A multi-level resolution fast imaging method for multi-channel high-resolution wide-swath SAR
Through the multi-stage resolution fast imaging method, low-resolution imaging is achieved using de-aberving processing and sub-aperture imaging, and high-resolution imaging is performed by combining constant virtual alarm detection and large bandwidth echo data, which solves the problem of high-component wide-frame synthetic aperture radar in large wide scenes The problem of slow imaging speed and low quality is achieved efficient and high-quality imaging.
Patent Information
- Application Number
- CN202210826682.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-07-13
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2042-07-13
AI Technical Summary
The existing high-component wide-frame synthetic aperture radar imaging technology has the problem of slow imaging speed and low imaging quality in large and wide scenes.
Using multi-stage resolution fast imaging method, low-resolution imaging is achieved through de-aberving and sub-aperture imaging, the target area is extracted using the constant alarm detection method, and then high-resolution imaging of the target area is performed, and the final combination of large bandwidth full-aperture echo data is performed. Imaging.
It improves imaging efficiency and quality, solves the problems of slow imaging speed and low quality, and is suitable for fast and high-resolution imaging in large scenarios.
Smart Images

Figure CN115436939B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of radar technology, and particularly relates to the field of high-resolution and wide-band synthetic aperture radar (SAR) imaging technology. Background Art
[0002] Synthetic aperture radar has the characteristics of all-day and all-weather, and is widely used in the fields of ocean monitoring and disaster warning. High-resolution and wide-width (HRWS) imaging is an important research direction of synthetic aperture radar (SAR) imaging. High resolution can provide more details in the observation area, while wide-width mapping can obtain information in a wider range, and can provide systematic observation of the dynamic process of the earth's surface, especially suitable for monitoring vast ocean areas. For wide-width mapping, such as a wide sea surface, if full-scene and high-precision imaging is to be performed, the amount of data to be processed is very large, resulting in a long imaging time, which cannot meet the actual application requirements of rapid detection of ships in the sea area, maritime rescue, and sea surface oil spill monitoring. Therefore, it is of great significance to study efficient imaging methods suitable for high-resolution wide-width SAR. For details, see the literature "Deng Yunkai, Yu Weidong, Wang Yu. High-resolution wide-width spaceborne SAR ocean monitoring and information inversion [J]. Science and Technology Review, 2017, 35(20): 69-76.".
[0003] For wide-swath imaging, multi-channel technology is usually used to solve the contradiction between traditional single-channel SAR in obtaining high resolution and wide-swath imaging at the same time. For details, see the literature "Guo Zhenyong, Yuan Xinzhe, Zhang Ping. A multi-channel SAR high-resolution wide-swath imaging algorithm [J]. Journal of Electronics and Information Technology, 2008(02):310-313.". At present, the research on high-resolution wide-swath multi-channel SAR fast imaging technology is still in its early stages. Most of the related fast imaging methods are based on converting multi-channel SAR echo data into multiple single-channel SAR data through compensation equivalent processing. Existing fast imaging methods can be divided into the following three categories: Fast SAR imaging methods based on frequency domain processing, mainly range-Doppler algorithm (RDA) and range migration algorithm (RMA). In the case of wide-width imaging, the slant distance between each target in the scene and the radar platform is much greater than that of a conventional SAR system, resulting in more complex target echo properties. A higher-order model is needed to characterize the echo model, which increases the complexity of system implementation. At the same time, a large number of approximations are used, which reduces the accuracy of imaging. The fast SAR imaging method based on time domain processing mainly includes the back projection (BP) algorithm. Since there is no approximation, it can well eliminate the distance-azimuth coupling. However, due to its high computational complexity, it cannot be applied in large-scale imaging scenarios of large data. The fast time domain methods that accelerate the BP algorithm mainly include the fast back projection (FBP) algorithm and the fast decomposed back projection (FFBP) algorithm. This type of algorithm decomposes the synthetic aperture into several sub-apertures, projects the distance compression data corresponding to the sub-aperture to the local polar coordinate network centered on the aperture to obtain a sub-image, and obtains the final image by fusing the sub-images. The image synthesis will cause errors to a certain extent, affecting the imaging quality.
[0004] Therefore, in order to solve the problems of slow imaging speed and low imaging quality in wide-width imaging of existing methods, the present invention proposes a multi-level resolution fast imaging method for multi-channel high-resolution wide-width SAR. This method first obtains the imaging result of the entire scene quickly at a low resolution, then quickly detects the target area and accurately extracts it, and then performs high-resolution imaging of the target area, thereby improving the imaging efficiency while maintaining high imaging quality. Summary of the invention
[0005] The invention belongs to the technical field of high-resolution and wide-band synthetic aperture radar (SAR) imaging, and discloses a multi-level resolution fast imaging method for multi-channel high-resolution and wide-band SAR. The method is based on the characteristics that the target area in the wide-band large scene area is relatively small and sparsely distributed, and the target has a high signal-to-noise ratio relative to the background, and is used to solve the problems of slow imaging speed and low imaging quality in the existing method in wide-band imaging. The method mainly includes three parts: fast low-resolution imaging of the entire scene, target area detection and extraction of low-resolution imaging results, and high-resolution imaging of the target area. For a wide-band target sparse scene, the present invention uses de-skewing processing in the distance upward and sub-aperture imaging in the azimuth direction to quickly realize low-resolution imaging, and then uses a constant false alarm detection method to detect and extract targets on the low-resolution image, and then uses large-bandwidth full-aperture echo data to image the target area, and finally obtains a high-resolution target image. Compared with the existing fast imaging method, the proposed method solves the contradiction between imaging quality and speed in wide-band imaging, and does not cause a decrease in imaging quality while obtaining fast imaging results.
[0006] In order to facilitate the description of the present invention, the following terms are first defined:
[0007] Definition 1. Pulse Repetition Frequency (PRF)
[0008] PRF (pulse repetition frequency) is the pulse repetition frequency, or repetition frequency for short. It is the number of pulses emitted per second and is the reciprocal of the pulse repetition interval (PRI). The pulse repetition interval is the time interval between one pulse and the next. The periodic pulse repetition interval is the pulse repetition period. For details on pulse repetition frequency, please refer to the literature "Radar Imaging Technology", edited by Bao Zheng et al., published by Electronic Industry Press.
[0009] Definition 2: Multi-channel High Resolution Wide Swath (HRWS) SAR
[0010] In order to obtain high resolution in azimuth, the traditional single-channel synthetic aperture radar system must meet the requirements of high pulse repetition frequency (PRF) while meeting the minimum azimuth antenna area. However, setting a limited PRF is a necessary condition to overcome distance ambiguity in order to obtain a wide swath. That is, in a single-channel synthetic aperture radar system, high resolution and wide swath are a pair of contradictory performance indicators. The application of multi-channel synthetic aperture radar system can solve this problem well. It combines digital beamforming technology (DBF) to efficiently achieve high-resolution wide swath ground observation, and has significant advantages in ground moving target detection, high-resolution wide-swath imaging and interference suppression. For details of multi-channel high-resolution wide-swath SAR, please refer to the literature: Guo Xiaojiang. Research on key technologies for high-resolution wide-swath SAR imaging processing [D]. Shanghai Jiaotong University, 2018.
[0011] Definition 3: Equivalent Phase Center (EPC) Principle
[0012] For a multi-channel SAR system, it can be regarded as a SAR that transmits and receives signals at the equivalent phase center (EPC) located between the transmitter and the receiver. The principle of the equivalent phase center is detailed in the literature G. Krieger, N. Gebert, and A. Moreira, "Unambiguous SAR signal reconstruction from nonuniform displaced phase center sampling," IEEE Geoscience and Remote Sensing Letters, vol. 1, no. 4, pp. 260–264, 2004.
[0013] Definition 4: Synthetic Aperture Radar Range Compression Method
[0014] The range compression method of synthetic aperture radar refers to the process of matching filtering using the transmission parameters of synthetic aperture radar, which mainly includes: using the range reference signal to match filter compression and range migration compensation on the range echo signal of synthetic aperture radar to obtain the imaging of synthetic aperture radar. For details of the traditional range compression method of synthetic aperture radar, please refer to the literature "Radar Imaging Technology", edited by Bao Zheng et al., published by Electronic Industry Press.
[0015] Definition 5. Norm
[0016] Let X be a number field On linear space, represents a complex number field. If it satisfies the following properties: ||X||≥0, and ||X||=0 only if X=0; ||aX||=|a|||X||, a is an arbitrary constant; ||X1+X2||≤||X1||+||X2||, then ||X|| is called the norm on the X space, where X1 and X2 are any two values on the X space. For the N×1 dimensional discrete signal vector X=[x1,x2,…,x N ] T , the LP norm expression of vector X is where x i is the i-th element of vector X, ∑|·| represents the absolute value summation symbol, and the L1 norm expression of vector X is The L2 norm of vector X is expressed as The L0 norm expression of vector X is And x i ≠0. For details on the norm, please refer to the literature “Matrix Theory”, edited by Huang Tingzhu et al., published by Higher Education Press.
[0017] Definition 6: Backprojection (BP) algorithm
[0018] The BP algorithm is a processing method borrowed from tomography technology. This method realizes the accumulation of energy at each scattering point by projecting the echo data point by point to each pixel in the image space. Compared with other methods, the BP method has a simple imaging principle, high motion compensation accuracy, no geometric distortion, and is suitable for SAR imaging of large scenes, large squints, arbitrary modes and arbitrary trajectories. The basic idea is to calculate the distance history between each pixel point in the imaging area and the SAR antenna platform within the aperture length, and then coherently accumulate the corresponding time domain echo signals to reconstruct the scattering coefficient of each pixel. Wherein, σ(m) represents the backscattering coefficient of the mth scattering point, m=1,2,…,M, M represents the total number of grid points of the image space division; s(r(n,m),n) represents the echo value obtained after interpolation of the mth scattering point at the nth slow time; r(n,m) represents the two-way slant range from the mth scattering point to the radar at the nth slow time; ω is the carrier frequency of the transmitted signal; N represents the total number of slow moments within a synthetic aperture.
[0019] In general, the BP imaging component of the mth pixel in an imaging scene using uniformly sampled echoes can be expressed as For details of the traditional back-projection algorithm, see S.Jun, Z.Xiaoling, Y.Jianyu, and W.Chen, “Apc trajectory design for one-active linear-array three-dimensional imaging sar,” IEEE Transactions on Geoscience and Remote Sensing, vol.48, pp.1470–1486, March 2010.
[0020] Definition 7: Subaperture Imaging Method
[0021] Sub-aperture imaging refers to the use of echo data from a partial aperture (i.e., sub-aperture) for imaging instead of the full aperture for the azimuth synthetic aperture of the SAR. Due to the small amount of data to be processed, the imaging speed will be greatly improved, but the azimuth resolution will be reduced. The azimuth resolution of the SAR system is Assume that the echo data of full aperture 1 / p is taken as the data used for imaging, that is, the length of the synthetic aperture used is That is, the horizontal aperture D' of a single antenna after equivalent x For D' x =pD x , then the corresponding azimuth resolution is For details of the traditional sub-aperture imaging method, please refer to the document "Radar Imaging Technology", edited by Bao Zheng et al., published by the Publishing House of Electronics Industry.
[0022] Definition 8: De-skewing method
[0023] The de-skewing method is proposed for linear frequency modulation signals. The method refers to mixing a reference signal with a fixed time and the same frequency and modulation frequency with the echo signal. The de-skewing echo will form a narrowband signal different from the original echo, and the echo data with low resolution of the range image can be obtained. For details of the traditional de-skewing method, please refer to the literature "Radar Imaging Technology", edited by Bao Zheng et al., published by Electronic Industry Press.
[0024] Definition 9: Constant false alarm detection method
[0025] CFAR detection technology refers to the technology that the radar system uses to distinguish the signal and noise output by the receiver to determine whether the target signal exists while keeping the false alarm probability constant. For details on the traditional CFAR detection method, please refer to the document "Principles of Synthetic Aperture Radar Imaging", edited by Pi Yiming et al., published by the University of Electronic Science and Technology of China Press.
[0026] Definition 10: SAR traditional theoretical imaging resolution
[0027] The traditional theoretical imaging resolution of SAR refers to the imaging resolution of the SAR system in the range, azimuth and track directions obtained by using the classical matched filter theory imaging algorithm. For a common antenna for transmission and reception, the resolution of SAR in the range direction is denoted by ρ r , the approximate expression is Where C is the speed of light in air, B r is the bandwidth of the SAR transmission signal; the resolution in azimuth is denoted by ρ a , the approximate expression is Where D a is the actual aperture of the antenna in azimuth; the resolution in the tangent track direction is denoted by ρ c , the approximate expression is Where λ is the wavelength of the SAR radar carrier frequency, R0 is the reference slant distance from the SAR platform to the center of the observation scene, and L is the antenna length. For details on the traditional theoretical imaging resolution of SAR, please refer to the reference "Shi, J., et al., APCtrajectory design for One-Active linear-array three-dimensional imaging SAR, IEEE Transactions on Geoscience and Remote Sensing, Vol. 48, No. 3, pp: 1470-1486, 2010".
[0028] Definition 11. Synthetic aperture radar raw echo simulation method
[0029] The synthetic aperture radar raw echo simulation method refers to a method of simulating the original signal with the characteristics of synthetic aperture radar echo signal under certain system parameter conditions based on the synthetic aperture radar imaging principle. For details of the traditional synthetic aperture radar raw echo simulation method, please refer to the literature: "Research on InSAR Echo Signal and System Simulation", Zhang Jianqi, master's thesis of Harbin Institute of Technology.
[0030] The present invention provides a multi-level resolution fast imaging method for multi-channel high-resolution wide-band SAR, which comprises the following steps:
[0031] Step 1: Initialize the multi-channel SAR system parameters required for high-resolution wide-band SAR multi-level resolution imaging:
[0032] Initialize the multi-channel SAR system parameters required for high-resolution wide-band SAR multi-level resolution imaging, including: the observation space of the multi-channel SAR is the ground three-dimensional coordinate system, denoted by XYZ, where X represents the horizontal axis, Y represents the horizontal axis, and Z represents the horizontal vertical axis; the bandwidth of the radar transmit signal is denoted by B, the pulse width of the radar transmit signal is denoted by T, and the frequency modulation slope of the radar transmit signal is denoted by K r, the sampling frequency of the radar receiving system is recorded as f s ; The multi-channel SAR imaging system moves in a uniform linear motion along the Y-axis parallel direction, and the platform velocity vector is recorded as The platform flight altitude is recorded as h; the number of channels of the multi-channel SAR is recorded as N, and the first channel transmitting the signal is recorded as T x , which is also the first channel R that receives the echo x1 ; The remaining channels only receive echoes, which are denoted by R x2 ,R x3 ,...,R xN ; The radar center frequency is denoted as f c , the pulse repetition frequency of the radar system's transmitted signal is recorded as PRF, and the pulse repetition time of the transmitted signal is recorded as PRI; the distance between each channel is recorded as d; the sampling time of the nth channel that satisfies uniform time sampling in the kth pulse repetition period is recorded as Where k represents the pulse repetition period, and its value is k=1,2,...,K; n represents the channel, and its value is 1,2,...,N; the echo is denoted by S r (τ(t,m),t), the echo after range compression is s r (τ(t,m),t), where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the mth pixel at the azimuth slow time t, and m is the mth pixel in the image domain. The uniform time sampling of the echo can be expressed as s r (τ(t k,n ,m),t k,n ), where t k,n is the azimuth slow time, τ(t k,n ,m) is the slow time t in azimuth k,n The distance to the fast time is the round-trip delay of the echo from the EPC to the mth pixel;
[0033] Step 2: Initialize the observation scene target space parameters of high-resolution wide-band SAR multi-level resolution imaging:
[0034] Initialize the scene space parameters of high-resolution wide-band SAR multi-level resolution imaging, including: the slant distance of the scene center is recorded as r o , the maximum observation distance of the radar system is recorded as r max The range width of the target area observed by the radar system is recorded as △r, and the number of pixels in the range is recorded as N r The number of pixels in the azimuth direction is recorded as N a , the total number of grids is recorded as M, the distance grid resolution is recorded as dr, and the azimuth grid resolution is recorded as da;
[0035] Step 3: Initialize the constant false alarm detection parameters:
[0036] The initial false alarm probability is denoted as P fa , the number of protection unit columns is recorded as N gc , the number of protection unit rows is recorded as N gr , the number of background clutter columns is recorded as N tc , the number of background clutter rows is recorded as N tr ;
[0037] Step 4: Generate received echo data and reference signal:
[0038] Using formula The reference signal is calculated and recorded as S ref (t), where t is the azimuth slow time, τ0 = 2r0 / c, r0 is the slant distance from the center of the scene, c is the speed of light, and T' = 2r max / c,r max is the maximum observation distance of the radar system, T is the pulse width of the radar transmission signal, and f c is the radar center frequency, K r is the frequency modulation slope of the radar transmission signal, rect[·] is the matrix window function;
[0039] The original echo data of multi-channel SAR is generated by using the original echo simulation method of synthetic aperture radar in Definition 11, denoted as S r (τ(t,m),t), i.e., the echo in step 1, where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain;
[0040] Step 5: Perform range de-skewing to obtain a small bandwidth echo:
[0041] For the original echo signal S obtained in step 4 r (τ(t,m),t) and the reference signal S ref (t), using the formula The small bandwidth echo after de-slanting is calculated and recorded as S o (t), where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain. is the reference signal S ref (t) conjugation;
[0042] Using formula The narrow bandwidth of the echo after de-slanting is calculated and recorded as B', where K r is the radar system transmission modulation slope, △r is the range width of the radar system observation target area, and c is the speed of light;
[0043] Step 6: Azimuth sub-aperture BP imaging
[0044] For the narrowband bandwidth B' obtained in step 5, use the formula The resolution of the low-resolution data after range deslanting is calculated and is denoted as ρ', where c is the speed of light;
[0045] Using formula The resolution of the high-resolution data before deslanting in the range direction is calculated and recorded as ρ, where c is the speed of light and B is the bandwidth of the radar transmission signal;
[0046] The formula p = ρ' / ρ is used to calculate the sampling multiple of the sub-aperture imaging resolution relative to the full-aperture imaging resolution, which is denoted as p.
[0047] Using formula The number of pixels in the range direction of low-resolution BP imaging is calculated and recorded as N r ';
[0048] Using formula Calculate the number of azimuth pixels of low-resolution BP imaging, denoted as N a ';
[0049] Using formula The total number of grids for low-resolution BP imaging is calculated and recorded as M';
[0050] The range grid resolution of low-resolution BP imaging is calculated using the formula dr' = p·dr, denoted as dr';
[0051] The azimuth grid resolution of low-resolution BP imaging is calculated using the formula da' = p·da, denoted as da', where N r is the number of distance pixels defined in step 2, N a is the number of pixels in the azimuth direction defined in step 2, M is the total number of grids defined in step 2, dr is the grid resolution in the distance direction defined in step 2, and da is the grid resolution in the azimuth direction defined in step 2;
[0052] For the de-slanted echo S obtained in step 5 o (t), using the formula The sub-aperture echo is calculated and recorded as S o (t'), where t' is the azimuth slow time, t is the azimuth slow time, and K is the total number of sampling periods of the azimuth echo;
[0053] The sub-aperture echo data S is compressed using the range-direction pulse compression method of the standard synthetic aperture radar echo data in Definition 4. o (t') performs range pulse compression to obtain multi-channel SAR echo data after range compression, denoted as s o(τ(t',m),t'), where t' is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t', and m is the m-th pixel in the image domain;
[0054] Using formula The sub-image of the mth scattering point in channel n is calculated and recorded as I n (m), where f c is the center frequency of the transmitted signal, m is the mth pixel in the image domain, and m is M is the total number of grids declared in step 2, s o (τ(t' k,n ,m),t' k,n ) is the uniform time sampling of the echo declared in step 1, τ(t' k,n ,m) is the slow time t' in azimuth k,n The round-trip delay of the echo from the EPC to the mth pixel is the distance to the fast time, is the sampling time point of the nth channel in the azimuth direction at the kth pulse repetition period, T r is the pulse repetition time, K is the total number of sampling periods of the azimuth echo;
[0055] Using formula The BP imaging result of the mth scattering point is calculated and recorded as I(m), where m is the mth pixel in the image domain, and a low-resolution imaging image I is obtained;
[0056] Step 7: Calculate the target area of the low-resolution image
[0057] Using formula The threshold factor is calculated and recorded as α, where P fa is the false alarm probability defined in step 3;
[0058] The unit to be detected is denoted as I(x0,y0), where x0 represents the number of rows of the matrix corresponding to the image, and y0 represents the number of columns of the matrix corresponding to the image. The formula is used The calculated mean value of the background clutter on the left is X, using the formula The calculated background clutter mean on the right is Y, where x represents the number of rows of the matrix corresponding to the image, y represents the number of columns of the matrix corresponding to the image, and N gc is the number of protection unit columns defined in step 3, N gr is the number of protection unit rows defined in step 3, N tc is the number of background clutter columns defined in step 3, N tr is the number of background clutter rows defined in step 3;
[0059] Using the formula Z = max(X,Y), the background clutter mean is calculated to be Z;
[0060] The detection threshold is calculated as T using the formula T = αZ;
[0061] The target region is extracted from the low-resolution image I obtained in step 6. If I(x0,y0)>T, it is considered that there is a target at the image I(x0,y0), and the center positions of multiple target regions {c1,…,c q ,…,c Q}、 Distance width {rl1,…,rl q ,…,rl Q} and the azimuth width {al1,…,al q ,…,al Q}, where q is the qth target extracted, and the value of q is 1, 2, ..., Q;
[0062] Step 8: Full-aperture data BP imaging of the target area
[0063] Using formula The number of pixels in the range direction of high-resolution BP imaging is calculated as NR, using the formula The calculated number of pixels in the azimuth direction of high-resolution BP imaging is NA, where N r is the number of distance pixels defined in step 2, N a is the number of pixels in the azimuth direction defined in step 2, M is the total number of grids defined in step 2, dr is the grid resolution in the distance direction defined in step 2, da is the grid resolution in the azimuth direction defined in step 2, rl q is the distance width of the qth target area obtained in step 7, a1 q is the azimuth width of the qth target area obtained in step 7;
[0064] The original echo data S obtained in step 4 is processed by using the standard synthetic aperture radar echo data range pulse compression method in definition 4. r (τ(t,m),t) performs range pulse compression to obtain multi-channel SAR echo data after range compression, denoted as s r (τ(t,m),t), i.e., the echo after range compression in step 1, where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain;
[0065] Using formula The sub-image of the mth scattering point in channel n is calculated and recorded as where f cis the center frequency of the transmitted signal, m is the mth pixel in the image domain, and m is s r (τ(t k,n ,m),t k,n ) is the uniform time sampling of the echo declared in step 1, τ(t k,n ,m) is the slow time t in azimuth k,n The round-trip delay of the echo from the EPC to the mth pixel is the distance to the fast time, is the sampling time point of the nth channel in the azimuth direction at the kth pulse repetition period, T r is the pulse repetition time, K is the total number of sampling periods of the azimuth echo;
[0066] Using formula The BP imaging result of the mth scattering point is calculated and recorded as I q (m), obtain the high-resolution image I of the qth target area q ;
[0067] This completes the entire method.
[0068] The innovation of the present invention is that the target of interest in the high-resolution wide-band SAR observation scene is sparse in space and has a high signal-to-noise ratio relative to the background. The target position is found and extracted by a detection method, and a multi-level resolution fast imaging method for multi-channel high-resolution wide-band SAR is provided. The method uses de-slant and sub-aperture imaging to obtain low-resolution echo data for fast imaging, then uses the constant false alarm method to extract the target, and finally uses the original echo data to perform high-resolution imaging of the target, overcoming the problem of slow imaging speed of the traditional BP algorithm for the imaging of sparse targets in large scenes for the entire observation scene, and is more suitable for high-resolution wide-band SAR fast imaging of large scenes with unknown observation targets in actual situations, and improves the computational efficiency relative to the BP algorithm, and improves the imaging quality relative to other fast imaging algorithms.
[0069] The advantage of the present invention is that it realizes a multi-level resolution fast imaging algorithm for multi-channel high-resolution wide-band SAR through target detection based on the BP algorithm, and adopts the method of performing target detection on the low-resolution imaging results to obtain the region of interest and then performing high-resolution imaging instead of directly performing high-resolution imaging on the entire scene. This algorithm not only retains the advantage of the BP algorithm in imaging quality over the frequency domain algorithm, but also overcomes the disadvantage of the BP algorithm being slow in high-resolution imaging of large scenes. Compared with other algorithms, this algorithm has the advantages of high imaging quality and high computational efficiency. The present invention can be applied to fields such as synthetic aperture radar imaging and earth remote sensing. BRIEF DESCRIPTION OF THE DRAWINGS
[0070] Figure 1The present invention provides a schematic block diagram of the processing flow of the high-resolution, wide-width, multi-level resolution SAR rapid imaging method.
[0071] Figure 2 This is a geometric relationship diagram of multi-channel SAR imaging.
[0072] Figure 2 The middle rectangular box represents the antenna, d is the channel spacing, R1, R2, R3, and R4 represent four channels, x, y, and z represent the horizontal, vertical, and height coordinates in the observed scene target space, respectively, and 0 represents the coordinate origin in the observed scene target space.
[0073] Figure 3 This is a table of simulation parameters of a multi-channel SAR system used in a specific implementation of the present invention.
[0074] Figure 4 This is a time table for the multi-level resolution rapid imaging method of the present invention. DETAILED DESCRIPTION
[0075] The present invention is mainly verified by simulation experiment method, and all steps and conclusions are verified correct on MATLAB R2019a. The specific implementation steps are as follows:
[0076] Step 1: Initialize the multi-channel SAR system parameters required for simulation:
[0077] Initialize the values of multi-channel SAR system parameters such as Figure 3 As shown, including: motion platform velocity vector The number of antenna channels N = 4, the beam squint angle θ = 0, the platform height H = 0m, and the initial position vector of each antenna channel Where n is the serial number of the nth channel antenna, n = 1, 2, ..., N, N = 4, the maximum baseline length L = 0.48m, the channel spacing d = 0.16m, the radar center frequency f c =10GHz, the bandwidth of the radar transmission signal B = 150MHz, the pulse width of the radar transmission signal T r =6×10 -6 s, radar transmission signal frequency modulation slope K r =2.5×10 13 Hz / s, sampling frequency f of radar receiving system s =210MHz, the pulse repetition frequency of the radar system transmitting signal PRF = 637Hz, the effective aperture length of the antenna in azimuth direction D a =2m, the speed of light in air is C = 3×10 8 m / s; the sampling time of the nth channel satisfying uniform time sampling in the kth pulse repetition period is Where k represents the pulse repetition period, and takes values of k=1, 2, ..., K, where K=5000 is the total number of sampling periods of the azimuth echo, and n represents the channel, and takes values of 1, 2, ..., N, where N=4 is the number of channels of the multi-channel SAR stated above; the echo is denoted by S r (τ(t,m),t), the echo after range compression is s r (τ(t,m),t), where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the mth pixel at the azimuth slow time t, and m is the mth pixel in the image domain. The uniform time sampling of the echo can be expressed as s r (τ(t k,n ,m),t k,n ), where t k,n is the azimuth slow time, τ(t k,n ,m) is the slow time t in azimuth k,n The distance to the fast time of the round-trip delay of the echo from EPC to the mth pixel at the time; Step 2, initialize the observation scene target space parameters of the multi-channel SAR:
[0078] Initialize the scene space parameters of high-resolution wide-band SAR multi-level resolution imaging, including: scene center slant distance r o =100km, the maximum observation distance of the radar system r max = 105km, distance to the number of pixels N r =10000, the number of pixels in the azimuth direction is N a =10000, total number of grids M = 10 8 , the grid resolution in distance direction is dr = 1m, and the grid resolution in azimuth direction is da = 1m;
[0079] Add multiple simulated scattered aircraft targets in the observation scene target space Ω, and their scattering coefficient values are all 1; the range width of the radar system observation target area △r = 2km; the scattering coefficient of the cell that does not contain a point target in the observation scene target space Ω is set to 0; in this simulated observation scene target space Ω, only the scattering coefficient value α of the cell containing the scattered aircraft target is set to 1, and the scattering coefficients of the remaining cells are all 0;
[0080] Step 3: Initialize the constant false alarm detection parameters:
[0081] False alarm probability P fa =0.001, number of protection unit columns N gc =30, number of protection unit rows N gr =30, number of background clutter columns N tc =30, number of background clutter rows N tr =30;
[0082] Step 4: Generate received echo data and reference signal:
[0083] Using formula The reference signal is calculated and recorded as S ref (t), where t is the azimuth slow time, τ0 = 2r0 / c, r0 is the slant distance from the center of the scene, c is the speed of light, and T' = 2r max / c,r max is the maximum observation distance of the radar system, T is the pulse width of the radar transmission signal, and f c is the radar center frequency, K r is the frequency modulation slope of the radar transmission signal, rect[·] is the matrix window function;
[0084] The original echo data of multi-channel SAR is generated by using the original echo simulation method of synthetic aperture radar in Definition 11, denoted as S r (τ(t,m),t), the echo declared in step 1, where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain;
[0085] Step 5: Perform range de-skewing to obtain a small bandwidth echo:
[0086] Using formula Using the original echo signal S obtained in step 4 r (τ(t,m),t) and the reference signal S ref (t), the small bandwidth echo after de-slanting is calculated and recorded as S o (t), where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain. is the reference signal S ref (t) conjugation;
[0087] Using formula The narrow bandwidth of the echo after de-slanting is calculated and recorded as B', where K r is the radar system transmission modulation slope, △r is the range width of the radar system observation target area, and c is the speed of light;
[0088] Step 6: Azimuth sub-aperture BP imaging
[0089] Using formula Using the narrowband bandwidth B' obtained in step 5, calculate the resolution of the low-resolution data after range deskewing, denoted as ρ', where c is the speed of light;
[0090] Using formula The resolution of the high-resolution data before deslanting in the range direction is calculated and recorded as ρ, where c is the speed of light and B is the bandwidth of the radar transmission signal;
[0091] The formula p = ρ' / ρ is used to calculate the sampling multiple of the sub-aperture imaging resolution relative to the full-aperture imaging resolution, which is denoted as p.
[0092] Using formula The number of pixels in the range direction of low-resolution BP imaging is calculated and recorded as N r ', using the formula Calculate the number of azimuth pixels of low-resolution BP imaging, denoted as N a ', using the formula The total number of grids for low-resolution BP imaging is calculated, denoted as M'. The range grid resolution of low-resolution BP imaging is calculated using the formula dr' = p·dr, denoted as dr'. The azimuth grid resolution of low-resolution BP imaging is calculated using the formula da' = p·da, denoted as da', where N r is the number of distance pixels defined in step 2, N a is the number of pixels in the azimuth direction defined in step 2, M is the total number of grids defined in step 2, dr is the grid resolution in the distance direction defined in step 2, and da is the grid resolution in the azimuth direction defined in step 2;
[0093] In order to ensure the uniformity of the resolution in the range and azimuth directions, the formula is used Using the de-slanted echo S obtained in step 5 o (t), the sub-aperture echo is calculated and recorded as S o (t'), where t' is the azimuth slow time, t is the azimuth slow time, and K = 5000 is the total number of sampling periods of the azimuth echo;
[0094] The sub-aperture echo data S is compressed using the range-direction pulse compression method of the standard synthetic aperture radar echo data in Definition 4. o (t') performs range pulse compression to obtain multi-channel SAR echo data after range compression, denoted as s o (τ(t',m),t'), where t' is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t', and m is the m-th pixel in the image domain;
[0095] Using formula The sub-image of the mth scattering point in channel n is calculated and recorded as I n (m), where f c =10GHz is the center frequency of the transmitted signal, m is the mth pixel in the image domain, and m is M is the total number of grids declared in step 2, s o (τ(t' k,n ,m),t' k,n ) is the uniform time sampling of the echo declared in step 1, τ(t' k,n ,m) is the slow time t' in azimuth k,n The round-trip delay of the echo from the EPC to the mth pixel is the distance to the fast time, is the sampling time point of the nth channel in the azimuth direction at the kth pulse repetition period, T r is the pulse repetition time, K=5000 is the total number of sampling cycles of the azimuth echo;
[0096] Using formula The BP imaging result of the mth scattering point is calculated and recorded as I(m), where m is the mth pixel in the image domain, and a low-resolution imaging image I is obtained;
[0097] Step 7: Calculate the target area of the low-resolution image
[0098] Using formula The threshold factor is calculated and recorded as α, where P fa is the false alarm probability defined in step 3;
[0099] The unit to be detected is denoted as I(x0,y0), where x0 represents the number of rows of the matrix corresponding to the image, and y0 represents the number of columns of the matrix corresponding to the image. The formula is used The mean value of the background clutter on its left side is calculated to be X, using the formula The mean value of the background clutter on the right side is calculated to be Y, where x represents the number of rows of the matrix corresponding to the image, y represents the number of columns of the matrix corresponding to the image, and N gc is the number of protection unit columns defined in step 3, N gr is the number of protection unit rows defined in step 3, N tc is the number of background clutter columns defined in step 3, N tr is the number of background clutter rows defined in step 3;
[0100] The background clutter mean is calculated using the formula Z = max(X,Y), and the detection threshold is calculated using the formula T = αZ.
[0101] The target region is extracted from the low-resolution image I obtained in step 6. If I(x0,y0)>T, it is considered that there is a target at the image I(x0,y0), and the center positions of multiple target regions {c1,…,c q ,…,c Q}、 Distance width {rl1,…,rl q ,…,rl Q} and the azimuth width {al1,…,al q ,…,al Q}, where q is the qth target extracted, and the value of q is 1, 2, ..., Q;
[0102] Step 8: Full-aperture data BP imaging of the target area
[0103] Using formula The number of pixels in the range direction of high-resolution BP imaging is calculated as NR, using the formula The calculated azimuth pixel number of high-resolution BP imaging is NA, where N r is the number of distance pixels defined in step 2, N a is the number of pixels in the azimuth direction defined in step 2, M is the total number of grids defined in step 2, dr is the grid resolution in the distance direction defined in step 2, da is the grid resolution in the azimuth direction defined in step 2, rl q is the distance width of the qth target area obtained in step 7, a1 q is the azimuth width of the qth target area obtained in step 7;
[0104] The original echo data S obtained in step 4 is processed by using the standard synthetic aperture radar echo data range pulse compression method in definition 4. r (τ(t,m),t) performs range pulse compression to obtain multi-channel SAR echo data after range compression, denoted as s r (τ(t,m),t), i.e., the range-compressed echo declared in step 1, where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain;
[0105] Using formula The sub-image of the mth scattering point in channel n is calculated and recorded as where f c =10GHz is the center frequency of the transmitted signal, m is the mth pixel in the image domain, and m is s r (τ(t k,n ,m),t k,n ) is the uniform time sampling of the echo declared in step 1, τ(t k,n ,m) is the slow time t in azimuth k,n The round-trip delay of the echo from the EPC to the mth pixel is the distance to the fast time, is the sampling time point of the nth channel in the azimuth direction at the kth pulse repetition period, T r is the pulse repetition time, K=5000 is the total number of sampling cycles of the azimuth echo;
[0106] Then use the formula The BP imaging result of the mth scattering point is calculated and recorded as I q (m), obtain the high-resolution image I of the qth target area q ;
[0107] This completes the entire method.
[0108] It can be seen from the specific implementation methods of the present invention that the present invention proposes a multi-level resolution fast imaging method for multi-channel high-resolution wide-width SAR, that is, for high-resolution wide-width sparse scene imaging, firstly perform low-resolution fast imaging on the entire scene, then quickly detect the target area based on the imaging result, and then perform high-resolution imaging on the target area, thereby quickly obtaining a high-resolution image of the target in the large scene. The method first uses de-skewing processing in the distance direction to obtain a small bandwidth echo, and uses sub-aperture imaging in the azimuth direction to reduce the amount of calculation to obtain fast low-resolution imaging, and uses constant false alarm as a detection method based on this result to detect and extract the target area, and finally performs high-resolution imaging of the target area with large bandwidth and full-aperture echo data. Figure 4 As shown in the figure, the proposed method has a much higher imaging efficiency than traditional imaging methods while maintaining high-resolution imaging, which is an obvious advantage over traditional imaging methods.
Claims
1. A multi-level resolution fast imaging method for multi-channel high-resolution wide-band SAR, characterized in that it The following steps are involved: Step 1: Initialize the multi-channel SAR system parameters required for high-resolution wide-band SAR multi-level resolution imaging: Initialize the multi-channel SAR system parameters required for high-resolution wide-band SAR multi-level resolution imaging, including: the observation space of the multi-channel SAR is the ground three-dimensional coordinate system, denoted by XYZ, where X represents the horizontal axis, Y represents the horizontal axis, and Z represents the horizontal vertical axis; the bandwidth of the radar transmit signal is denoted by B, the pulse width of the radar transmit signal is denoted by T, and the frequency modulation slope of the radar transmit signal is denoted by K r , the sampling frequency of the radar receiving system is recorded as f s ; The multi-channel SAR imaging system moves in a uniform linear motion along the Y-axis parallel direction, and the platform velocity vector is recorded as The platform flight altitude is recorded as h; the number of channels of the multi-channel SAR is recorded as N, and the first channel transmitting the signal is recorded as T x , which is also the first channel R that receives the echo x1 ; The remaining channels only receive echoes, which are denoted by R x2 ,R x3 ,...,R xN ; The radar center frequency is denoted as f c , the pulse repetition frequency of the radar system's transmitted signal is recorded as PRF, and the pulse repetition time of the transmitted signal is recorded as PRI; the distance between each channel is recorded as d; the sampling time of the nth channel that satisfies uniform time sampling in the kth pulse repetition period is recorded as Where k represents the pulse repetition period, and its value is k=1,2,...,K; n represents the channel, and its value is 1,2,...,N; the echo is denoted by S r (τ(t,m),t), the echo after range compression is s r (τ(t,m),t), where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the mth pixel at the azimuth slow time t, and m is the mth pixel in the image domain; then the uniform time sampling of the echo can be expressed as s r (τ(t k,n ,m),t k,n ), where t k,n is the azimuthal slow time, τ(t k,n ,m) is the slow time t in azimuth k,n The distance to the fast time is the round-trip delay of the echo from the EPC to the mth pixel; Step 2: Initialize the observation scene target space parameters of high-resolution wide-band SAR multi-level resolution imaging: Initialize the scene space parameters of high-resolution wide-band SAR multi-level resolution imaging, including: the slant distance of the scene center is recorded as r o , the maximum observation distance of the radar system is recorded as r max The range width of the target area observed by the radar system is recorded as △r, and the number of pixels in the range is recorded as N r The number of pixels in the azimuth direction is recorded as N a , the total number of grids is recorded as M, the distance grid resolution is recorded as dr, and the azimuth grid resolution is recorded as da; Step 3: Initialize the constant false alarm detection parameters: The initial false alarm probability is denoted as P fa , the number of protection unit columns is recorded as N gc , the number of protection unit rows is recorded as N gr , the number of background clutter columns is recorded as N tc , the number of background clutter rows is recorded as N tr ; Step 4: Generate received echo data and reference signal: Using formula The reference signal is calculated and recorded as S ref (t), where t is the azimuth slow time, τ0 = 2r0 / c, r0 is the slant distance from the center of the scene, c is the speed of light, and T' = 2r max / c,r max is the maximum observation distance of the radar system, T is the pulse width of the radar transmission signal, and f c is the radar center frequency, K r is the frequency modulation slope of the radar transmission signal, rect[·] is the matrix window function; The traditional synthetic aperture radar raw echo simulation method is used to generate the raw echo data of multi-channel SAR, denoted as S r (τ(t,m),t), i.e., the echo in step 1, where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain; Step 5: Perform range de-skewing to obtain a small bandwidth echo: For the original echo signal S obtained in step 4 r (τ(t,m),t) and the reference signal S ref (t), using the formula The small bandwidth echo after de-slanting is calculated and recorded as S o (t), where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain. is the reference signal S ref (t) conjugation; Using formula The narrow bandwidth of the echo after de-slanting is calculated and recorded as B', where K r is the radar system transmission modulation slope, △r is the range width of the radar system observation target area, and c is the speed of light; Step 6: Azimuth sub-aperture BP imaging For the narrowband bandwidth B' obtained in step 5, use the formula The resolution of the low-resolution data after range deslanting is calculated and is denoted as ρ', where c is the speed of light; Using formula The resolution of the high-resolution data before deslanting in the range direction is calculated and recorded as ρ, where c is the speed of light and B is the bandwidth of the radar transmission signal; The formula p = ρ' / ρ is used to calculate the sampling multiple of the sub-aperture imaging resolution relative to the full-aperture imaging resolution, which is denoted as p. Using formula The number of pixels in the range direction of low-resolution BP imaging is calculated and recorded as N r '; Using formula Calculate the number of azimuth pixels of low-resolution BP imaging, denoted as N a '; Using formula The total number of grids for low-resolution BP imaging is calculated and recorded as M'; The range grid resolution of low-resolution BP imaging is calculated using the formula dr' = p·dr, denoted as dr'; The azimuth grid resolution of low-resolution BP imaging is calculated using the formula da' = p·da, denoted as da', where N r is the number of pixels in the distance direction in step 2, N a is the number of pixels in the azimuth direction in step 2, M is the total number of grids in step 2, dr is the grid resolution in the distance direction in step 2, and da is the grid resolution in the azimuth direction in step 2; For the de-slanted echo S obtained in step 5 o (t), using the formula The sub-aperture echo is calculated and recorded as S o (t'), where t' is the azimuth slow time, t is the azimuth slow time, and K is the total number of sampling periods of the azimuth echo; The standard synthetic aperture radar echo data range pulse compression method is used to compress the sub-aperture echo data S o (t') performs range pulse compression to obtain multi-channel SAR echo data after range compression, denoted as s o (τ(t',m),t'), where t' is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t', and m is the m-th pixel in the image domain; Using formula The sub-image of the mth scattering point in channel n is calculated and recorded as I n (m), where f c is the center frequency of the transmitted signal, m is the mth pixel in the image domain, and m is M is the total number of grids declared in step 2, s o (τ(t' k,n ,m),t' k,n ) is the uniform time sampling of the echo declared in step 1, τ(t' k,n ,m) is the slow time t' in azimuth k,n The round-trip delay of the echo from the EPC to the mth pixel is the distance to the fast time, is the sampling time point of the nth channel in the azimuth direction at the kth pulse repetition period, T r is the pulse repetition time, K is the total number of sampling periods of the azimuth echo; Using formula The BP imaging result of the mth scattering point is calculated and recorded as I(m), where m is the mth pixel in the image domain, and a low-resolution imaging image I is obtained; Step 7: Calculate the target area of the low-resolution image Using formula The threshold factor is calculated and recorded as α, where P fa is the false alarm probability in step 3; The unit to be detected is denoted as I(x0,y0), where x0 represents the number of rows of the matrix corresponding to the image, and y0 represents the number of columns of the matrix corresponding to the image. The formula is used The calculated mean value of the background clutter on the left is X, using the formula The calculated background clutter mean on the right is Y, where x represents the number of rows of the matrix corresponding to the image, y represents the number of columns of the matrix corresponding to the image, and N gc is the number of protection unit columns in step 3, N gr is the number of protection unit rows in step 3, N tc is the number of background clutter columns in step 3, N tr is the number of background clutter rows in step 3; Using the formula Z=max(X,Y), the background clutter mean is calculated to be Z; The detection threshold is calculated as T using the formula T = αZ; The target region is extracted from the low-resolution image I obtained in step 6. If I(x0,y0)>T, it is considered that there is a target at the image I(x0,y0), and the center positions of multiple target regions {c1,…,c q ,…,c Q }、 Distance width {rl1,…,rl q ,…,rl Q } and the azimuth width {al1,…,al q ,…,al Q }, where q is the qth target extracted, and the value of q is 1, 2, ..., Q; Step 8: Full-aperture data BP imaging of the target area Using formula The number of pixels in the range direction of high-resolution BP imaging is calculated as NR, using the formula The calculated azimuth pixel number of high-resolution BP imaging is NA, where N r is the number of pixels in the distance direction in step 2, N a is the number of pixels in the azimuth direction in step 2, M is the total number of grids in step 2, dr is the grid resolution in the distance direction in step 2, da is the grid resolution in the azimuth direction in step 2, rl q is the distance width of the qth target area obtained in step 7, a1 q is the azimuth width of the qth target area obtained in step 7; The original echo data S obtained in step 4 is processed by using the standard synthetic aperture radar echo data range pulse compression method. r (τ(t,m),t) performs range pulse compression to obtain multi-channel SAR echo data after range compression, denoted as s r (τ(t,m),t), i.e., the echo after range compression in step 1, where t is the azimuth slow time, τ is the range fast time of the round-trip delay of the echo from the EPC to the m-th pixel at the azimuth slow time t, and m is the m-th pixel in the image domain; Using formula The sub-image of the mth scattering point in channel n is calculated and recorded as where f c is the center frequency of the transmitted signal, m is the mth pixel in the image domain, and m is s r (τ(t k,n ,m),t k,n ) is the uniform time sampling of the echo declared in step 1, τ(t k,n ,m) is the slow time t in azimuth k,n The round-trip delay of the echo from the EPC to the mth pixel is the distance to the fast time, is the sampling time point of the nth channel in the azimuth direction at the kth pulse repetition period, T r is the pulse repetition time, K is the total number of sampling periods of the azimuth echo; Using formula The BP imaging result of the mth scattering point is calculated and recorded as I q (m), obtain the high-resolution image I of the qth target area q ; This completes the entire method.
