Terahertz ground moving target recalibration imaging method, device, equipment and medium
Through the multi-constraint processing technology of terahertz band video synthesis aperture radar, the azimuth scale variation problem in ground motion target imaging and recalibration is solved, and efficient recalibration and accurate imaging of curve motion targets is achieved. It is suitable for Dynamic city reconnaissance and surveillance.
Patent Information
- Application Number
- CN202411308056.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-19
- Publication Date
- 2024-12-17
- Estimated Expiration
- 2044-09-19
AI Technical Summary
In the prior art, in dynamic urban reconnaissance and moving target indication, there are difficulties in imaging and recalibration of ground moving targets, especially the azimuth scale variation caused by curved moving targets, resulting in blurring of targets and low imaging quality.
The terahertz band video synthesis aperture radar is used to obtain the original echo data, and through aperture division and multi-constraint processing, including distance-dimensional compensation, azimuth focus and self-focus reconstruction, combining the minimum image entropy and target geometric constraints, the efficient dynamic target is achieved. Recalibrate.
It achieves efficient and robust recalibration of non-cooperative sports goals, adaptively estimates motion errors, reduces system complexity, improves imaging quality and accuracy, and is suitable for various motion modes and background interference.
Smart Images
Figure CN118818500B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of radar signal processing technology, and in particular to a terahertz ground moving target recalibration imaging method, device, equipment and medium. Background Art
[0002] Synthetic Aperture Radar (SAR) can obtain high-quality images of static areas at almost any time, weather and lighting conditions. The demand for dynamic situational awareness and moving target detection has driven the development of SAR, which has led to the emergence of ViSAR. ViSAR expands the temporal resolution of images under given imaging resolution conditions, so that areas of interest can be monitored through image sequences. Facing high-resolution radar imaging scenarios with urgent applications, ViSAR is widely used in dynamic urban reconnaissance and surveillance, as well as moving target indication and imaging.
[0003] Ground moving targets are usually blurred in SAR images, but they carry high-value / high-threat information for target identification and tracking. The motion models of moving targets can generally be divided into two categories. The first category is defined as linear motion, such as uniform speed or uniform acceleration motion. The second category refers to curved motion, such as turning, U-turn, etc. The basis for this classification is that the former usually does not change the azimuth dimension of the moving target, while when the moving target is in the latter, the relative rotation will cause additional azimuth dimension scale changes, so that the target is severely stretched in the azimuth dimension. Therefore, imaging and recalibration of ground moving targets are of urgent significance. Summary of the invention
[0004] Based on this, it is necessary to provide a terahertz ground moving target recalibration imaging method, device, equipment and medium with efficient and robust characteristics to address the above technical problems.
[0005] A terahertz ground moving target recalibration imaging method, the method comprising:
[0006] Acquiring original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in a terahertz frequency band;
[0007] Performing aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0008] Extracting a moving target signal from the sub-aperture image, performing distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then performing self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraint to obtain a moving target focused image;
[0009] splicing the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated;
[0010] According to all the sub-aperture images that have completed the target recalibration, a complete target recalibration sequence SAR image is obtained.
[0011] In one embodiment, the moving target in the sub-aperture image includes one or more moving targets.
[0012] In one embodiment, when extracting moving target signals, clutter suppression is also performed on the sub-aperture image;
[0013] Among them, clutter suppression is performed by using any one of the methods, or a combination of multiple methods, including space-time adaptive processing methods, Doppler filtering methods, adaptive clutter cancellation methods, image processing methods, constant false alarm rate detection methods or principal component analysis methods.
[0014] In one embodiment, when performing range dimension compensation on the moving target signal, range movement alignment is used as a constraint condition in the range compression domain.
[0015] In one embodiment, when performing distance dimension compensation on the moving target signal, the constraint condition is expressed as:
[0016] ;
[0017] In the above formula, represents the estimated radial velocity, represents the radial velocity, represents the new wave number, Indicates that the distance movement is aligned. Indicates along the range beam The inverse Fourier transform of is the wavenumber domain signal of the moving target, For slow time, is the phase compensation function obtained based on the radial velocity of the moving target.
[0018] In one embodiment, after performing distance dimension compensation, when performing azimuth dimension focusing on the moving target signal, a self-focusing constraint condition is constructed based on minimum image entropy, and is solved using an ADMM algorithm.
[0019] In one embodiment, the self-focusing constraint is expressed as:
[0020] ;
[0021] In the above formula, Indicates the initial moving target focusing signal, is the distance profile matrix Stacked in rows, is the subaperture image matrix Stacked in rows, represents the phase error matrix, represents the Fourier transform matrix, represents the regularization coefficient.
[0022] In one embodiment, when the self-focusing constraint condition is iteratively solved using the ADMM algorithm, it is decomposed into three sub-problems, and the three sub-problems are alternately solved in each iteration. The three sub-problems are expressed as follows:
[0023] ;
[0024] In the above formula, represents the Lagrange multiplier domain penalty coefficient, represents the complex soft threshold function, the superscript represents the number of iterations, Expressed as Lagrange multipliers in matrix form, represents the target image in matrix form, Represents the auxiliary variables of the ADMM algorithm in matrix form.
[0025] In one embodiment, the preliminary moving target focusing signal is reconstructed by self-focusing based on the moving target geometric deformation constraint, and the moving target geometric constraint is expressed as:
[0026] ;
[0027] In the above formula, represents the moving target focused image, Represents the target geometry constraint.
[0028] In one embodiment, when extracting the moving target signal in the sub-aperture image:
[0029] After the sub-aperture image is subjected to static area imaging processing, a defocused region of interest related to the moving target is obtained based on a moving target model and imaging feature analysis, and moving target signals are extracted based on the defocused region of interest.
[0030] In one embodiment, when the moving target focused image and the corresponding sub-aperture image are stitched together:
[0031] After performing static area imaging processing on the sub-aperture image, a residual background image is obtained;
[0032] The moving target focused image is stitched with the corresponding residual background image to obtain a recalibrated sub-aperture image of the target.
[0033] The present application also provides a terahertz ground moving target recalibration imaging device, the device comprising:
[0034] The original echo data acquisition module is used to acquire the original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in the terahertz frequency band;
[0035] A sub-aperture image obtaining module, used for performing aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0036] A moving target refocusing module under a three-constraint framework is used to extract the moving target signal in the sub-aperture image, perform distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then perform self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraints to obtain a moving target focused image;
[0037] A sub-aperture image reconstruction module is used to splice the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated;
[0038] The complete sequence SAR image acquisition module is used to obtain a complete target recalibrated sequence SAR image based on all sub-aperture images that have completed target recalibration.
[0039] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:
[0040] Acquiring original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in a terahertz frequency band;
[0041] Performing aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0042] Extracting a moving target signal from the sub-aperture image, performing distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then performing self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraint to obtain a moving target focused image;
[0043] splicing the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after target recalibration;
[0044] According to all the sub-aperture images that have completed the target recalibration, a complete target recalibration sequence SAR image is obtained.
[0045] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:
[0046] Acquiring original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in a terahertz frequency band;
[0047] Performing aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0048] Extracting a moving target signal from the sub-aperture image, performing distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then performing self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraint to obtain a moving target focused image;
[0049] splicing the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated;
[0050] According to all the sub-aperture images that have completed the target recalibration, a complete target recalibration sequence SAR image is obtained.
[0051] The above-mentioned terahertz ground moving target recalibration imaging method, device, equipment and medium divide the original echo data obtained by detecting the ground area by the video synthetic aperture radar in the terahertz frequency band into apertures to obtain multiple sub-aperture images, extract the moving target signal in the sub-aperture image, compensate the moving target signal in the distance dimension and focus the azimuth dimension through two constraints in turn, obtain the preliminary moving target focusing signal, and then reconstruct the preliminary moving target focusing signal by self-focusing based on the moving target geometric constraint to obtain the moving target focusing image, splice the moving target focusing image with the corresponding sub-aperture image to obtain the sub-aperture image after the target is recalibrated, and finally obtain a complete target recalibration sequence SAR image according to all the sub-aperture images that have completed the target recalibration. The method can adaptively estimate the motion error and residual airborne motion error from non-cooperative moving targets, without the need to additionally consider the motion model of unknown targets, and at the same time does not increase the complexity of the system, and has the characteristics of high efficiency and robustness. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] Figure 1 A schematic diagram of a process of a terahertz ground moving target recalibration imaging method in one embodiment;
[0053] Figure 2 A schematic flow chart of the steps of a terahertz ground moving target recalibration imaging method in another embodiment;
[0054] Figure 3 A schematic diagram of an experimental scene in an experiment;
[0055] Figure 4 A schematic diagram of the large scene imaging results of the outdoor terahertz ViSAR in an experiment;
[0056] Figure 5 This is a schematic diagram of the range compression domain imaging results before and after walking compensation in an experiment, where: Figure 5 (a) is a schematic diagram of the range compression domain imaging results before walking compensation. Figure 5 (b) is a schematic diagram of the range compression domain imaging results after movement compensation;
[0057] Figure 6 The original and proposed imaging results of moving target 1 and moving target 2 in an experiment, where: Figure 6 (a) is a schematic diagram showing the original imaging results. Figure 6 (b) Schematic diagram showing the proposed imaging results;
[0058] Figure 7 This is a visual salient feature map of an armored vehicle turning around in an experiment in 5 consecutive frames, where: Figure 7 (a) and Figure 7 (b) are schematic diagrams of the original imaging result and the imaging result after motion compensation, respectively. Figure 7 (c) is a schematic diagram of the imaging results after constraints 1 and 2. Figure 7 (d) is a schematic diagram of the imaging results after three constraints;
[0059] Figure 8 It is a structural block diagram of a terahertz ground moving target recalibration imaging device in one embodiment;
[0060] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0061] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0062] In the radar mission of ground moving target indication and imaging, there are challenges including: non-cooperative moving targets may have arbitrary motion patterns, so different imaging processing methods are required. Especially under ViSAR continuous observation, the motion pattern of the same moving target will change as the radar line of sight changes. The signal-to-noise ratio of a moving target may vary significantly depending on the target speed and the scene it is in. For example, this phenomenon is common in narrow beam coverage areas in the terahertz band. Under long dwell times in the microwave band, the motion pattern of MTs is difficult to keep single and fixed, which adds additional processing complexity. Therefore, the existing technology still lacks robustness for moving targets with different motion patterns or low signal-to-noise ratio situations.
[0063] In response to the above problems, Figure 1 As shown, a terahertz ground moving target recalibration imaging method is provided, comprising the following steps:
[0064] Step S100, obtaining original echo data, where the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in the terahertz frequency band.
[0065] Step S110, performing aperture division on the original echo data to obtain a plurality of sub-aperture images.
[0066] Step S120, extract the moving target signal in the sub-aperture image, perform distance dimension compensation and azimuth dimension focusing on the moving target signal through two constraints in turn, obtain a preliminary moving target focusing signal, and then perform self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraints to obtain a moving target focused image.
[0067] Step S130 , stitching the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated.
[0068] Step S140: obtaining a complete target recalibrated sequence of SAR images based on all sub-aperture images that have completed target recalibration.
[0069] In this embodiment, an efficient method for imaging and recalibrating ground moving targets is provided. The refocusing process of the moving target is refined into three constraints: Constraint 1 is the distance movement correction, which completes the refocusing of the distance dimension; Constraint 2 is the minimum image entropy constraint, which completes the refocusing of the azimuth dimension; Constraint 3 uses the target shape as a constraint to complete the recalibration of the azimuth dimension of the moving target. Based on these three constraints, the motion error and residual airborne motion error from non-cooperative moving targets can be adaptively estimated without additionally considering the motion model of unknown targets. At the same time, the system complexity is not increased, and it has the characteristics of high efficiency and robustness.
[0070] Furthermore, the method processes the raw echo data obtained by continuous detection of a ground area by video synthetic aperture radar in the terahertz frequency band. Compared with the microwave band, terahertz ViSAR has an impressive low-latency (0.2s) imaging capability at an azimuth resolution of 0.2m, which can ensure that the motion pattern of the moving target is as relatively fixed as possible during the dwell time. Therefore, the range unit migration caused by the radial velocity of the moving target is manifested as range walk, and the constraint 1 of range walk alignment is constructed. It should be noted that the range unit migration correction does not operate directly in the complete echo, but in the echo obtained by inverting the defocused region of interest data, thereby allowing the simultaneous processing of multiple moving targets and reducing the computational burden. Subsequently, in view of the inherent sparsity of MTs at high frequencies, an ADMM framework based on the minimum image entropy as constraint 2 is established to adaptively compensate for the phase error caused by the azimuth velocity. At this point, the focusing of the moving target has been basically completed. However, in some motion modes, including turning and U-turning, the azimuth dimension may show obvious geometric distortion. To this end, an azimuth recalibration with the target geometry as constraint 3 is designed by pre-establishing an azimuth broadening coefficient table that considers the rotation radius, target speed, and rotation center.
[0071] In step S100, a video synthetic aperture radar continuously detects a ground area in a terahertz frequency band to obtain SAR echo signals arranged in time sequence.
[0072] In step S110, after aperture segmentation is performed on the original echo data, a plurality of continuous sub-aperture images, namely SAR images, are obtained, each of which includes one or more moving targets.
[0073] In step S120, when extracting the moving target signal, clutter suppression is also performed on the sub-aperture image, wherein clutter suppression is performed by using any one of a space-time adaptive processing method, a Doppler filtering method, an adaptive clutter cancellation method, an image processing method, a constant false alarm rate detection method or a principal component analysis method, or a combination of multiple methods.
[0074] In this embodiment, when extracting the moving target signal in the sub-aperture image: after the sub-aperture image is subjected to static area imaging processing, a defocused region of interest related to the moving target is obtained based on the moving target model and imaging feature analysis, and the moving target signal is extracted based on the defocused region of interest.
[0075] In this embodiment, in the subsequent target reconstruction under the three-constraint framework, only the moving target signals extracted from each sub-aperture image are targeted. When there are multiple moving target signals, a separate target reconstruction is performed under the three constraints for each moving target signal to obtain a moving target focused image corresponding to each moving target. In this way, instead of directly processing in the complete echo data, the region of interest (RoI) data containing the defocused moving target is detected and inverted into the echo, which can reduce the amount of calculation.
[0076] In this embodiment, when performing range dimension compensation on moving target signals, range movement alignment is used as a constraint condition in the range compression domain.
[0077] Specifically, since this method is aimed at the demultiplexing frequency modulation signal of the terahertz system, for a stationary point target, its received signal can be expressed as:
[0078] (1)
[0079] In formula (1), represents the signal amplitude, and Represent the distance frequency and carrier frequency respectively. Indicates the signal bandwidth. and Represent the rectangular window and azimuth window respectively, represents the speed of light. represents the linear frequency modulation distance, which satisfies:
[0080] (2)
[0081] In formula (2), Indicates the shortest distance between the target and the aircraft trajectory, Represents the reference distance, which is pre-selected and fixed. It is expressed as the lower viewing angle. The azimuth satisfies In fact, the distance history of a moving target is different from that of a stationary target. Taking uniform motion as an example, it can be expressed as:
[0082] (3)
[0083] In formula (3), and are the azimuth and radial velocities, respectively. Substituting formula (3) into formula (1) and converting it to the wavenumber domain, we can obtain:
[0084] (4)
[0085] In formula (4), represents the new wave number, and the frequency is It can be seen that the radial velocity will cause the first-order phase error, which is mainly the range walk and azimuth offset, while the azimuth velocity will cause the second-order phase error, resulting in azimuth broadening.
[0086] On the basis of the above, the rationality of range migration being expressed as range walk under terahertz ViSAR is further analyzed.
[0087] In the real world, the speed of a typical ground moving target usually does not exceed 40 m / s, and the time required to reach 100 km / h is usually , corresponding to the maximum acceleration When the range resolution is 0.2m and the dwell time is 0.2s, it can be inferred that when the radial velocity exceeds 1m / s, the range unit migration of the moving target will exceed one range unit. Therefore, the moving target will inevitably produce range migration. When the maximum value of the second-order range reaches one range unit, the required acceleration is approximately , indicating that the second-order distance can generally be ignored.
[0088] In summary, the range unit migration of moving targets under terahertz ViSAR is mainly manifested as range walk. Therefore, the range walk alignment is used as a constraint condition in the range compression domain. When the compensation process, that is, the range dimension compensation of the moving target signal, is performed, the constraint condition can be constructed as follows:
[0089] (5)
[0090] in, ;
[0091] In formula (5), represents the estimated radial velocity, represents the radial velocity, represents the new wave number, Indicates that the distance movement is aligned. Indicates along the range beam The inverse Fourier transform of is the wavenumber domain signal of the moving target, For slow time, is the phase compensation function obtained based on the radial velocity of the moving target. In fact, the radial velocity of the moving target can be roughly calculated according to the slope of the inclined line in the range compression domain, which can speed up the processing efficiency.
[0092] In this embodiment, after performing distance dimension compensation, when performing azimuth dimension focusing on the moving target signal, a self-focusing constraint condition is constructed based on the minimum image entropy, and the ADMM algorithm is used to solve it.
[0093] Specifically, unlike range dimension compensation, azimuth dimension focusing needs to consider the phase error caused by azimuth velocity, the potential high-order motion of the moving target, and the residual airborne motion error in the terahertz band. The necessity of the latter two errors can be explained as follows: First, robust moving target imaging should adaptively consider the phase error caused by high-order motion models such as acceleration, rather than being limited to the uniform motion model described above. Second, in the terahertz frequency band, terahertz ViSAR requires submillimeter motion compensation accuracy, which is difficult to meet with existing inertial navigation systems or global positioning systems, and may produce residual airborne motion errors. Considering the sparsity of moving targets in SAR images, a self-focusing framework based on minimum image entropy is designed through ADMM.
[0094] Furthermore, after distance travel compensation, the distance profile of the moving target can be written as:
[0095] (6)
[0096] In formula (6), represents the azimuth phase error. Further, the range profile can be discretized as:
[0097] (7)
[0098] In formula (7), and are the coefficients of the pulse and range cells, respectively. Therefore, the range profile can be rewritten as:
[0099] (8)
[0100] In formula (8), is the distance profile matrix Stacked along the columns, that is to say Similarly, is the sub-aperture SAR image matrix Stacked in rows. represents the phase error matrix, where is the phase error vector and is the docking matrix. is the Fourier transform matrix, where the mth column is . is stacking noise.
[0101] Then, considering the sparsity of the moving target, the optimization problem based on the L1 norm, that is, the self-focus constraint condition can be established as:
[0102] (9)
[0103] In formula (9), Indicates the initial moving target focusing signal, is the distance profile matrix Stacked in rows, is the subaperture image matrix Stacked in rows, represents the phase error matrix, represents the Fourier transform matrix, represents the regularization coefficient.
[0104] Furthermore, formula (9) can be efficiently solved by ADMM, which introduces auxiliary variables , and transform the optimization problem (Formula (9)) into:
[0105] (10)
[0106] Its augmented Lagrangian function can be expressed as:
[0107] (11)
[0108] In formula (11), and Respectively represent the Lagrange multiplier domain penalty coefficient. The solution to the above problem can be obtained by taking partial derivatives of each variable separately, decomposing it into three sub-problems, and solving these three sub-problems alternately in each iteration. The three sub-problems are expressed as:
[0109] (12)
[0110] In formula (12), Represents the complex soft threshold function.
[0111] Considering , and convert formula (12) into matrix form, expressed as:
[0112] (13)
[0113] In formula (13), represents the Lagrange multiplier domain penalty coefficient, represents the complex soft threshold function, the superscript represents the number of iterations, Expressed as Lagrange multipliers in matrix form, represents the target image in matrix form, Represents the auxiliary variables of the ADMM algorithm in matrix form.
[0114] It should be noted that in each iteration, the unknown phase error needs to be estimated adaptively. Using the minimum image entropy as the constraint, it can be expressed as
[0115] (14)
[0116] In formula (14), represents the sub-image energy, which is independent of the phase. It means adding all elements. represents the Hadamard product. Take the partial derivative of the above equation and set it equal to 0. Finally, the iterative process of phase error can be obtained as
[0117] (15)
[0118] In formula (15), It represents the operation of obtaining the phase. By iterating formula (13) and formula (15), a basically focused moving target can be obtained.
[0119] In this embodiment, after completing the compensation processing based on the constraints in the distance compression domain and constructing the self-focusing constraints based on the minimum image entropy, the focusing performance of the moving target is significantly improved in both distance and azimuth dimensions. However, it is found under real measured data that: due to the formation of the local rotating aperture, moving targets with special motion patterns, such as turning and U-turning, still suffer from severe geometric distortion. The size of the moving target obtained from these imaging results will be far from the ground truth. It should be emphasized that this phenomenon is rarely reported in conventional microwave bands. This is because in the microwave band, the moving target is usually considered to be a point target, while in the terahertz band, MT appears as an extended target, and its outline, etc. are revealed.
[0120] Therefore, here we establish a priori sparse table of azimuth expansion to perform recalibration. , moving target speed , rotate the initial angle It is related. By traversing all possible values offline in advance, it can be established in advance. The process can be expressed as:
[0121] (16)
[0122] In formula (16), and Can be clamped between -90 and 90°. The direction can be determined in advance by the shadow left by the ground target on the ground. It can be determined by the road geometry. Finally, the preliminary moving target focusing signal is self-focused and reconstructed based on the moving target geometric deformation constraint. The moving target geometric constraint, that is, the final moving target imaging result obtained under the moving target geometric constraint can be expressed as:
[0123] (17)
[0124] In formula (17), represents the moving target focused image, Represents the target geometry constraint.
[0125] After the above three-dimensional constraints, a well-focused moving target imaging result can be obtained. The above operation is performed on each moving target signal on each sub-aperture image to obtain the corresponding moving target focused image. The above operation is performed on each sub-aperture image sequentially or simultaneously to complete all the moving target focusing steps.
[0126] In step S130, the moving target focused image is stitched into the corresponding sub-aperture image again to obtain the sub-aperture image after the target is recalibrated.
[0127] Specifically, in step S120, when extracting the moving target signal from the sub-aperture image, after the sub-aperture image is subjected to static region imaging processing, a residual background image is also obtained. In this way, the moving target focused image and the corresponding residual background image are stitched together to obtain the sub-aperture image after the target is recalibrated.
[0128] Finally, in step S140, the same moving targets are superimposed in time sequence to obtain a complete target recalibrated sequence SAR image.
[0129] In one embodiment, the steps of the above-mentioned terahertz ground moving target recalibration imaging method are as follows: Figure 2 shown.
[0130] This method is also verified by experiments. Taking a terahertz radar system with a carrier frequency of 216 GHz as an example, the target is a tank and armored vehicle moving at an intersection. The corresponding experimental scene diagram is shown in the figure below. Figure 3 In the experiment, the terahertz radar system has a carrier frequency of 216 GHz, a bandwidth of 0.9 GHz, a pulse repetition frequency of 16,000 Hz, 8,192 sampling points in each pulse, and an observation time of 0.2 seconds for each frame. The experimental target is placed at a crossroads 3,200 meters away from the radar and moves back and forth along the road repeatedly.
[0131] First, by Figure 4 The results of large-scale scene imaging of outdoor terahertz ViSAR are shown. Before motion compensation, the overall scene is blurred, the road edge is unclear, and the target shadow is difficult to distinguish. After motion compensation, the shadow of the moving template is clearly visible and the road can be clearly distinguished. In addition, it can be found that there are two moving targets, whose imaging results are blurred and deviate from the center of the imaging scene.
[0132] Take moving target 1 as an example. Figure 5 The range compression domain imaging results before and after range motion compensation were compared. It was further verified that the moving target mainly behaves linearly under short dwell time. Subsequently, it can be found that after the motion compensation is completed, the tilted straight line has been significantly leveled, which is helpful for the subsequent azimuth refocusing.
[0133] Then, the two moving targets are focused in the azimuth dimension based on constraints 2 and 3. The corresponding imaging results are as follows: Figure 6 As shown in the figure, it can be found that the two moving targets are well reconstructed, the contours of the targets are restored, and the key points are revealed. It should be noted that at this time, the two moving targets are moving in a straight line along the road surface, so the azimuth scale of the moving targets has not changed significantly.
[0134] Then, taking the armored vehicle turning around as an example, Figure 7 (a) and Figure 7 (b) are the original imaging results before and after motion compensation of the armored vehicle. It can be noticed that the target size is greatly enlarged. After completing the compensation of constraints 1 and 2, the focusing performance of MT is improved, as shown in Figure 2. Figure 7 (c) shows that the azimuth scale imbalance still exists, which leads to a large deviation from the cognition. After estimating the speed and the rotation center angle of the dispatching target, it can be determined that the azimuth broadening coefficient is close to 4, so Figure 7 (d) shows the imaging result after the final orientation recalibration, which restores the outline of the armored vehicle and obtains accurate size information.
[0135] The above-mentioned terahertz ground moving target recalibration imaging method integrates the framework of ground moving target imaging and recalibration, which can adaptively refocus moving targets, especially curved moving targets, without being disturbed by motion models and background. This method takes ground moving targets as the object, first determines the characteristics of moving targets under terahertz ViSAR, and uses this as the basis. The range migration of moving targets is mainly manifested as range walk, so the range walk correction is established as constraint 1. Taking advantage of the sparsity of moving targets in the overall image, an ADMM framework based on minimum image entropy is established, and this is used as constraint 2. Considering that curved moving targets will undergo scale changes, a recalibration method with target geometry as constraint 3 is established, and the azimuth broadening coefficient is determined a priori. Finally, the refocused target image is spliced into the residual background. This method can robustly estimate the self-focusing capability of residual airborne motion from motion errors of non-cooperative moving targets.
[0136] Furthermore, this method can robustly handle various scenarios of moving target imaging, is not limited to the motion model of moving targets, can adaptively consider background clutter, and has low time dependence. This method directly starts from the perspective of the moving target sub-image, does not cause additional information loss, reduces the amount of data to be processed, and can consider multiple moving targets in parallel, thereby bringing stronger computational benefits. At the same time, this method has the characteristics of good stability and universality, and does not increase the complexity of the system and algorithm while obtaining better compensation performance and recalibration, and has good practicality.
[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. Moreover, 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 one embodiment, Figure 8 As shown, a terahertz ground moving target recalibration imaging device is provided, comprising: an original echo data acquisition module 200, a sub-aperture image acquisition module 210, a moving target refocusing module 220 under a three-constraint framework, a sub-aperture image reconstruction module 230 and a complete sequence SAR image acquisition module 240, wherein:
[0139] The original echo data acquisition module 200 is used to acquire the original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in the terahertz frequency band;
[0140] A sub-aperture image obtaining module 210 is used to perform aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0141] A moving target refocusing module 220 under a three-constraint framework is used to extract a moving target signal from the sub-aperture image, perform distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then perform self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraints to obtain a moving target focused image;
[0142] The sub-aperture image reconstruction module 230 is used to stitch the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated;
[0143] The complete sequence SAR image obtaining module 240 is used to obtain a complete target recalibrated sequence SAR image according to all sub-aperture images that have completed target recalibration.
[0144] For the specific definition of the terahertz ground moving target recalibration imaging device, please refer to the definition of the terahertz ground moving target recalibration imaging method mentioned above, which will not be repeated here. Each module in the above-mentioned terahertz ground moving target recalibration imaging device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0145] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a terahertz ground moving target recalibration imaging method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.
[0146] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0147] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:
[0148] Acquiring original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in a terahertz frequency band;
[0149] Performing aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0150] Extracting a moving target signal from the sub-aperture image, performing distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then performing self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraint to obtain a moving target focused image;
[0151] splicing the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated;
[0152] According to all the sub-aperture images that have completed the target recalibration, a complete target recalibration sequence SAR image is obtained.
[0153] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
[0154] Acquiring original echo data, wherein the original echo data is obtained by detecting the ground area using a video synthetic aperture radar in a terahertz frequency band;
[0155] Performing aperture division on the original echo data to obtain a plurality of sub-aperture images;
[0156] Extracting a moving target signal from the sub-aperture image, performing distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints to obtain a preliminary moving target focusing signal, and then performing self-focusing reconstruction on the preliminary moving target focusing signal based on the moving target geometric constraint to obtain a moving target focused image;
[0157] splicing the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated;
[0158] According to all the sub-aperture images that have completed the target recalibration, a complete target recalibration sequence SAR image is obtained.
[0159] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0160] The technical features of the above embodiments may 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.
[0161] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.
Claims
1. A terahertz ground moving target recalibration imaging method, characterized in that: The method comprises: Acquiring original echo data, wherein the original echo data is obtained by detecting a ground area containing a moving target using a video synthetic aperture radar in a terahertz frequency band; Performing aperture division on the original echo data to obtain a plurality of sub-aperture images; Extracting the moving target signal in the sub-aperture image, performing distance dimension compensation and azimuth dimension focusing on the moving target signal in sequence through two constraints, to obtain a preliminary moving target focused signal, wherein when performing distance dimension compensation, the distance movement alignment is used as a constraint condition in the distance compression domain, when performing azimuth dimension compensation, a self-focusing constraint condition is constructed based on the minimum image entropy, and then the preliminary moving target focused signal is self-focused and reconstructed based on the moving target geometric constraint to obtain a moving target focused image; splicing the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated; According to all the sub-aperture images that have completed the target recalibration, a complete target recalibration sequence SAR image is obtained.
2. The terahertz ground moving target recalibration imaging method according to claim 1, characterized in that: The moving targets in the sub-aperture image include one or more moving targets.
3. The terahertz ground moving target recalibration imaging method according to claim 1, characterized in that: When extracting moving target signals, clutter suppression is also performed on the sub-aperture image; Among them, clutter suppression is performed by using any one of the methods, or a combination of multiple methods, including space-time adaptive processing methods, Doppler filtering methods, adaptive clutter cancellation methods, image processing methods, constant false alarm rate detection methods or principal component analysis methods.
4. The terahertz ground moving target recalibration imaging method according to claim 3, characterized in that: When performing distance dimension compensation on the moving target signal, the constraint condition is expressed as: In the above formula, represents the estimated radial velocity, represents the radial velocity, represents the new wave number, Indicates that the distance movement is aligned. Indicates along the range beam The inverse Fourier transform of is the wavenumber domain signal of the moving target, For slow time, is the phase compensation function obtained based on the radial velocity of the moving target.
5. The terahertz ground moving target recalibration imaging method according to claim 1, characterized in that: After performing distance dimension compensation, when performing azimuth dimension focusing on the moving target signal, a self-focusing constraint condition is constructed based on the minimum image entropy, and the ADMM algorithm is used to solve it.
6. The terahertz ground moving target recalibration imaging method according to claim 5, characterized in that: The self-focusing constraint is expressed as: In the above formula, Indicates the initial moving target focusing signal, is the distance profile matrix Stacked in rows, is the subaperture image matrix Stacked in rows, represents the phase error matrix, represents the Fourier transform matrix, represents the regularization coefficient.
7. The terahertz ground moving target recalibration imaging method according to claim 6, characterized in that: When the ADMM algorithm is used to iteratively solve the self-focusing constraint condition, it is decomposed into three sub-problems, and the three sub-problems are solved alternately in each iteration. The three sub-problems are expressed as: In the above formula, represents the Lagrange multiplier domain penalty coefficient, represents the complex soft threshold function, the superscript represents the number of iterations, Expressed as Lagrange multipliers in matrix form, represents the target image in matrix form, Represents the auxiliary variables of the ADMM algorithm in matrix form.
8. The terahertz ground moving target recalibration imaging method according to claim 1, characterized in that: The preliminary moving target focusing signal is reconstructed by self-focusing based on the moving target geometric deformation constraint, and the moving target geometric constraint is expressed as: In the above formula, represents the moving target focused image, Represents the target geometry constraint.
9. The terahertz ground moving target recalibration imaging method according to any one of claims 1 to 8, characterized in that: When extracting the moving target signal in the sub-aperture image: After the sub-aperture image is subjected to static area imaging processing, a defocused region of interest related to the moving target is obtained based on a moving target model and imaging feature analysis, and moving target signals are extracted based on the defocused region of interest.
10. The terahertz ground moving target recalibration imaging method according to claim 9, characterized in that: When the moving target focused image and the corresponding sub-aperture image are stitched together: After performing static area imaging processing on the sub-aperture image, a residual background image is obtained; The moving target focused image is stitched with the corresponding residual background image to obtain a recalibrated sub-aperture image of the target.
11. A terahertz ground moving target recalibration imaging device, characterized in that: The device comprises: The original echo data acquisition module is used to acquire the original echo data, wherein the original echo data is obtained by detecting the ground area containing the moving target by the video synthetic aperture radar in the terahertz frequency band; A sub-aperture image obtaining module, used for performing aperture division on the original echo data to obtain a plurality of sub-aperture images; A moving target refocusing module under a three-constraint framework is used to extract a moving target signal from the sub-aperture image, and to perform distance dimension compensation and azimuth dimension focusing on the moving target signal in turn through two constraints, to obtain a preliminary moving target focusing signal, wherein when performing distance dimension compensation, the distance movement alignment is used as a constraint condition in the distance compression domain, and when performing azimuth dimension compensation, a self-focusing constraint condition is constructed based on the minimum image entropy, and then the preliminary moving target focusing signal is self-focused and reconstructed based on the moving target geometric constraint to obtain a moving target focused image; A sub-aperture image reconstruction module is used to stitch the moving target focused image with the corresponding sub-aperture image to obtain a sub-aperture image after the target is recalibrated; The complete sequence SAR image acquisition module is used to obtain a complete target recalibrated sequence SAR image based on all sub-aperture images that have completed target recalibration.
12. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 10 are implemented.
13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.