A terahertz video SAR fast imaging method and device

Abstract

The application provides a terahertz video SAR fast imaging method and device, the method comprises the following steps: obtaining terahertz video SAR echo data; performing sub-aperture division under a rectangular coordinate system with a full-aperture center as an origin; performing Dechirp processing on the echo data of the sub-aperture; performing two-dimensional interpolation on the data after Dechirp processing, realizing conversion from polar coordinates to rectangular coordinates, and obtaining wave number domain echo data of each sub-aperture; splicing adjacent two sub-aperture data in the wave number domain to obtain wave number domain data of a longer sub-aperture, and repeating the operation until all sub-apertures are spliced to obtain full-aperture wave number domain data; performing two-dimensional inverse Fourier transform on the full-aperture wave number domain data to obtain a full-resolution image; and performing phase error correction on the full-resolution image to obtain a terahertz video SAR image. The application can effectively improve operation efficiency while ensuring imaging quality.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and more specifically, to a terahertz video SAR fast imaging method and apparatus. Background Technology

[0002] Terahertz (THz) waves refer to electromagnetic waves with a spectrum between 100 GHz and 10 THz, characterized by high carrier frequency, large bandwidth, and good penetration. Compared with microwave synthetic aperture radar (SAR) imaging, terahertz video synthetic aperture radar (THz-ViSAR) imaging has significant advantages such as higher resolution, higher frame rate, higher detection probability, and easier identification, thus attracting increasing attention in the field of modern radar imaging.

[0003] Although various imaging algorithms exist, the two most commonly used imaging algorithms for THz-ViSAR are the Polar Format Algorithm (PFA) and the Back Projection Algorithm (BPA). PFA, based on the plane wavefront assumption, introduces residual phase errors, leading to geometric distortion and defocusing in SAR images, making it unsuitable for large scenes and high-resolution scenarios. BPA is a time-domain imaging algorithm applicable to any imaging mode and arbitrary trajectory. However, this algorithm requires point-by-point traversal, resulting in extremely high computational complexity, thus limiting its widespread application. To address this, several acceleration algorithms have been proposed, the most representative being the Fast Back Projection (FBP) and Fast Factorized Back Projection (FFBP). FBP first performs sub-aperture segmentation and reconstructs sub-images in local polar coordinates. Then, it transforms the sub-images to Cartesian coordinates and performs coherent superposition to obtain a full-resolution image. It sacrifices image quality for improved computational efficiency. FFBP uses the same processing method as FBP to obtain sub-images, and then obtains the final image through step-by-step image fusion. This process involves a large number of two-dimensional interpolations to map coordinates, inevitably introducing interpolation errors. This makes it difficult to balance image quality and computational efficiency in practical applications. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a terahertz video SAR fast imaging method and apparatus that can effectively improve computational efficiency while ensuring imaging quality.

[0005] To solve the above problems, the technical solution of the present invention is as follows:

[0006] A terahertz video SAR fast imaging method includes the following steps:

[0007] Acquire terahertz video SAR echo data;

[0008] Sub-aperture division is performed in a rectangular coordinate system with the center of the full aperture as the origin;

[0009] Decirp processing is performed on the echo data of the sub-aperture;

[0010] Two-dimensional interpolation is performed on the Decirp-processed data to achieve the conversion from polar coordinates to rectangular coordinates and obtain the wavenumber domain echo data of each sub-aperture.

[0011] The wavenumber domain data of each two adjacent sub-apertures are stitched together to obtain the wavenumber domain data of the longer sub-aperture. This process is repeated until all sub-apertures are stitched together to obtain the wavenumber domain data of the full aperture.

[0012] A two-dimensional inverse Fourier transform is performed on the wavenumber domain data of the full aperture to obtain a full-resolution image;

[0013] Phase error correction is performed on the full-resolution image to obtain a terahertz video SAR image.

[0014] Preferably, the step of acquiring terahertz video SAR echo data specifically includes: the terahertz video SAR transmitting a linear frequency modulated pulse signal, the expression of which is: Among them, t r For distance to fast time, T r For pulse width, f c The center frequency is γ = B / T. r B is the bandwidth, and rect(·) is the rectangular window function; for any point target P(x,y,z) within the illumination area, the instantaneous slant range from the radar platform to that point is: Among them, t a For azimuth, time slows down, x a Let v be the radar's x-axis coordinate. a H represents the radar-carrying aircraft's flight speed, and H represents the radar-carrying aircraft's flight altitude; the echo signal of a point target can be expressed as: Wherein, the two-way delay τ = 2R p / c, where c is the speed of light.

[0015] Preferably, the step of dividing the aperture into sub-apertures in a rectangular coordinate system with the center of the full aperture as the origin specifically includes: if the length of the full aperture is L a If the number of sub-apertures is N, then the length of the sub-aperture is l = L. a / N, at this time the echo data of the i-th sub-aperture can be expressed as: in, Let be the azimuth time corresponding to the i-th sub-aperture, and its range is:

[0016] Preferably, the step of performing Decirp processing on the echo data of the sub-aperture specifically includes the following steps:

[0017] Construct a reference signal with the same modulation frequency and center frequency as the transmitted signal, and mix the reference signal with the echo signal to obtain the echo difference frequency signal.

[0018] The echo difference frequency signal is transformed to the range frequency domain by using a range-to-Fourier transform;

[0019] Remove the remaining video phase term and the skewed phase term in the distance frequency domain;

[0020] The signal is then transformed to the range-time domain using an inverse Fourier transform in the range direction, resulting in the Decirp-processed echo signal of the i-th sub-aperture: Among them, wave number Differential Slope R ref Let be the reference slope distance for the i-th sub-aperture, and The center time of the sub-aperture.

[0021] Preferably, the step of performing two-dimensional interpolation on the Decirp-processed data to convert polar coordinates to rectangular coordinates and obtain wavenumber domain echo data for each sub-aperture specifically includes: interpolating the differential slant range... After performing a Taylor series expansion and ignoring terms of second order and higher, it can be expressed as: in, Let θ be the elevation angle and θ be the azimuth angle. At this point, the echo signal from the i-th sub-aperture is converted to: Among them, the range wave number K x and azimuth wave number K y They are respectively: Echo signal in polar coordinate format echo signal s in rectangular coordinate format i (K x ,K y The transformation can be achieved through two-dimensional interpolation.

[0022] Preferably, the step of stitching together every two adjacent sub-aperture data in the wavenumber domain to obtain wavenumber domain data for a longer sub-aperture, and repeating this process until all sub-apertures are stitched together to obtain wavenumber domain data for the full aperture, specifically includes: if the total number of sub-aperture stitching stages in the algorithm is G, for stage g (g = 1, 2…G), the number of sub-apertures is represented by K. g =2 G-g express, The wavenumber domain echo data of the qth sub-aperture in stage g (g = 1, 2…G) can be used to represent the wavenumber domain stitching process of two adjacent sub-apertures as follows: in The wavenumber domain echo data for the 2q-1th sub-aperture of stage g-1 has magnitudes M in the range and azimuth directions, respectively. r ×M a , The wavenumber domain echo data for the 2qth sub-aperture of stage g-1 has magnitudes M in the range and azimuth directions, respectively. r ×M a , The magnitudes of the distance and azimuth directions are M respectively. r ×2M a Repeat this process until all sub-apertures are stitched together to obtain the full aperture wavenumber domain data s(K). x ,K y ).

[0023] Preferably, the step of performing a two-dimensional inverse Fourier transform on the wavenumber domain data of the entire aperture to obtain a full-resolution image specifically includes: performing a two-dimensional inverse Fourier transform on the wavenumber domain data of the entire aperture to obtain a full-resolution image: I(x,y)=∫∫s(K x ,K y )exp[-j(xK x +yK y )]dxdy.

[0024] Preferably, the step of performing phase error correction on the full-resolution image to obtain a terahertz video SAR image specifically includes:

[0025] Through formula Determine whether secondary phase error correction is needed for the imaging scene. If the imaging scene is larger than r... π / 4 If the phase error is corrected by spatial post-filtering, then a secondary phase error correction is performed; otherwise, no correction is performed. In the formula, ρ... a Here, λ represents the azimuth resolution, and R represents the wavelength. c The minimum slant distance for the radar to reach the center of the scene.

[0026] Geometric distortion correction is achieved through image domain resampling, where the mapping relationship between the true ground coordinates (x, y) and the actual coordinates (x', y') in the image after imaging is as follows: in

[0027] Furthermore, the present invention also provides a terahertz video SAR rapid imaging apparatus, characterized in that the apparatus includes a processor and a memory for storing executable instructions of the processor, the processor being configured to execute the terahertz video SAR rapid imaging method as described above by executing the executable instructions.

[0028] Compared with existing technologies, the advantages of this invention are as follows: In the initial imaging stage, a global Cartesian coordinate system with a simpler geometric configuration and a more efficient polar coordinate algorithm are used to process sub-aperture data, greatly reducing the computational load and making it easier to implement; the fusion stage is achieved through simple wavenumber domain stitching, avoiding the introduction and accumulation of coordinate mapping and interpolation errors. Therefore, the terahertz video SAR fast imaging method described in this invention effectively improves computational efficiency while ensuring imaging quality. Attached Figure Description

[0029] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0030] Figure 1 This is a flowchart of a terahertz video SAR fast imaging method provided in an embodiment of the present invention;

[0031] Figure 2 This is a geometric schematic diagram of terahertz video SAR provided in an embodiment of the present invention. Detailed Implementation

[0032] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0033] Specifically, Figure 1 This is a flowchart of a terahertz video SAR fast imaging method provided in an embodiment of the present invention; Figure 2 This is a geometrical schematic diagram of a terahertz video SAR provided in an embodiment of the present invention, as shown below. Figure 1 and Figure 2 As shown, the terahertz video SAR fast imaging method includes the following steps:

[0034] S1: Acquire terahertz video SAR echo data;

[0035] Specifically, in step S1, the terahertz video SAR transmits a linear frequency modulated pulse signal, the expression of which is:

[0036]

[0037] Among them, t r For distance to fast time, T r For pulse width, f c The center frequency is γ = B / T. r B is the bandwidth, and rect(·) is the rectangular window function.

[0038] For any point target P(x,y,z) within the illumination area, the instantaneous slant range from the radar platform to that point is:

[0039]

[0040] Among them, t a For azimuth, time slows down, x a Let v be the radar's x-axis coordinate. a H represents the speed of the radar-carrying aircraft, and H represents the altitude of the radar-carrying aircraft.

[0041] The echo signal of a point target can be represented as:

[0042]

[0043] Wherein, the two-way delay τ = 2R p / c, where c is the speed of light.

[0044] S2: Sub-aperture division is performed in a rectangular coordinate system with the center of the full aperture as the origin;

[0045] Specifically, the sub-apertures are divided in a rectangular coordinate system with the center of the full aperture as the origin. If the length of the full aperture is L... a If the number of sub-apertures is N, then the length of the sub-aperture is l = L. a / N. The echo data of the i-th sub-aperture can then be expressed as:

[0046]

[0047] in, The azimuth time corresponding to the i-th sub-aperture, its range is...

[0048]

[0049] S3: Perform Decirp processing on the echo data of the sub-aperture;

[0050] Specifically, the Decirp processing of the echo data from the sub-aperture includes the following steps:

[0051] Step 1: Construct a reference signal with the same modulation frequency and center frequency as the transmitted signal. Mix the reference signal with the echo signal to obtain the echo difference frequency signal.

[0052] Step 2: Transform the echo difference frequency signal to the range frequency domain using a range-to-Fourier transform;

[0053] Step 3: Remove the remaining video phase term and the skewed phase term in the distance frequency domain;

[0054] Step 4: Then, transform the above signal to the range time domain using an inverse Fourier transform in the range direction to obtain the Decirp-processed i-th sub-aperture echo signal:

[0055]

[0056] Among them, wave number Differential Slope R ref Let be the reference slope distance for the i-th sub-aperture, and The center time of the sub-aperture.

[0057] S4: Perform two-dimensional interpolation on the Decirp-processed data to convert polar coordinates to rectangular coordinates and obtain wavenumber domain echo data for each sub-aperture.

[0058] Specifically, two-dimensional interpolation is performed on the data processed by Decirp to achieve the conversion from polar coordinates to rectangular coordinates.

[0059] Differential Slope Range After performing a Taylor series expansion and ignoring terms of second order and higher, it can be expressed as:

[0060]

[0061] in, θ is the pitch angle, and θ is the azimuth angle.

[0062] Substituting equation (7) into equation (6), we get:

[0063]

[0064] Among them, the range wave number K x and azimuth wave number K y They are respectively

[0065]

[0066] At this point, the echo signal of the i-th sub-aperture is transformed from equation (6) to equation (9), and the conversion from polar coordinate format data to rectangular coordinate format can be achieved through two-dimensional interpolation.

[0067] Steps S3 and S4 are performed on all sub-aperture data respectively to obtain wavenumber domain echo data for each sub-aperture.

[0068] S5: Stitch together the wavenumber domain data of every two adjacent sub-apertures to obtain the wavenumber domain data of the longer sub-aperture. Repeat this process until all sub-apertures are stitched together to obtain the wavenumber domain data of the full aperture.

[0069] Specifically, the wavenumber domain data of every two adjacent sub-apertures are stitched together to obtain the wavenumber domain data of the longer sub-aperture; step S5 is repeated until all sub-apertures are stitched together to obtain the wavenumber domain data of the full aperture.

[0070] Specifically, if the total number of stages for sub-aperture stitching in the algorithm is G, then for stage g (g = 1, 2…G), the number of sub-apertures is represented by K. g =2 G-g This means, for example, that the number of sub-apertures in stage 1 is K1 = 2. G-1 . The wavenumber domain echo data of the qth sub-aperture in stage g (g = 1, 2…G) can be used to represent the wavenumber domain stitching process of two adjacent sub-apertures as follows:

[0071]

[0072] in The wavenumber domain echo data for the 2q-1th sub-aperture of stage g-1 has magnitudes M in the range and azimuth directions, respectively. r ×M a , The wavenumber domain echo data for the 2qth sub-aperture of stage g-1 has magnitudes M in the range and azimuth directions, respectively. r ×M a , The magnitudes of the distance and azimuth directions are M respectively. r ×2M a Repeat this process until all sub-apertures are stitched together to obtain the full aperture wavenumber domain data s(K). x ,K y ).

[0073] S6: Perform a two-dimensional inverse Fourier transform on the wavenumber domain data of the full aperture to obtain a full-resolution image;

[0074] Specifically, in step S6, a two-dimensional inverse Fourier transform is performed on the wavenumber domain data of the full aperture to obtain a full-resolution image, i.e.:

[0075] I(x,y)=∫∫s(K x ,K y )exp[-j(xK x +yK y )]dxdy (11)

[0076] S7: Perform phase error correction on the full-resolution image to obtain a terahertz video SAR image.

[0077] Specifically, in step S7, phase error correction is performed on the full-resolution image to obtain a defocus-free terahertz video SAR image.

[0078] In step S4, the differential slope distance is... When performing a Taylor series expansion, neglecting second-order and higher-order terms introduces linear and quadratic phase errors into the echo signals of each sub-aperture. Linear phase error causes image distortion, while quadratic phase error is spatially variable and becomes more severe with distance from the scene center.

[0079] The phase error correction specifically includes the following steps:

[0080] Step 1: Determine whether the imaging scene needs secondary phase error correction using formula (12). If the imaging scene is greater than r... π / 4 If the error is positive, then a secondary phase error correction is performed through spatial post-filtering; otherwise, no correction is performed.

[0081]

[0082] In the formula, ρ a Here, λ represents the azimuth resolution, and R represents the wavelength. c The minimum slant distance for the radar to reach the center of the scene.

[0083] Step 2: Geometric distortion correction is achieved through image domain resampling, where the mapping relationship between the true ground coordinates (x, y) and the actual coordinates (x', y') in the image after imaging is as follows:

[0084]

[0085] in

[0086] Compared with existing technologies, the advantages of this invention are as follows: In the initial imaging stage, a global Cartesian coordinate system with a simpler geometric configuration and a more efficient polar coordinate algorithm are used to process sub-aperture data, greatly reducing the computational load and making it easier to implement; the fusion stage is achieved through simple wavenumber domain stitching, avoiding the introduction and accumulation of coordinate mapping and interpolation errors. Therefore, the terahertz video SAR multi-level back projection fast imaging algorithm described in this invention effectively improves computational efficiency while ensuring imaging quality.

[0087] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A terahertz video SAR fast imaging method, characterized in that, The method includes the following steps: Acquire terahertz video SAR echo data; Sub-aperture division is performed in a rectangular coordinate system with the center of the full aperture as the origin; Decirp processing is performed on the echo data of the sub-aperture; Two-dimensional interpolation is performed on the Decirp-processed data to achieve the conversion from polar coordinates to rectangular coordinates and obtain the wavenumber domain echo data of each sub-aperture. The wavenumber domain data of each two adjacent sub-apertures are stitched together to obtain the wavenumber domain data of the longer sub-aperture. This process is repeated until all sub-apertures are stitched together to obtain the wavenumber domain data of the full aperture. A two-dimensional inverse Fourier transform is performed on the wavenumber domain data of the full aperture to obtain a full-resolution image; Phase error correction is performed on the full-resolution image to obtain a terahertz video SAR image.

2. The terahertz video SAR fast imaging method according to claim 1, characterized in that, The specific steps for acquiring terahertz video SAR echo data include: the terahertz video SAR transmits a linear frequency modulated pulse signal, the expression of which is: Among them, t r For distance to fast time, T r For pulse width, f c The center frequency is γ = B / T. r B is the bandwidth, and rect(·) is the rectangular window function; for any point target P(x,y,z) within the illumination area, the instantaneous slant range from the radar platform to that point is: Among them, t a For azimuth, time slows down, x a Let v be the radar's x-axis coordinate. a H represents the radar-carrying aircraft's flight speed, and H represents the radar-carrying aircraft's flight altitude; the echo signal of a point target can be expressed as: Wherein, the two-way delay τ = 2R p / c, where c is the speed of light.

3. The terahertz video SAR fast imaging method according to claim 2, characterized in that, The step of dividing the sub-aperture in a rectangular coordinate system with the center of the full aperture as the origin specifically includes: if the length of the full aperture is L a If the number of sub-apertures is N, then the length of the sub-aperture is l = L. a / N, at this time the echo data of the i-th sub-aperture can be expressed as: in, Let be the azimuth time corresponding to the i-th sub-aperture, and its range is:

4. The terahertz video SAR fast imaging method according to claim 1, characterized in that, The Decirp processing of the echo data from the sub-aperture specifically includes the following steps: Construct a reference signal with the same modulation frequency and center frequency as the transmitted signal, and mix the reference signal with the echo signal to obtain the echo difference frequency signal. The echo difference frequency signal is transformed to the range frequency domain by using a range-to-Fourier transform; Remove the remaining video phase term and the skewed phase term in the distance frequency domain; The signal is then transformed to the range-time domain using an inverse Fourier transform in the range direction, resulting in the Decirp-processed echo signal of the i-th sub-aperture: in, The azimuth time corresponding to the i-th sub-aperture; wavenumber c is the speed of light, f c Center frequency; differential slant distance for The instantaneous slant distance at time t, Let be the reference slope distance for the i-th sub-aperture, and x is the center time of the sub-aperture. a Let v be the radar's x-axis coordinate. a H represents the speed of the radar-carrying aircraft, and H represents the altitude of the radar-carrying aircraft.

5. The terahertz video SAR fast imaging method according to claim 4, characterized in that, The specific steps of performing two-dimensional interpolation on the Decirp-processed data to convert polar coordinates to rectangular coordinates and obtain wavenumber domain echo data for each sub-aperture include: interpolating the differential slant range... After performing a Taylor series expansion and ignoring terms of second order and higher, it can be expressed as: Where x and y represent the x-axis coordinates and y-axis coordinates of the point target, respectively. Let θ be the elevation angle and θ be the azimuth angle. The echo signal from the i-th sub-aperture is then converted to: Among them, the range wave number K x and azimuth wave number K y They are respectively: Echo signal in polar coordinate format echo signal s in rectangular coordinate format i (K x ,K y The transformation can be achieved through two-dimensional interpolation.

6. The terahertz video SAR fast imaging method according to claim 5, characterized in that, The step of stitching together the wavenumber domain data of every two adjacent sub-apertures to obtain wavenumber domain data for longer sub-apertures, and repeating this process until all sub-apertures are stitched together to obtain the wavenumber domain data for the full aperture, specifically includes: if the total number of sub-aperture stitching stages in the algorithm is G, for stage g (g = 1, 2…G), the number of sub-apertures is represented by K. g =2 G-g express, The wavenumber domain echo data of the qth sub-aperture in stage g (g = 1, 2…G) can be used to represent the wavenumber domain stitching process of two adjacent sub-apertures as follows: in The wavenumber domain echo data for the 2q-1th sub-aperture of stage g-1 has magnitudes M in the range and azimuth directions, respectively. r ×M a , The wavenumber domain echo data for the 2qth sub-aperture of stage g-1 has magnitudes M in the range and azimuth directions, respectively. r ×M a , The magnitudes of the distance and azimuth directions are M respectively. r ×2M a Repeat this process until all sub-apertures are stitched together to obtain the full aperture wavenumber domain data s(K). x ,K y ).

7. The terahertz video SAR fast imaging method according to claim 1, characterized in that, The step of performing a two-dimensional inverse Fourier transform on the wavenumber domain data of the entire aperture to obtain a full-resolution image specifically includes: performing a two-dimensional inverse Fourier transform on the wavenumber domain data of the entire aperture to obtain a full-resolution image: I(x,y)=∫∫s(K x ,K y )exp[-j(xK x +yK y )]dxdy, where x and y represent the x-axis coordinates and y-axis coordinates of the point target, respectively, K x K is the range wave number. y For the azimuth wave number, s(K x ,K y () represents the wavenumber domain data for the entire aperture.

8. The terahertz video SAR fast imaging method according to claim 1, characterized in that, The step of performing phase error correction on the full-resolution image to obtain a terahertz video SAR image specifically includes: Through formula Determine whether secondary phase error correction is needed for the imaging scene. If the imaging scene is larger than r... π / 4 If the phase error is corrected by spatial post-filtering, then a secondary phase error correction is performed; otherwise, no correction is performed. In the formula, ρ... a Here, λ represents the azimuth resolution, and R represents the wavelength. c The minimum slant distance for the radar to reach the center of the scene. Where x a H represents the radar's x-axis coordinate, and H represents the radar carrier aircraft's flight altitude. Geometric distortion correction is achieved through image domain resampling, where the true ground coordinates (x, y) and the actual coordinates (x, y) in the image after imaging are compared. ' ,y ' The mapping relationship between them is as follows: in It is the pitch angle.

9. A terahertz video SAR fast imaging device, characterized in that, The apparatus includes a processor and a memory for storing executable instructions of the processor, the processor being configured to perform the terahertz video SAR fast imaging method as described in any one of claims 1 to 8 by executing the executable instructions.

Images