, This is a single-view complex image of the auxiliary channel after mean filtering. The interferogram after removing the flat phase. This is the interferogram generated by the conjugate multiplication of the main and auxiliary channel complex images after mean filtering. The main channel single-view complex image was not processed. .

[0060] Step 3: Perform trace interferometry on the pre-processed image to obtain an interferometric phase map and a coherence coefficient map; constrain the interferometric phase map based on the coherence coefficient map to obtain a high-confidence interferometric phase.

[0061] The complex image pairs obtained after preliminary processing and ATI processing is performed, and the coherence coefficient map of the processed complex image pair is introduced as prior information to remove low-coherence noise regions. The threshold is selected based on the actual situation. The coherence coefficient map of the two-channel image after preliminary processing is calculated. A threshold decision mechanism is introduced to construct a high-confidence mask to eliminate invalid phases in low-coherence regions (typically corresponding to shadows or extremely decorrelational noise regions). Let the original interference phase be... The coherence coefficient threshold is Then the constrained interference phase Represented as:

[0062] ,

[0063] Among them, the interference phase obtained from the preliminary processing in step 2 , and The complex image pairs obtained from the preliminary processing in step 2, As an empirical threshold, for and The obtained coherence coefficient diagram.

[0064] Step 4: Process the complex image pairs after preliminary processing in Step 2. and The bias phase center antenna is processed to obtain the traditional DPCA differential result. The traditional DPCA result is then weighted and modulated using the weight function constructed from the high-confidence constrained interferometric phase obtained in step 3.

[0065] Constructed weight function The aim is to enhance regions with significant moving target phase characteristics while further suppressing clutter.

[0066] The weighted DPCA results obtained as follows:

[0067] ,

[0068] in, The weighting function is constructed as follows: (This is the interference phase constrained in step 3) , For traditional DPCA results, and The image pairs are the complex images obtained from the preliminary processing in step 2.

[0069] Step 5: Analyze the weighted DPCA results. Perform constant false alarm rate (CFAR) detection and output the detection results of moving targets. The maximum selected CFAR (Greatest Of-Constant False Alarm Rate) estimates the background power of the target cell by selecting the mean of a larger reference window. This can detect moving targets within the edge range of strong urban building clutter and is better adapted to non-uniform background noise compared to the cell average CFAR.

[0070] To demonstrate the effectiveness of this method, an experiment was conducted using SAR echo data from two channels of an airborne Ku-band satellite along its trajectory. The imaging results are as follows: Figure 2 As shown. The two cooperative objectives are respectively Figure 2 The two curved curves marked in red are due to the lack of range migration correction during imaging. The results of moving target detection using the traditional ATI and DPCA methods are shown in Figures 3(a) and 3(b), respectively. The ATI results with intermediate coherence constraints and the final weighted DPCA results obtained using the scheme of this invention are shown in Figures 3(c) and 3(d), respectively. The result of GO-CFAR detection using the weighted DPCA is... Figure 4 As shown in the results, the moving targets in the two red boxes can be extracted more effectively by this method compared to the traditional ATI and DPCA methods, proving the effectiveness of this method.

[0071] On the other hand, the present invention provides a multi-channel SAR moving target detection device for complex urban scenarios, the various modules of which can implement the various steps of the aforementioned method, specifically including:

[0072] The image preprocessing module is used to acquire multi-channel SAR data and preprocess it to obtain two single-view complex images after registration and channel equalization.

[0073] The image optimization module is used to perform mean filtering and flat phase removal processing on the two single-view complex images after registration and channel equalization to obtain a pre-processed image pair.

[0074] The interferometric processing and phase constraint module is used to perform trace interferometric processing on the pre-processed image pair to obtain an interferometric phase map and a coherence coefficient map, and constrain the interferometric phase map based on the coherence coefficient map to remove the interferometric phase corresponding to the low coherence region to obtain a high-confidence interferometric phase.

[0075] The weighted fusion module is used to perform offset phase center antenna processing on the pre-processed image pair to obtain differential amplitude results, and to construct a weight function using the high-confidence interferometric phase to perform weighted fusion on the differential amplitude results to obtain a weighted fusion result.

[0076] The constant false alarm rate (CFAR) detection module is used to perform CFAR detection on the weighted fusion result and output the moving target detection result.

[0077] Thirdly, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned multi-channel SAR moving target detection method for complex urban scenarios.

[0078] Fourthly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned multi-channel SAR moving target detection method for complex urban scenarios.

[0079] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-channel SAR moving target detection method for complex urban scenarios, characterized in that, The method includes: Step 1: Acquire multi-channel SAR data and preprocess it to obtain two single-view complex images after registration and channel equalization; Step 2: Perform mean filtering and flat phase removal processing on the two single-view complex images after registration and channel equalization to obtain a pre-processed image pair; Step 3: Perform trace interferometry on the pre-processed image pair to obtain an interferometric phase map and a coherence coefficient map. Then, constrain the interferometric phase map based on the coherence coefficient map, remove the interferometric phase corresponding to the low coherence region, and obtain a high-confidence interferometric phase. Step 4: Perform offset phase center antenna processing on the pre-processed image pair to obtain differential amplitude results, and use the high-confidence interferometric phase to construct a weight function to perform weighted fusion on the differential amplitude results to obtain a weighted fusion result; Step 5: Perform constant false alarm rate (CFAR) detection on the weighted fusion result and output the moving target detection result.

2. The multi-channel SAR moving target detection method for complex urban scenarios according to claim 1, characterized in that, Step 1 includes: using the main channel single-view complex image as a reference, performing sub-pixel level image registration on the auxiliary channel single-view complex image; performing two-dimensional frequency domain adaptive equalization on the registered auxiliary channel single-view complex image, and solving the equalization filter by minimizing the energy function of the difference between the two channels to make its spectral characteristics consistent with the main channel single-view complex image.

3. The multi-channel SAR moving target detection method for complex urban scenarios according to claim 1, characterized in that, Step 2 includes: applying the same mean filter operator to the two single-view complex images after registration and channel equalization to obtain a filtered image pair; multiplying the filtered image pair by conjugate to generate an interferogram; performing a Fourier transform on the interferogram along the distance direction and estimating the main peak frequency offset; shifting the spectral peak to the zero frequency position to eliminate the flat-ground phase effect; and adding the compensated flat-ground phase to the auxiliary channel single-view complex image.

4. The multi-channel SAR moving target detection method for complex urban scenarios according to claim 1, characterized in that, In step 3, constraining the interference phase map based on the coherence coefficient map specifically involves: setting a coherence coefficient threshold; for pixels with a coherence coefficient less than or equal to the threshold, setting their interference phase to zero; and for pixels with a coherence coefficient greater than the threshold, retaining their original interference phase.

5. The multi-channel SAR moving target detection method for complex urban scenarios according to claim 1, characterized in that, In step 4, the weight function is constructed as 1 minus the cosine value of the high-confidence interference phase. The weight function is then multiplied by the differential amplitude result obtained from the bias phase center antenna processing to obtain the weighted fusion result.

6. The multi-channel SAR moving target detection method for complex urban scenarios according to claim 1, characterized in that, In step 5, the constant false alarm rate (CFAR) detection adopts the maximum selection CFAR detection method, which estimates the background power of the unit to be detected by selecting a larger reference window mean.

7. The multi-channel SAR moving target detection method for complex urban scenarios according to claim 1, characterized in that, The multi-channel SAR data refers to the echo data acquired by a synthetic aperture radar system with two or more channels deployed along the flight path.

8. A multi-channel SAR moving target detection device for complex urban scenarios, characterized in that, include: The image preprocessing module is used to acquire multi-channel SAR data and preprocess it to obtain two single-view complex images after registration and channel equalization. The image optimization module is used to perform mean filtering and flat phase removal processing on the two single-view complex images after registration and channel equalization to obtain a pre-processed image pair. The interferometric processing and phase constraint module is used to perform trace interferometric processing on the pre-processed image pair to obtain an interferometric phase map and a coherence coefficient map, and constrain the interferometric phase map based on the coherence coefficient map to remove the interferometric phase corresponding to the low coherence region to obtain a high-confidence interferometric phase. The weighted fusion module is used to perform offset phase center antenna processing on the pre-processed image pair to obtain differential amplitude results, and to construct a weight function using the high-confidence interferometric phase to perform weighted fusion on the differential amplitude results to obtain a weighted fusion result. The constant false alarm rate (CFAR) detection module is used to perform CFAR detection on the weighted fusion result and output the moving target detection result.

9. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When one or more programs are executed by the one or more processors, the one or more processors implement the multi-channel SAR moving target detection method for complex urban scenarios as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, enable the processor to implement the multi-channel SAR moving target detection method for complex urban scenarios as described in any one of claims 1-7.

Images