Wide-angle staring SAR video imaging method and device based on trajectory error estimation

Abstract

The present application relates to a wide-angle staring SAR video imaging method and apparatus based on trajectory error estimation. The method comprises: performing range compression on the original echo signal received by a wide-angle staring SAR, dividing the entire aperture into multiple overlapping sub-apertures, and performing coarse imaging of the sub-apertures using a BP imaging algorithm; performing coarse correction using a backprojection autofocusing algorithm based on the coarse image; imaging the sub-apertures using a BP imaging algorithm based on the coarse correction trajectory, dividing the imaging area into multiple sub-areas, estimating the trajectory error using a backprojection autofocusing algorithm and a weighted least squares method for each sub-area, and accurately compensating the coarse correction trajectory; registering the precisely compensated trajectories of different sub-apertures using a Kabsch algorithm, and performing video frame imaging using an autoregressive BP imaging algorithm based on the registered trajectory positions. This method achieves video frame imaging, autofocusing, and stability between video frames under any imaging model.

Description

Technical Field

[0001] The present application relates to the field of radar signal processing technology, and in particular to a wide-angle staring SAR video imaging method and device based on trajectory error estimation. Background Art

[0002] Synthetic Aperture Radar (SAR) has become a hot topic in radar research in recent years. With increasing research, its application scope is expanding, including terrain mapping and information support for post-disaster rescue. Wide-angle Staring SAR (WasSAR) is a special SAR imaging mode that achieves long-term imaging observation of key areas by flying along a curve or circle.

[0003] Video SAR (ViSAR) enables high-frame-rate radar imagery surveillance of a designated area. While traditional strip SAR imaging modes only produce static SAR images, ViSAR continuously displays SAR images at a constant frame rate, providing a more intuitive visual representation of the target's motion within the scene through video, yielding richer and more accurate detection information. Because ViSAR requires long-term observation of key areas, it is often combined with the wide-angle staring mode, resulting in the WasSAR video imaging mode.

[0004] Frequency-domain imaging algorithms are typically the preferred algorithms for SAR imaging, but since WasSAR video imaging geometry is no longer a straight line, frequency-domain imaging algorithms are generally no longer applicable. Time-domain imaging algorithms, such as backprojection algorithms, have no restrictions on imaging geometry and are therefore commonly used for WasSAR imaging processing.

[0005] The quality of ViSAR imaging depends significantly on the accuracy of the trajectory of the radar-carrying platform. While the platform's trajectory is typically measured using a Global Positioning System (GPS) or an Inertial Navigation System (INS), errors are inevitable, leading to defocus and geometric distortion in the resulting image. This requires further fine-grained motion compensation using an autofocus algorithm based on radar echo data. Current SAR autofocus algorithms primarily include the Map Drift (MD) algorithm and the Phase Gradient Autofocus (PGA) algorithm. However, these algorithms can cause image drift and jitter when applied to ViSAR. Recently, researchers have proposed a ViSAR video frame stabilization method that combines image registration. This method assumes that image drift is directly related to trajectory error and infers the trajectory position based on the positional relationship between features in the image and adjacent frames. However, this method can only correct for translational and rotational errors in the trajectory and is applicable only to the terahertz band, not to lower bands such as the X and Ku bands, which require overlapping subapertures. Summary of the Invention

[0006] Based on this, it is necessary to provide a wide-angle staring SAR video imaging method and device based on trajectory error estimation to address the above technical problems.

[0007] A wide-angle staring SAR video imaging method based on trajectory error estimation, the method comprising:

[0008] Wide-angle staring SAR is used to receive the original echo signal, which is range-compressed and the entire aperture is divided into multiple overlapping sub-apertures. The BP imaging algorithm is used to perform coarse imaging on the sub-apertures.

[0009] The phase error is estimated by using a back-projection autofocus algorithm according to the coarse imaging image. The sub-aperture trajectory is roughly corrected according to the relationship between the phase error and the slant range error to obtain a coarse correction trajectory.

[0010] According to the coarse correction trajectory, the BP imaging algorithm is used to image the sub-aperture, and the imaging area is divided into multiple sub-areas. The trajectory error is estimated for each sub-area using the back-projection autofocus algorithm and weighted least squares method. The coarse correction trajectory is then accurately compensated to obtain the precise compensated trajectory.

[0011] Based on overlapping sub-apertures, the Kabsch algorithm is used to register the precise compensation trajectories of different sub-apertures and obtain the registered trajectory positions.

[0012] The autoregressive BP imaging algorithm is used to perform video frame imaging according to the registered trajectory position.

[0013] A wide-angle staring SAR video imaging device based on trajectory error estimation, the device comprising:

[0014] The coarse imaging module is used to receive the original echo signal using the wide-angle staring SAR, perform range compression on the original echo signal, divide the entire aperture into multiple overlapping sub-apertures, and use the BP imaging algorithm to perform coarse imaging on the sub-apertures.

[0015] The autofocus coarse correction module is used to estimate the phase error based on the coarse imaging image using the back-projection autofocus algorithm; and to perform coarse correction on the sub-aperture trajectory according to the relationship between the phase error and the slant range error to obtain a coarse correction trajectory.

[0016] The autofocus precise compensation module is used to image the sub-aperture according to the coarse correction trajectory using the BP imaging algorithm, divide the imaging area into multiple sub-areas, estimate the trajectory error for each sub-area using the back-projection autofocus algorithm and weighted least squares method, and accurately compensate the coarse correction trajectory to obtain a precise compensation trajectory.

[0017] The overlapping part registration module is used to register the precise compensation trajectories of different sub-apertures based on the overlapping sub-apertures using the Kabsch algorithm to obtain the registered trajectory position.

[0018] The video frame imaging module is used to perform video frame imaging using an autoregressive BP imaging algorithm according to the registered trajectory position.

[0019] The wide-angle staring SAR video imaging method and device based on trajectory error estimation include: receiving raw echo signals using a wide-angle staring SAR, performing range compression on the raw echo signals, dividing the entire aperture into multiple overlapping sub-apertures, and performing coarse imaging of the sub-apertures using a BP imaging algorithm; estimating phase error using a backprojection autofocusing algorithm based on the coarse image; coarsely correcting the sub-aperture trajectory based on the relationship between phase error and slant range error; imaging the sub-aperture using a BP imaging algorithm based on the coarsely corrected trajectory, dividing the imaging area into multiple sub-areas, estimating trajectory error using a backprojection autofocusing algorithm and weighted least squares method for each sub-area, and accurately compensating the coarsely corrected trajectory; registering the precisely compensated trajectories of different sub-apertures using the Kabsch algorithm based on the overlapping sub-apertures to obtain the registered trajectory position; and performing video frame imaging using an autoregressive BP imaging algorithm based on the registered trajectory position. This method achieves video frame imaging, autofocusing, and stability between video frames under any imaging model. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 1 is a flow chart of a wide-angle staring SAR video imaging method based on trajectory error estimation in one embodiment;

[0021] Figure 2 Schematic diagram of a wide-angle staring SAR video imaging geometric model based on trajectory error estimation in another embodiment;

[0022] Figure 3 A schematic diagram of a video imaging geometric model based on multiple sub-regions in another embodiment;

[0023] Figure 4 An optical image of an experimental scene in another embodiment;

[0024] Figure 5 is the imaging result of the first frame of image in another embodiment, wherein Figure 5 (a) Figure 5 (f) Figure 5 (k) are the imaging results of the measurement trajectory, the imaging results of the rough compensation trajectory, and the imaging results of the precise compensation trajectory, respectively. Figure 5 (b) to Figure 5 (e) Figure 5 (a) The enlarged image of the imaging result of area AD, Figure 5 (g) to Figure 5 (j) Figure 5 (f) corresponds to Figure 5 (a) The enlarged image of the imaging result of area AD, Figure 5 (l) to Figure 5 (o) Figure 5 (k) corresponds to Figure 5 (a) Enlarged image of the AD imaging results in the middle area;

[0025] Figure 6 Schematic diagram of subaperture trajectory before and after compensation in another embodiment, wherein Figure 6 (a) is a schematic diagram of the subaperture trajectory before compensation. Figure 6 (b) is a schematic diagram of the sub-aperture trajectory after compensation;

[0026] Figure 7 Schematic diagram of imaging results before and after sub-aperture trajectory registration in another embodiment, wherein Figure 7 (a) is the schematic diagram before registration. Figure 7 (b) to Figure 7 (e) Figure 7 (a) corresponds to Figure 5 (a) The enlarged image of the imaging result of area AD, Figure 7 (f) is the schematic diagram after registration. Figure 7 (g) to Figure 7 (j) Figure 7 (f) corresponds to Figure 5 (a) Enlarged image of the imaging results of area AD;

[0027] Figure 8 Schematic diagram of imaging results of the 1st and 69th frames of images in another embodiment, wherein Figure 8 (a) is a schematic diagram of the imaging results. Figure 8 (b) to Figure 8 (e) Figure 8 (a) corresponds to Figure 5 (a) Enlarged image of the imaging results of area AD. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0029] This application proposes a wide-angle staring SAR video imaging method based on trajectory error estimation. This method establishes a wide-angle staring SAR imaging model in the presence of trajectory error and divides the entire aperture into multiple overlapping sub-apertures. A time-domain backpropagation algorithm is used to perform coarse imaging of the sub-apertures. Next, for each sub-aperture, a phase error model is established based on the trajectory error to estimate the phase error of each echo. Autofocus back projection (ABP) has been proven to be a practical phase error estimation method. The trajectory error is coarsely compensated based on the relationship between phase error and slant range error. Furthermore, considering the spatially varying nature of trajectory error, the imaging area is divided into multiple sub-areas. For each sub-area, the phase error model is used to estimate the phase error of each echo within that area, thereby obtaining the slant range error from the trajectory to that sub-area. Subsequently, the weighted least squares (WLS) method is used to estimate the slant range error of all sub-areas and accurately compensate for the trajectory error. Finally, based on the overlapping sub-apertures, the Kabsch algorithm is used to connect the subsequent sub-aperture with the previous one. Finally, based on the connected trajectories, the autoregressive BP (ARBP) algorithm is used to perform video frame imaging. This method achieves video frame imaging, self-focusing, and stabilization between video frames under any imaging model.

[0030] In one embodiment, Figure 1 As shown, a wide-angle staring SAR video imaging method based on trajectory error estimation is provided, which includes the following steps:

[0031] Step 100: using a wide-angle staring SAR to receive original echo signals, performing range compression on the original echo signals, dividing the entire aperture into multiple overlapping sub-apertures, and using a BP imaging algorithm to perform coarse imaging on the sub-apertures.

[0032] Step 102: Estimate the phase error using a back-projection autofocus algorithm based on the coarse imaging image; perform a coarse correction on the sub-aperture trajectory based on the relationship between the phase error and the slant range error to obtain a coarse correction trajectory.

[0033] Step 104: Based on the coarse correction trajectory, the sub-aperture is imaged using the BP imaging algorithm, and the imaging area is divided into multiple sub-areas. The trajectory error is estimated for each sub-area using the back-projection autofocus algorithm and the weighted least squares method, and the coarse correction trajectory is accurately compensated to obtain a precise compensated trajectory.

[0034] Step 106: Based on the overlapping sub-apertures, the Kabsch algorithm is used to align the precise compensation trajectories of different sub-apertures to obtain the aligned trajectory positions.

[0035] Step 108: Perform video frame imaging using an autoregressive BP imaging algorithm according to the registered trajectory position.

[0036] The wide-angle staring SAR video imaging method and apparatus based on trajectory error estimation include: receiving raw echo signals using a wide-angle staring SAR, performing range compression on the raw echo signals, dividing the entire aperture into multiple overlapping sub-apertures, and performing coarse imaging of the sub-apertures using a BP imaging algorithm; estimating phase error using a backprojection autofocusing algorithm based on the coarse image; coarsely correcting the sub-aperture trajectory based on the relationship between phase error and slant range error; imaging the sub-aperture using a BP imaging algorithm based on the coarsely corrected trajectory, dividing the imaging area into multiple sub-areas, estimating trajectory error using a backprojection autofocusing algorithm and weighted least squares method for each sub-area, and accurately compensating the coarsely corrected trajectory; registering the precisely compensated trajectories of different sub-apertures using a Kabsch algorithm based on the overlapping sub-apertures to obtain the registered trajectory position; and performing video frame imaging using an autoregressive BP imaging algorithm based on the registered trajectory position. This method achieves video frame imaging, autofocusing, and stability between video frames under any imaging model.

[0037] In one embodiment, step 100 includes: receiving original echo signals using a wide-angle staring SAR, performing range compression on the original echo signals; and dividing the entire aperture into a plurality of overlapping sub-apertures.

[0038] Specifically, the geometric model of wide-angle staring SAR video imaging based on trajectory error estimation is shown in the figure below: Figure 2 As shown, let the fast time be , slow time is , magenta dots are imaging grids , the green dot is the location of the point target .exist At this moment, the actual trajectory is , the trajectory measured by the global positioning system or inertial navigation system is , the radar transmission carrier frequency is , the bandwidth is , the pulse width is The original echo signal is from point The reflected signal is frequency modulated to baseband; the original echo signal is:

[0039] ;

[0040] in, , for The actual trajectory position at the moment to the point target position distance, is the speed of light, is the two-norm.

[0041] After the original echo signal is range-compressed, the echo signal after range compression is obtained:

[0042] ;

[0043] in, Represents the echo signal after compression.

[0044] Let the integration time of the entire aperture be The entire aperture is divided into overlapping subapertures, where The starting time of each sub-aperture is , the end time is . Adjacent sub-apertures need to satisfy .

[0045] make is the overlap ratio of the overlapping sub-apertures. The overlap ratio does not need to be set according to the actual video frame rate and the azimuth resolution of the image, but it is necessary to ensure the stability of the subsequent trajectory connection. .

[0046] According to the measured trajectory position, the actual trajectory position and the position of the point target, the BP imaging algorithm is used to perform coarse imaging on the sub-aperture to obtain a coarse imaging image of each sub-aperture; The expression of the coarse imaging image of a sub-aperture is:

[0047] ;

[0048] ;

[0049] in, express time, Indicates the The coarse imaging image of the sub-aperture, Represents the imaging result of a single echo, express The trajectory position measured at each moment, express The trajectory position measured at each moment to the imaging grid distance, represents the imaging grid, express The distance from the actual trajectory position to the point target position at the moment, represents the position of the point target, express Actual trajectory position at the moment, is the speed of light, is the two-norm, Indicates bandwidth, Indicates the radar transmission carrier frequency, .

[0050] In one embodiment, step 102 includes: estimating a phase error caused by a trajectory error using a backprojection autofocus algorithm based on the coarse image to obtain a phase error estimation result; and determining a slant range error based on the phase error estimation result. The slant range error expression is:

[0051] ;

[0052] in, is the slant range error caused by the phase error, is the phase error estimation result, For the The time corresponding to the emission of a pulse, is the phase unwrapping function.

[0053] The sub-aperture trajectory is roughly corrected according to the slant range error to obtain the rough correction trajectory; The position expression of each sub-aperture trajectory is:

[0054] ;

[0055] in, Indicates the The first sub-aperture The trajectory position of the pulse coarse compensation, Indicates the The trajectory position of each pulse measurement, Indicates the The slant range error caused by the trajectory error of each pulse is Representation Grid center.

[0056] In one embodiment, estimating the phase error caused by the trajectory error using a backprojection autofocus algorithm based on the coarse imaging image to obtain a phase error estimation result includes: estimating the phase error caused by the trajectory error using the backprojection autofocus algorithm based on the coarse imaging image, and determining that the phase error to be estimated is:

[0057] ;

[0058] ;

[0059] in, represents the phase error to be estimated, is a function for evaluating image sharpness, for The number of pulses, 、 Respectively represent The start and end times of the sub-apertures must satisfy , For the The slow time corresponding to each pulse is Represents the imaging result of a single echo; express time, Represents the imaging grid.

[0060] The phase error to be estimated is regarded as a point in multidimensional space and is estimated using the coordinate descent method to obtain the phase error estimation result.

[0061] Specifically, for Actual trajectory position at the moment to the imaging grid Without considering the envelope error, the phase error caused by the trajectory error is:

[0062] ;

[0063] in, is the slant range error caused by the trajectory error. In order to correct the phase error, it is generally assumed that the phase error is independent of the position of the imaging grid, that is:

[0064] ;

[0065] The autofocus algorithm requires an estimation of the phase error, , and compensated into the echo. The imaging result after phase compensation of each sub-aperture is:

[0066] ;

[0067] The ABP algorithm is used to estimate the phase error. Based on the idea of ​​image quality optimization, the ABP algorithm is based on the sharpest image criterion, that is, to find the phase error to be estimated. ; The function for evaluating image sharpness is: .

[0068] The phase error to be estimated is considered as a point in multidimensional space and estimated using coordinate descent. This algorithm is an iterative non-gradient optimization algorithm that searches for local extreme values ​​along the current dimension in each iteration. The result of the iteration is: ;So No. The estimated results are:

[0069] ; By continuously iterating the above formula, the final phase estimation result can be obtained .

[0070] According to the phase error to be estimated The slope range error can be inferred .

[0071] Since the phase error is assumed to be independent of the position of the imaging grid, we can assume that for The measured trajectory and the actual trajectory to the center of the scene at all times The slant range error is calculated. By roughly compensating the trajectory based on the slant range error, the focused imaging result can be obtained. For the The trajectory after the sub-aperture is roughly compensated is The trajectory position of the pulse coarse compensation is as shown in the above The trajectory position expression of the coarse compensation pulse is shown in FIG.

[0072] In one embodiment, step 104 includes: imaging the sub-aperture using the BP imaging algorithm according to the coarse correction trajectory, dividing the imaging result after phase compensation into multiple sub-areas; estimating the phase of each sub-area using the back-projection autofocus algorithm. time Slope range error to the center of the corresponding sub-area; The slant range error estimation result expression corresponding to each sub-area is:

[0073] ;

[0074] ;

[0075] in, Indicates the The slant range error estimation results corresponding to the sub-areas are: represents the coefficient vector, for Moment The trajectory error vector of each sub-aperture, They are Moment The trajectory error vector of each sub-aperture is x 、 y 、 z The value of the direction, Indicates the The coarse compensation trajectory position vector of each sub-aperture, represents the actual trajectory position vector, for time arrive The instantaneous oblique angle of for time arrive pitch angle.

[0076] According to the slant range error estimation results corresponding to all sub-areas, the overdetermined equation for the trajectory error vector is obtained as follows:

[0077] ;

[0078] The weighted least square method is used to solve the overdetermined equation to obtain the trajectory error estimate; the coarse compensation trajectory position is compensated according to the trajectory error estimate to obtain the precise compensation trajectory.

[0079] In one embodiment, the coarsely compensated trajectory position is compensated according to the trajectory error estimate to obtain a precisely compensated trajectory; the precisely compensated trajectory expression is:

[0080] ;

[0081] in, Indicates the The sub-aperture accurately compensates the trajectory vector, Indicates the The trajectory position vector of the sub-aperture coarse compensation, represents the trajectory error vector.

[0082] Specifically, according to the imaging principle of BP imaging algorithm, the trajectory position based on coarse compensation , the echo after distance compression Projected onto the imaging grid, the imaging result of a single echo is obtained, namely:

[0083] ;

[0084] in, For the Track position after sub-aperture coarse compensation To the imaging grid For the first The imaging result is:

[0085] ;

[0086] in, express Imaging result of a single echo at time t.

[0087] When the imaging scene is large or the beam width is large, the slant range error caused by the trajectory error will no longer vary at different grid points. Therefore, it is assumed that will no longer apply. Although based on the assumption The autofocus algorithm can still estimate the phase error well, but the entire imaging scene will be difficult to focus at the same time. Instead, some areas will be focused and some areas will be defocused.

[0088] The schematic diagram of the video imaging geometric model based on multiple sub-regions is as follows Figure 3 The entire imaging area is divided into sub-regions, The center of the sub-region is . Use ABP algorithm to estimate time arrive Slope range error: , the estimated result is At the same time, according to the geometric relationship:

[0089] ;

[0090] in, for Moment The trajectory error vector of each sub-aperture, for time arrive The instantaneous oblique angle of for time arrive pitch angle.

[0091] pass sub-regions, we can obtain the overdetermined equation for the trajectory error vector as shown above. The overdetermined equation for the trajectory error vector can be rewritten as:

[0092] ;

[0093] The trajectory error vector can be obtained by weighted least squares method:

[0094] ;

[0095] in , for according to The variance of the change, is a diagonal matrix. Finally, the trajectory that can make the whole image well focused, that is, the precise compensation trajectory is:

[0096] ;

[0097] in, Indicates the The trajectory of the sub-aperture is accurately compensated.

[0098] In one embodiment, step 106 includes: Hedi The precise compensation trajectory vector of each sub-aperture and the number of pulses of the overlapping trajectory are calculated. Hedi The center of mass of the two trajectories of the sub-aperture; Hedi The center of mass expression of the two trajectories of a sub-aperture is:

[0099] ;

[0100] ;

[0101] in, Indicates the The centroid of the subaperture trajectory, Indicates the The centroid of the subaperture trajectory, for Number of internal pulses, and Respectively Hedi The precise compensation trajectory vector of each sub-aperture, 、 Respectively represent The start and end times of the sub-apertures must satisfy , .

[0102] According to Hedi The precise compensation trajectory vector of each sub-aperture and the corresponding trajectory center of mass are calculated. Hedi The covariance matrix of the accurate compensation trajectory vector of the sub-aperture is obtained; the covariance matrix is ​​subjected to singular value decomposition, and the rotation matrix is ​​estimated based on the decomposition result; based on the estimated result of the rotation matrix, the Hedi The center of mass of the sub-aperture trajectory is obtained to estimate the translation matrix; according to the The accurate compensation trajectory vector of each sub-aperture, the estimated rotation matrix and translation matrix are obtained to complete the registration. The trajectory of the sub-aperture is completed The trajectory expression of each sub-aperture is:

[0103] ;

[0104] in, Indicates the completion of the registration The trajectory of the sub-aperture, represents the estimated rotation matrix, represents the estimated translation matrix.

[0105] According to the trajectories of all registered sub-apertures, the position of the registered trajectory is obtained; the expression of the position of the registered trajectory is:

[0106] ;

[0107] ;

[0108] in, Represents the trajectory position after registration.

[0109] Specifically, for Hedi The sub-apertures perform self-focusing in step 104 and step 106 to obtain the trajectory and .for of satisfy ,for of satisfy , and there are .

[0110] for , and The following conversion relationship exists:

[0111] ;

[0112] Where, is the rotation matrix, is the translation matrix, is the noise vector. 、 Find the center of mass of the two trajectories, where is the number of pulses in the overlapping trajectory. Then calculate and The covariance of is:

[0113] ;

[0114] right Perform singular value decomposition, we have:

[0115] ;

[0116] in, is a diagonal matrix, are two unitary matrices.

[0117] Estimated rotation matrix for:

[0118] ;

[0119] Translation Matrix for:

[0120] ;

[0121] Complete the registration The trajectory of each subaperture is:

[0122] .

[0123] The final trajectory can be obtained from the registered sub-aperture trajectory. The registered trajectory position shown in the above expression of the registered trajectory position is the final trajectory.

[0124] In one embodiment, step 108 includes: using an autoregressive BP imaging algorithm based on the registered trajectory position, projecting the echo signal after range compression onto an imaging grid to obtain an imaging result of a single echo; the imaging result of a single echo is expressed as:

[0125] ;

[0126] in, express Imaging results of a single echo at time t; for Trajectory after time registration To the imaging grid distance, express The actual trajectory position at the moment to the point target position distance, represents the speed of light, Indicates the radar transmission carrier frequency, Indicates bandwidth.

[0127] According to the imaging results of a single echo, we can get The video frame image at the moment, The video frame image expression at the moment is:

[0128] ;

[0129] Where, express The video frame image at the moment, Represents the imaging result of a single echo, is the order of the autoregressive model, are the parameters of the autoregressive model.

[0130] Specifically, according to the imaging principle of the BP algorithm, based on the trajectory position after registration , the echo signal after range compression Projected onto the imaging grid, the imaging result of a single echo is obtained, namely:

[0131] ;

[0132] Video frame image at time As mentioned above The video frame image expression at the moment is shown as follows.

[0133] In one embodiment, the order and parameters of the autoregressive model are determined by equivalent echo coefficients and window functions.

[0134] Specifically, when the window function is a rectangular window, .make ,but ;in, Indicates the equivalent echo number.

[0135] When the window function is Bartlett window, .make , ,but , .

[0136] When the window function is Hanning window, .make , , ,but , , .

[0137] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. In addition, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0138] In a verification example, to verify the effectiveness of the proposed algorithm, the project team conducted a WasSAR video imaging experiment using a Ku-band multi-rotor drone in Qingdao, China. The experimental scene is a village, and the optical image of the scene is as follows: Figure 4 As shown, the imaging scene size is 400m × 400m. The WasSAR system is installed on a multirotor drone. Considering computer memory limitations, the complete echo data is divided into 8192 echoes per subaperture, with an overlap of 75% between adjacent subapertures. In step 104, the entire imaging area is divided into 64 8×8 subregions, each measuring 50m × 50m, along the east and north directions. The window function selected for imaging is the Hanning window.

[0139] First, to verify the effectiveness of steps 100 to 104 in the proposed algorithm, we will use the imaging results of the first frame to compare the imaging results of the trajectory measured by the positioning system, the trajectory after coarse compensation, and the trajectory after fine compensation. The three imaging results correspond to steps 100 to 104, respectively. Figure 5 (a) Figure 5 (f) Figure 5 (k) are schematic diagrams of the imaging results of the measurement trajectory, the rough compensation trajectory, and the precise compensation trajectory. Figure 5 (b) to Figure 5 (e) Figure 5 (a) The enlarged image of the imaging result of area AD, Figure 5 (g) to Figure 5(j) Figure 5 (f) corresponds to Figure 5 (a) The enlarged image of the imaging result of area AD, Figure 5 (l) to Figure 5 (o) Figure 5 (k) corresponds to Figure 5 (a) is an enlarged image of the imaging result of area AD. By comparison, it can be seen that Figure 5 (f) to Figure 5 (j) Image quality ratio Figure 5 (a) to Figure 5 (e) has been significantly improved, proving that coarse compensation can effectively improve image quality, but Figure 5 The image of region A in (g) is still defocused. Figure 5 (k) to Figure 5 (o) The image quality is better than Figure 5 (f) to Figure 5 (j), at the same time Figure 5 The image quality of region A in (l) is significantly better than Figure 5 (g), demonstrating that accurate compensation can achieve focusing of the entire image.

[0140] Next, to verify the effectiveness of step 106 in the proposed algorithm, two adjacent subapertures with a 75% overlap were used to compare the imaging results before and after trajectory registration. The two subapertures correspond to the first and eighth frames of the image, respectively, and the corresponding trajectories have both been precisely compensated through steps 100 to 104. Figure 6 (a) is the subaperture trajectory before compensation, Figure 6 (b) is the sub-aperture trajectory after compensation. Figure 7 (a) is based on Figure 6 (a) Schematic diagram of the imaging results of the trajectory, where the magenta imaging result corresponds to the blue trajectory and the green imaging result corresponds to the red trajectory. Figure 7 (b) to Figure 7 (e) Figure 7 (a) corresponds to Figure 5 (a) A magnified image of the imaging results of area AD. It can be seen that there is a significant offset between the imaging results of the two sub-apertures. Figure 7 (f) is based on Figure 6 (b) Imaging results of the trajectory, Figure 7 (g) to Figure 7 (j) Figure 7 (f) corresponds to Figure 5 (a) A magnified image of the imaging result of area AD. By comparison, it can be seen that after trajectory registration, there is no offset between the two sub-aperture images, proving that trajectory registration can achieve stable image sequences.

[0141] Finally, in order to verify the effectiveness of the proposed algorithm, we use the 1st frame image and the 69th frame image for verification. Figure 8 (a) is the imaging result of two frames of images, where the magenta image is the 1st frame and the green image is the 69th frame; Figure 8 (b) to Figure 8 (e) Figure 8 (a) corresponds to Figure 5 (a) A magnified image of the imaging result of region AD. It can be seen that the images of frames 1 and 69 are well focused and have no offset, demonstrating the effectiveness of the proposed algorithm in imaging, motion compensation, and video frame stabilization.

[0142] In one embodiment, a wide-angle staring SAR video imaging device based on trajectory error estimation is provided, comprising: a coarse imaging module, a coarse autofocus correction module, a precise autofocus compensation module, an overlapping portion registration module, and a video frame imaging module, wherein:

[0143] The coarse imaging module is used to receive the original echo signal using the wide-angle staring SAR, perform range compression on the original echo signal, divide the entire aperture into multiple overlapping sub-apertures, and use the BP imaging algorithm to perform coarse imaging on the sub-apertures.

[0144] The autofocus coarse correction module is used to estimate the phase error based on the coarse imaging image using the back-projection autofocus algorithm; and to perform coarse correction on the sub-aperture trajectory according to the relationship between the phase error and the slant range error to obtain a coarse correction trajectory.

[0145] The autofocus precise compensation module is used to image the sub-aperture according to the coarse correction trajectory using the BP imaging algorithm, divide the imaging area into multiple sub-areas, estimate the trajectory error for each sub-area using the back-projection autofocus algorithm and weighted least squares method, and accurately compensate the coarse correction trajectory to obtain a precise compensation trajectory.

[0146] The overlapping part registration module is used to register the precise compensation trajectories of different sub-apertures based on the overlapping sub-apertures using the Kabsch algorithm to obtain the registered trajectory position.

[0147] The video frame imaging module is used to perform video frame imaging using an autoregressive BP imaging algorithm according to the registered trajectory position.

[0148] In one embodiment, the coarse imaging module is further used to receive the original echo signal using a wide-angle staring SAR and perform range compression on the original echo signal; divide the entire aperture into multiple overlapping sub-apertures; wherein the overlap ratio of the overlapping sub-apertures is as shown in the above overlap ratio expression.

[0149] According to the measured trajectory position, the actual trajectory position and the position of the point target, the BP imaging algorithm is used to perform coarse imaging on the sub-aperture to obtain a coarse imaging image of each sub-aperture; The coarse imaging image of each sub-aperture is shown in the above The coarse imaging image expression of each sub-aperture is shown in FIG.

[0150] In one embodiment, the autofocus coarse correction module is further configured to estimate the phase error caused by the trajectory error using a back-projection autofocus algorithm based on the coarse imaging image to obtain a phase error estimation result; add the phase error estimation result to the echo to obtain an imaging result of each sub-aperture after phase compensation; and determine the slant range error as shown in the above slant range error expression based on the phase error estimation result. The sub-aperture trajectory is coarsely corrected based on the slant range error to obtain a coarse correction trajectory; The trajectory position of the pulse coarse compensation is as shown in the above The trajectory position expression of the coarse compensation pulse is shown in FIG.

[0151] In one embodiment, estimating a phase error caused by a trajectory error using a backprojection autofocus algorithm based on a coarse image to obtain a phase error estimation result includes: estimating the phase error caused by the trajectory error using the backprojection autofocus algorithm based on the coarse image to determine a phase error to be estimated as shown in the above-mentioned expression for the phase error to be estimated. Treating the phase error to be estimated as a point in a multidimensional space and estimating it using a coordinate descent method to obtain a phase error estimation result.

[0152] In one embodiment, the autofocus precise compensation module is further used to image the sub-aperture using the BP imaging algorithm according to the coarse correction trajectory, and divide the imaging result after phase compensation into multiple sub-areas; and estimate the phase of each sub-area using the back-projection autofocus algorithm. time Slope range error to the center of the corresponding sub-area; The slant range error estimation results corresponding to the sub-area are shown in the above The slant range error estimation results for each sub-region are shown in the expression. Based on the slant range error estimation results for all sub-regions, the overdetermined equation for the trajectory error vector is obtained, as shown above. The overdetermined equation is solved using the weighted least squares method to obtain the trajectory error estimate. The coarsely compensated trajectory position is compensated based on the trajectory error estimate to obtain the precisely compensated trajectory.

[0153] In one embodiment, the autofocus precise compensation module is further configured to compensate the coarsely compensated trajectory position according to the trajectory error estimation value to obtain a precise compensation trajectory as shown in the precise compensation trajectory expression above.

[0154] In one embodiment, the overlapping portion registration module is further configured to: Hedi The precise compensation trajectory vector of each sub-aperture and the number of pulses of the overlapping trajectory are calculated as above. Hedi The centroid expression of the two trajectories of the sub-aperture is shown in the Hedi The center of mass of the two trajectories of the sub-aperture; according to the Hedi The precise compensation trajectory vector of each sub-aperture and the corresponding trajectory center of mass are calculated. Hedi The covariance matrix of the accurate compensation trajectory vector of the sub-aperture is obtained; the covariance matrix is ​​subjected to singular value decomposition, and the rotation matrix is ​​estimated based on the decomposition result; based on the estimated result of the rotation matrix, the Hedi The center of mass of the sub-aperture trajectory is obtained to estimate the translation matrix; according to the The accurate compensation trajectory vector of each sub-aperture, the estimated rotation matrix and translation matrix are obtained as described above to complete the registration. The trajectory expression of the sub-aperture is shown as the first According to the trajectories of all the registered sub-apertures, the registered trajectory position is obtained as shown in the above expression of the registered trajectory position.

[0155] In one embodiment, the video frame imaging module is further configured to project the echo signal after distance compression onto the imaging grid using an autoregressive BP imaging algorithm according to the registered trajectory position, and obtain an imaging result of a single echo as shown in the above-mentioned imaging result expression of a single echo; and obtain an imaging result of a single echo as shown in the above-mentioned imaging result expression according to the imaging result of the single echo. The video frame image expression at the moment is shown as The video frame image at the moment.

[0156] In one embodiment, the order and parameters of the autoregressive model in the video frame imaging module are determined by the equivalent echo coefficient and the window function.

[0157] Regarding the specific limitations of the wide-angle staring SAR video imaging device based on trajectory error estimation, please refer to the limitations of the wide-angle staring SAR video imaging method based on trajectory error estimation above, and will not be repeated here. Each module in the above-mentioned wide-angle staring SAR video imaging device based on trajectory error estimation can be implemented in whole or in part through software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each of the above modules.

[0158] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0159] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A wide-angle staring SAR video imaging method based on trajectory error estimation, characterized in that: The method comprises: A wide-angle staring SAR is used to receive the original echo signal, the original echo signal is range-compressed, the entire aperture is divided into multiple overlapping sub-apertures, and the sub-apertures are coarsely imaged using a BP imaging algorithm; The phase error is estimated using a back-projection autofocus algorithm based on the coarse imaging image. Based on the relationship between the phase error and the slant range error, the sub-aperture trajectory is roughly corrected to obtain a coarse correction trajectory. According to the coarse correction trajectory, the sub-aperture is imaged using a BP imaging algorithm, the imaging area is divided into multiple sub-areas, a back-projection autofocus algorithm and a weighted least squares method are used to estimate the trajectory error for each sub-area, and the coarse correction trajectory is accurately compensated to obtain a precise compensation trajectory; Based on overlapping sub-apertures, the precise compensation trajectories of different sub-apertures are registered using the Kabsch algorithm to obtain the registered trajectory positions; The autoregressive BP imaging algorithm is used to perform video frame imaging according to the registered trajectory position.

2. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 1, characterized in that: A wide-angle staring SAR is used to receive the original echo signal, and the original echo signal is range-compressed. The entire aperture is divided into multiple overlapping sub-apertures, and the sub-apertures are coarsely imaged using a BP imaging algorithm, including: Using a wide-angle staring SAR to receive original echo signals, and performing distance compression on the original echo signals; The entire aperture is divided into multiple overlapping sub-apertures; the overlap ratio of the overlapping sub-apertures is: in, represents the overlap ratio of overlapping subapertures, 、 Respectively represent The start and end times of the sub-apertures must satisfy ; According to the measured trajectory position, the actual trajectory position and the position of the point target, the BP imaging algorithm is used to perform coarse imaging on the sub-aperture to obtain a coarse imaging image of each sub-aperture; The coarse imaging image of each sub-aperture is: in, express time, Indicates the The coarse imaging image of the sub-aperture, Represents the imaging result of a single echo, express The trajectory position measured at each moment, express The trajectory position measured at each moment to the imaging grid distance, represents the imaging grid, express The distance from the actual trajectory position to the point target position at the moment, represents the position of the point target, express Actual trajectory position at the moment, is the speed of light, is the two-norm, Indicates bandwidth, Indicates the radar transmission carrier frequency.

3. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 1, characterized in that: The phase error is estimated using a back-projection autofocus algorithm based on the coarse imaging image. Based on the relationship between the phase error and the slant range error, the sub-aperture trajectory is roughly corrected to obtain a coarse correction trajectory, including: The phase error caused by the trajectory error is estimated using the back-projection autofocus algorithm based on the coarse imaging image to obtain the phase error estimation result; According to the phase error estimation result, the slant range error is determined as: in, is the slope range error caused by trajectory error, is the phase error estimation result, For the The moment the echo is emitted, is the phase unwrapping function; The sub-aperture trajectory is roughly corrected according to the slant range error to obtain a roughly corrected trajectory; The trajectory position of the pulse coarse compensation is: in, Indicates the Subaperture No. The trajectory position of the pulse coarse compensation, Indicates the The trajectory position of each pulse measurement, Indicates the The slant range error caused by the trajectory error of each pulse is Representing the imaging grid center.

4. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 3, characterized in that: The phase error caused by the trajectory error is estimated using the back-projection autofocus algorithm based on the coarse imaging image, and the phase error estimation results are obtained, including: The phase error caused by the trajectory error is estimated using the back-projection autofocus algorithm based on the coarse imaging image, and the phase error to be estimated is determined as: in, represents the phase error to be estimated, is a function for evaluating image sharpness, for The number of pulses, 、 Respectively represent The start and end times of the sub-apertures must satisfy , For the The slow time corresponding to each pulse is Represents the imaging result of a single echo; express time, represents the imaging grid; The phase error to be estimated is regarded as a point in multidimensional space and is estimated using the coordinate descent method to obtain the phase error estimation result.

5. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 1, characterized in that: According to the coarse correction trajectory, the sub-aperture is imaged using the BP imaging algorithm, the imaging area is divided into multiple sub-areas, the trajectory error is estimated for each sub-area using the back-projection autofocus algorithm and the weighted least squares method, and the coarse correction trajectory is accurately compensated to obtain an accurate compensation trajectory, including: According to the rough correction trajectory, the sub-aperture is imaged using the BP imaging algorithm, and the imaging results after rough trajectory correction are divided into sub-regions; The back-projection autofocus algorithm is used to estimate the time Slope range error to the center of the corresponding sub-area; The slant range error estimation result corresponding to each sub-area is: in, Indicates the The slant range error estimation results corresponding to the sub-areas are: represents the coefficient vector, for Moment The trajectory error vector of each sub-aperture, They are Moment The trajectory error vector of each sub-aperture is x 、 y 、 z The value of the direction, Indicates the The trajectory position vector of the sub-aperture coarse compensation, represents the actual trajectory position vector, for time arrive The instantaneous oblique angle of for time arrive Pitch angle; According to the slant range error estimation results corresponding to all sub-areas, the overdetermined equation for the trajectory error vector is obtained as follows: The weighted least squares method is used to solve the overdetermined equation to obtain the trajectory error estimate ; The coarsely compensated trajectory position is compensated according to the trajectory error estimate to obtain a precisely compensated trajectory.

6. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 5, characterized in that: The coarse compensation trajectory position is compensated according to the trajectory error estimation value to obtain the precise compensation trajectory: in, Indicates the The sub-aperture accurately compensates the trajectory vector, Indicates the The trajectory position vector of the sub-aperture coarse compensation, represents the trajectory error vector.

7. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 1, characterized in that: Based on overlapping sub-apertures, the precise compensation trajectories of different sub-apertures are registered using the Kabsch algorithm to obtain the registered trajectory positions, including: According to Hedi The precise compensation trajectory vector of each sub-aperture and the number of pulses of the overlapping trajectory are calculated. Hedi The centroid of the two trajectories of the subaperture is: in, for Number of internal pulses, and Respectively Hedi The precise compensation trajectory vector of each sub-aperture, Indicates the The centroid of the subaperture trajectory, Indicates the The centroid of the subaperture trajectory, and The center of mass 、 Respectively represent The start and end times of the sub-apertures must satisfy , ; According to Hedi The precise compensation trajectory vector of each sub-aperture and the corresponding trajectory center of mass are calculated. Hedi The covariance matrix of the accurate compensation trajectory vector of each sub-aperture; Perform singular value decomposition on the covariance matrix and estimate the rotation matrix based on the decomposition result ; According to the estimated results of the rotation matrix, Hedi The center of mass of the sub-aperture trajectory is obtained to estimate the translation matrix ; According to The accurate compensation trajectory vector of each sub-aperture, the estimated rotation matrix and translation matrix are obtained to complete the registration. The trajectory of each subaperture is: in, Indicates the completion of the registration The trajectory of the sub-aperture, represents the estimated rotation matrix, represents the estimated translation matrix; According to the trajectories of all sub-apertures registered, the trajectory position after registration is obtained as: in, Represents the trajectory position after registration.

8. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 1, characterized in that: The autoregressive BP imaging algorithm is used to perform video frame imaging based on the registered trajectory position, including: According to the registered trajectory position, the autoregressive BP imaging algorithm is used to project the echo signal after range compression onto the imaging grid, and the imaging result of a single echo is obtained as follows: in, express Imaging results of a single echo at time; for Trajectory after time registration To the imaging grid distance, express The distance from the actual trajectory position at the moment to the point target position, represents the speed of light, Indicates the radar transmission carrier frequency, Indicates bandwidth, According to the imaging results of a single echo, we can get The video frame image at the moment is: Where, express The video frame image at the moment, Represents the imaging result of a single echo, is the order of the autoregressive model, are the parameters of the autoregressive model.

9. The wide-angle staring SAR video imaging method based on trajectory error estimation according to claim 8, characterized in that: The order and parameters of the autoregressive model are determined by the equivalent echo coefficient and window function.

10. A wide-angle staring SAR video imaging device based on trajectory error estimation, characterized in that: The device comprises: A coarse imaging module is used to receive original echo signals using a wide-angle staring SAR, perform range compression on the original echo signals, divide the entire aperture into multiple overlapping sub-apertures, and perform coarse imaging on the sub-apertures using a BP imaging algorithm; The autofocus coarse correction module is used to estimate the phase error based on the coarse imaging image using the back-projection autofocus algorithm; based on the relationship between the phase error and the slant range error, the sub-aperture trajectory is coarsely corrected to obtain a coarse correction trajectory; an autofocus precision compensation module, configured to image the subaperture using a BP imaging algorithm according to the coarse correction trajectory, divide the imaging area into a plurality of sub-areas, estimate the trajectory error using a back-projection autofocus algorithm and a weighted least squares method for each sub-area, and precisely compensate the coarse correction trajectory to obtain a precisely compensated trajectory; An overlapping part registration module, configured to register the precise compensation trajectories of different sub-apertures using a Kabsch algorithm based on overlapping sub-apertures to obtain a registered trajectory position; The video frame imaging module is used to perform video frame imaging using an autoregressive BP imaging algorithm according to the registered trajectory position.

Images