CN111239731A - Synthetic aperture radar rapid imaging method and device based on neural network - Google Patents
Synthetic aperture radar rapid imaging method and device based on neural network Download PDFInfo
- Publication number
- CN111239731A CN111239731A CN202010008334.9A CN202010008334A CN111239731A CN 111239731 A CN111239731 A CN 111239731A CN 202010008334 A CN202010008334 A CN 202010008334A CN 111239731 A CN111239731 A CN 111239731A
- Authority
- CN
- China
- Prior art keywords
- imaging
- resolution
- neural network
- synthetic aperture
- aperture radar
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/417—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section involving the use of neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention discloses a synthetic aperture radar rapid imaging method based on a neural network, which comprises the steps of using a back projection algorithm to carry out low-resolution synthetic aperture radar imaging, and using an image reconstruction model to carry out high-resolution reconstruction on the obtained low-resolution image to obtain a corresponding high-resolution image; the image reconstruction model is obtained by pre-training through the following method: respectively carrying out low-resolution and high-resolution synthetic aperture radar imaging on the target by using a back projection algorithm to obtain a training sample; and then training the neural network by respectively taking the low-resolution and high-resolution images in a series of different training samples as input values and real values. The invention also discloses a synthetic aperture radar rapid imaging device based on the neural network. Compared with the traditional BP imaging algorithm, the method has the advantages of small calculated amount, high operation rate, good robustness and strong adaptability.
Description
Technical Field
The invention relates to the technical field of radar imaging, in particular to a synthetic aperture radar imaging method.
Background
Radar imaging technology is widely applied to various fields of military use and civil use. With the increasing diversification of imaging scenes and the increasing demand of applications, radar imaging technology needs to be advanced accordingly.
Synthetic Aperture Radar (SAR) is a commonly used radar imaging method, and the imaging resolution is improved by constructing an equivalent large observation aperture. In synthetic aperture imaging, frequency domain algorithms such as Range Doppler (RD) and ω -k need to have an ideal uniform aperture, which has strict requirements on the motion state of a radar platform. The Back Projection (BP) algorithm is an imaging algorithm based on time domain processing, and ensures that echo information corresponding to each point is aligned in time by calculating and compensating the two-way time delay between a receiving and transmitting antenna and a target, and then overlapping imaging is carried out on imaging pixel points. Therefore, the imaging algorithm can be suitable for various imaging systems and scenes while obtaining high imaging precision, and is easy to compensate motion.
Compared with other SAR imaging algorithms, the BP algorithm is simple to implement and does not need complex signal processing such as motion compensation and the like. At present, the BP imaging algorithm is widely applied to the fields of through-wall radar, airborne radar, ground penetrating radar and the like. But instead of the other end of the tube
As is known, the defects of BP imaging are large operation amount, low operation efficiency and slow imaging time. How to realize rapid BP
Algorithm imaging is popular research content in the field of synthetic aperture radar, and is proposed by a plurality of experts and scholars at home and abroad
And a corresponding fast implementation method. BOAG in the document "Boag A, Bresler Y, Michielsen E. A multilevel domain composition for fast project of tomographic images [ J ]. Image Processing, IEEE Transactions on, 2000, 9(9): 1573-1582" proposes a fast BP imaging algorithm, which is based on the principle of blocking an imaging scene area, and then reduces the calculation amount of the BP imaging algorithm by a step-by-step hierarchical coherent accumulation method to achieve the purpose of reducing the imaging time. The document 'Wangma, Zhou Zheng Europe, Liting Jun et al, ultra wide band pulse through-the-wall radar cross-correlation BP imaging [ J ]. proceedings of university of electronic technology, 2011, 40(1): 16-19.' the Wangma et al, based on the characteristics of ultra wide band pulse ground penetrating radar imaging, provides an improved Fast Back projection imaging algorithm (FBP), which obviously improves the BP imaging operation speed. The disadvantage of these algorithms is that although a certain amount of BP algorithm computation is reduced, the requirements of real-time high-precision imaging in the aspects of practicability, robustness, imaging quality and the like are still not met.
Disclosure of Invention
The invention aims to overcome the defects of the traditional BP algorithm and provides a synthetic aperture radar rapid imaging method based on a neural network.
The invention specifically adopts the following technical scheme to solve the technical problems:
the method comprises the steps of using a back projection algorithm to carry out low-resolution synthetic aperture radar imaging, and using an image reconstruction model to carry out high-resolution reconstruction on an obtained low-resolution image to obtain a corresponding high-resolution image; the image reconstruction model is obtained by pre-training through the following method: respectively carrying out low-resolution and high-resolution synthetic aperture radar imaging on the target by using a back projection algorithm to obtain a training sample; and then training the neural network by respectively taking the low-resolution and high-resolution images in a series of different training samples as input values and real values.
According to the same inventive concept, the following technical scheme can be obtained:
synthetic aperture radar fast imaging device based on neural network includes:
the BP imaging module is used for carrying out low-resolution synthetic aperture radar imaging by using a back projection algorithm;
the image reconstruction model is used for carrying out high-resolution reconstruction on the obtained low-resolution image to obtain a corresponding high-resolution image; the image reconstruction model is obtained by pre-training through the following method: respectively carrying out low-resolution and high-resolution synthetic aperture radar imaging on the target by using a back projection algorithm to obtain a training sample; and then training the neural network by respectively taking the low-resolution and high-resolution images in a series of different training samples as input values and real values.
Preferably, the neural network is a convolutional neural network.
Further preferably, the convolutional neural network comprises five convolutional layers.
Preferably, the low-resolution and high-resolution synthetic aperture radar imaging is performed on the target by using a back projection algorithm, specifically as follows: detecting a given imaging area and a target by using M real apertures, and respectively performing matched filtering processing on received echoes to obtain M one-dimensional range profiles of the imaging area; dividing an imaging area into imaging grids consisting of N multiplied by N pixels, and then respectively mapping the N multiplied by N pixels and the N multiplied by N pixels in the imaging grids with M one-dimensional distance images through a back projection algorithm to respectively obtain M coarse images, wherein M, N, N are integers which are more than 2 and N is more than N; and finally, accumulating the obtained coarse images one by one to obtain a high-resolution image of the NxN pixel points and a low-resolution image of the nxn pixel points.
Preferably, the loss function of the neural network is a mean square error.
Compared with the prior art, the technical scheme of the invention has the following beneficial effects:
the invention reduces the operation cost of BP algorithm and improves the imaging rate by reducing the number of the needed imaging grids; and the method of the neural network ensures that the imaging quality is still higher while the imaging is carried out rapidly, and the method has good robustness and strong adaptability.
Drawings
FIG. 1 is a schematic diagram of a method for rapid imaging of a synthetic aperture radar based on a neural network according to the present invention;
FIG. 2 is a low resolution image before high resolution reconstruction in an embodiment of the present invention;
fig. 3 is a high-resolution image after high-resolution reconstruction according to an embodiment of the present invention.
Detailed Description
Aiming at the problems of large computation amount, low computation efficiency and slow imaging time of the traditional BP imaging algorithm, the basic idea of the invention is as follows: the calculation cost of the BP algorithm is reduced by reducing the number of grid points participating in imaging, so that the calculation rate is greatly improved; by utilizing the neural network, the imaging quality of the proposed algorithm is ensured while the operation rate is improved.
The synthetic aperture radar fast imaging method based on the neural network uses a back projection algorithm to carry out low-resolution synthetic aperture radar imaging, and uses an image reconstruction model to carry out high-resolution reconstruction on the obtained low-resolution image to obtain a corresponding high-resolution image; the image reconstruction model is obtained by pre-training through the following method: respectively carrying out low-resolution and high-resolution synthetic aperture radar imaging on the target by using a back projection algorithm to obtain a training sample; and then training the neural network by respectively taking the low-resolution and high-resolution images in a series of different training samples as input values and real values.
The neural network can adopt a BP neural network, an RBF neural network, a circulating neural network and the like. However, the effects of different neural network models on different application scenes are greatly different, and therefore, through a large number of experiments, the convolutional neural network is preferably adopted to construct the image reconstruction model, the convolutional neural network can directly process the two-dimensional image, and the lattice-point characteristics can be stably and effectively learned with a small calculated amount.
For the public understanding, the technical scheme of the invention is further explained in detail by a specific embodiment:
firstly, a training sample set needs to be acquired and an image reconstruction model needs to be trained, the principle of which is shown in fig. 1, and the specific process is as follows:
s1, carrying out BP imaging on dense pixel points for the pre-imaging area to obtain a training set high resolution image HR;
in the embodiment, M real apertures are used for detecting a given imaging area and a target, and the received echoes are respectively subjected to matched filtering processing to obtain M one-dimensional range profiles of the imaging area; dividing an imaging region into an imaging grid composed of N × N pixels; then mapping the imaging grid and the M one-dimensional range profiles by an interpolation method in a BP algorithm to obtain M rough images; and finally, accumulating the obtained coarse images one by one to obtain a final high-resolution imaging result HR of the NxN pixel points.
The parameters may be specifically determined according to the specific imaging system, for example, the number of real apertures M is set to 15, and the number of grid points N × N is set to 256 × 256.
S2, selecting a small number of pixel points in the same imaging area to perform BP imaging to obtain a training set low-resolution image LR;
in this embodiment, for the same imaging scene as in S1, the real aperture number is unchanged, and the received echo is subjected to matched filtering processing to obtain M one-dimensional range profiles; keeping the size of the imaging area set in the S1 unchanged, only selecting n multiplied by n pixel points in the original imaging grid to participate in imaging calculation, and mapping the divided imaging grid and M one-dimensional range profiles with each other by an interpolation method in a BP algorithm to obtain M coarse images; and finally, accumulating the obtained coarse images one by one to obtain a low-resolution imaging result LR consisting of n multiplied by n pixel points.
The number of grid points n × n selected in this step may be set to 64 × 64.
S3, constructing a neural network, corresponding LR-HR one by one, respectively serving as an input value and a true value of the neural network, and training the neural network;
in this embodiment, before performing the neural network, training samples need to be obtained, and network structure setting, loss function setting, and parameter setting are performed, which are specifically as follows:
s31, obtaining a training sample;
in this embodiment, imaging targets are randomly placed in the same imaging area, and after each placement, two steps S1 and S2 are performed on the imaging area, respectively, to obtain a set of corresponding HR and LR imaging results; and changing the target positions for H times, calculating to obtain H composition image results, and corresponding the H high-resolution images and the H low-resolution images one by one to form H training samples.
Preferably, the number H of training samples is set to 1000.
S32, constructing a neural network, and setting a network structure;
in this embodiment, the neural network includes three parts, namely, feature extraction, nonlinear mapping, and image reconstruction. A total of five convolutional layers are included, the first four of which are followed by a ReLU layer, which introduces non-linearity in the network. The last convolutional layer is followed by a regression layer to reconstruct the image and output it as a network.
And S33, constructing a loss function.
The invention aims to construct a mapping from a low-resolution image (LR) to a high-resolution image (HR), so that the difference between the output image of the network and the original high-resolution image HR is minimized, which is the basis for constructing a Loss function in the invention, and therefore, the neural network adopts mean square error to carry out Loss function setting.
And S34, setting network parameters.
In this embodiment, the initial learning rate is set to 0.1, training is finished when the learning rate is attenuated to the termination learning rate, the attenuation multiple of the learning rate is set to 0.1, a Loss function is optimized by using a random gradient descent momentum (SGDM) optimizer, the weight value and the offset value of the neural network are determined, and the mapping relationship is determined.
In the actual imaging process of the synthetic aperture radar, a high-quality imaging result can be quickly obtained only by selecting a small number of pixel points of a target to be imaged to perform BP imaging and then inputting the obtained low-resolution image into a trained neural network (namely, the image reconstruction model).
In this embodiment, when imaging a target in a designated pre-imaging area, a low-resolution imaging result is obtained according to S2, and then a corresponding high-resolution imaging result can be obtained quickly by using the neural network trained in S3. Fig. 2 and 3 show a low-resolution image and a high-resolution image before and after high-resolution reconstruction, respectively. Compared with the traditional BP imaging algorithm, the method has the advantages that the operation rate is greatly improved, and the final imaging quality is ensured.
Claims (10)
1. A synthetic aperture radar fast imaging method based on a neural network is characterized in that a back projection algorithm is used for carrying out low-resolution synthetic aperture radar imaging, and an image reconstruction model is used for carrying out high-resolution reconstruction on an obtained low-resolution image to obtain a corresponding high-resolution image; the image reconstruction model is obtained by pre-training through the following method: respectively carrying out low-resolution and high-resolution synthetic aperture radar imaging on the target by using a back projection algorithm to obtain a training sample; and then training the neural network by respectively taking the low-resolution and high-resolution images in a series of different training samples as input values and real values.
2. The method for rapid imaging of synthetic aperture radar based on neural network as claimed in claim 1, wherein the neural network is a convolutional neural network.
3. The method of claim 2 wherein the convolutional neural network comprises five convolutional layers.
4. The method for fast imaging of synthetic aperture radar based on neural network as claimed in claim 1, wherein the low resolution and high resolution synthetic aperture radar imaging is performed on the target by using a back projection algorithm, respectively, specifically as follows: detecting a given imaging area and a target by using M real apertures, and respectively performing matched filtering processing on received echoes to obtain M one-dimensional range profiles of the imaging area; dividing an imaging area into imaging grids consisting of N multiplied by N pixels, and then respectively mapping the N multiplied by N pixels and the N multiplied by N pixels in the imaging grids with M one-dimensional distance images through a back projection algorithm to respectively obtain M coarse images, wherein M, N, N are integers which are more than 2 and N is more than N; and finally, accumulating the obtained coarse images one by one to obtain a high-resolution image of the NxN pixel points and a low-resolution image of the nxn pixel points.
5. The method for rapid imaging of synthetic aperture radar based on neural network as claimed in claim 1, wherein the loss function of the neural network is mean square error.
6. A synthetic aperture radar fast imaging device based on a neural network is characterized by comprising:
the BP imaging module is used for carrying out low-resolution synthetic aperture radar imaging by using a back projection algorithm;
the image reconstruction model is used for carrying out high-resolution reconstruction on the obtained low-resolution image to obtain a corresponding high-resolution image; the image reconstruction model is obtained by pre-training through the following method: respectively carrying out low-resolution and high-resolution synthetic aperture radar imaging on the target by using a back projection algorithm to obtain a training sample; and then training the neural network by respectively taking the low-resolution and high-resolution images in a series of different training samples as input values and real values.
7. The synthetic aperture radar fast imaging apparatus based on neural network as claimed in claim 6, wherein the neural network is a convolutional neural network.
8. The neural network-based synthetic aperture radar fast imaging apparatus as claimed in claim 7, wherein the convolutional neural network comprises five convolutional layers.
9. The apparatus according to claim 6, wherein the low-resolution and high-resolution synthetic aperture radar imaging is performed on the target by using a back projection algorithm, specifically as follows: detecting a given imaging area and a target by using M real apertures, and respectively performing matched filtering processing on received echoes to obtain M one-dimensional range profiles of the imaging area; dividing an imaging area into imaging grids consisting of N multiplied by N pixels, and then respectively mapping the N multiplied by N pixels and the N multiplied by N pixels in the imaging grids with M one-dimensional distance images through a back projection algorithm to respectively obtain M coarse images, wherein M, N, N are integers which are more than 2 and N is more than N; and finally, accumulating the obtained coarse images one by one to obtain a high-resolution image of the NxN pixel points and a low-resolution image of the nxn pixel points.
10. The neural network-based synthetic aperture radar fast imaging apparatus as claimed in claim 6, wherein the loss function of the neural network is a mean square error.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010008334.9A CN111239731A (en) | 2020-01-06 | 2020-01-06 | Synthetic aperture radar rapid imaging method and device based on neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010008334.9A CN111239731A (en) | 2020-01-06 | 2020-01-06 | Synthetic aperture radar rapid imaging method and device based on neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111239731A true CN111239731A (en) | 2020-06-05 |
Family
ID=70870801
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010008334.9A Pending CN111239731A (en) | 2020-01-06 | 2020-01-06 | Synthetic aperture radar rapid imaging method and device based on neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111239731A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113433130A (en) * | 2021-07-12 | 2021-09-24 | 清华大学 | Method and device for generating confocal imaging by wide-field imaging |
CN113435510A (en) * | 2021-06-28 | 2021-09-24 | 西安电子科技大学 | Inverse synthetic aperture radar image classification method based on sequence adjustment network |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101639937A (en) * | 2009-09-03 | 2010-02-03 | 复旦大学 | Super-resolution method based on artificial neural network |
CN106228512A (en) * | 2016-07-19 | 2016-12-14 | 北京工业大学 | Based on learning rate adaptive convolutional neural networks image super-resolution rebuilding method |
CN107680113A (en) * | 2017-10-27 | 2018-02-09 | 武汉大学 | The image partition method of multi-layer segmentation network based on Bayesian frame edge prior |
US20180260688A1 (en) * | 2014-07-16 | 2018-09-13 | Deep Learning Analytics, LLC | Systems and Methods for Recognizing Objects in Radar Imagery |
CN108734660A (en) * | 2018-05-25 | 2018-11-02 | 上海通途半导体科技有限公司 | A kind of image super-resolution rebuilding method and device based on deep learning |
CN108872988A (en) * | 2018-07-12 | 2018-11-23 | 南京航空航天大学 | A kind of inverse synthetic aperture radar imaging method based on convolutional neural networks |
CN109270525A (en) * | 2018-12-07 | 2019-01-25 | 电子科技大学 | Through-wall radar imaging method and system based on deep learning |
CN109685717A (en) * | 2018-12-14 | 2019-04-26 | 厦门理工学院 | Image super-resolution rebuilding method, device and electronic equipment |
RU2698649C1 (en) * | 2018-01-16 | 2019-08-29 | Акционерное общество "Федеральный научно-производственный центр "Нижегородский научно-исследовательский институт радиотехники" | Method of detecting and classifying small objects on images obtained by synthetic aperture radar stations |
US20190333199A1 (en) * | 2018-04-26 | 2019-10-31 | The Regents Of The University Of California | Systems and methods for deep learning microscopy |
-
2020
- 2020-01-06 CN CN202010008334.9A patent/CN111239731A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101639937A (en) * | 2009-09-03 | 2010-02-03 | 复旦大学 | Super-resolution method based on artificial neural network |
US20180260688A1 (en) * | 2014-07-16 | 2018-09-13 | Deep Learning Analytics, LLC | Systems and Methods for Recognizing Objects in Radar Imagery |
CN106228512A (en) * | 2016-07-19 | 2016-12-14 | 北京工业大学 | Based on learning rate adaptive convolutional neural networks image super-resolution rebuilding method |
CN107680113A (en) * | 2017-10-27 | 2018-02-09 | 武汉大学 | The image partition method of multi-layer segmentation network based on Bayesian frame edge prior |
RU2698649C1 (en) * | 2018-01-16 | 2019-08-29 | Акционерное общество "Федеральный научно-производственный центр "Нижегородский научно-исследовательский институт радиотехники" | Method of detecting and classifying small objects on images obtained by synthetic aperture radar stations |
US20190333199A1 (en) * | 2018-04-26 | 2019-10-31 | The Regents Of The University Of California | Systems and methods for deep learning microscopy |
CN108734660A (en) * | 2018-05-25 | 2018-11-02 | 上海通途半导体科技有限公司 | A kind of image super-resolution rebuilding method and device based on deep learning |
CN108872988A (en) * | 2018-07-12 | 2018-11-23 | 南京航空航天大学 | A kind of inverse synthetic aperture radar imaging method based on convolutional neural networks |
CN109270525A (en) * | 2018-12-07 | 2019-01-25 | 电子科技大学 | Through-wall radar imaging method and system based on deep learning |
CN109685717A (en) * | 2018-12-14 | 2019-04-26 | 厦门理工学院 | Image super-resolution rebuilding method, device and electronic equipment |
Non-Patent Citations (5)
Title |
---|
GUANQUN SUN; FANGZHENG ZHANG; SHILONG PAN: "《Improved Incoherent Back Projection Imaging based on Self-amplitude Weighting and Multiplicative Tomography Weighting》", 《2019 IEEE MTT-S INTERNATIONAL WIRELESS SYMPOSIUM (IWS)》 * |
李皓月: "《视频合成孔径雷达动目标成像方法》", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 * |
杨晓波 等著: "《太赫兹雷达》", 31 December 2017, 国防工业出版社 * |
杨欣 著: "《图像超分辨率技术原理及应用》", 30 April 2013, 国防工业出版社 * |
龙永新; 叶秀清; 顾伟康: "《多准则图象重建的神经网络模型及实现》", 《信号处理》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113435510A (en) * | 2021-06-28 | 2021-09-24 | 西安电子科技大学 | Inverse synthetic aperture radar image classification method based on sequence adjustment network |
CN113435510B (en) * | 2021-06-28 | 2024-02-09 | 西安电子科技大学 | Reverse synthetic aperture radar image classification method based on sequence adjustment network |
CN113433130A (en) * | 2021-07-12 | 2021-09-24 | 清华大学 | Method and device for generating confocal imaging by wide-field imaging |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107229048B (en) | High-resolution wide-range SAR moving target speed estimation and imaging method | |
CN109669184B (en) | Synthetic aperture radar azimuth ambiguity elimination method based on full convolution network | |
CN110095775B (en) | Hybrid coordinate system-based bump platform SAR (synthetic Aperture Radar) rapid time domain imaging method | |
CN112099008A (en) | SA-ISAR imaging and self-focusing method based on CV-ADMMN | |
CN111239731A (en) | Synthetic aperture radar rapid imaging method and device based on neural network | |
CN113567982B (en) | Directional periodic sampling data sparse SAR imaging method and device based on mixed norm | |
CN104316923B (en) | Self-focusing method aiming at synthetic aperture radar (Back Projection) imaging | |
CN110208796B (en) | Scanning radar super-resolution imaging method based on singular value inverse filtering | |
CN103364770B (en) | Radar target detecting system based on matrix filling and detecting method thereof | |
CN106405552A (en) | WVD-PGA algorithm based SAR object focusing method | |
CN103698765A (en) | ISAR (inverse synthetic aperture radar) imaging orientation calibration method | |
CN112147608A (en) | Rapid Gaussian gridding non-uniform FFT through-wall imaging radar BP method | |
CN109856636B (en) | Curve synthetic aperture radar self-adaptive three-dimensional imaging method | |
CN108957452A (en) | A kind of adaptive FFBP imaging method of synthetic aperture radar | |
CN109188436B (en) | Efficient bistatic SAR echo generation method suitable for any platform track | |
CN106097292A (en) | Sequential SAR space-time neighborhood Gauss Weighted median filtering speckle is made an uproar suppression fast algorithm | |
CN107831473A (en) | Based on Gaussian process recurrence apart from instantaneous Doppler image sequence noise-reduction method | |
CN105759264B (en) | Fine motion target defect echo high-resolution imaging method based on time-frequency dictionary | |
CN108594196B (en) | Method and device for extracting parameters of target scattering center | |
CN106910178B (en) | Multi-angle SAR image fusion method based on tone statistical characteristic classification | |
CN107783111A (en) | A kind of radar foresight super-resolution imaging method based on maximum entropy criterion | |
CN116243313A (en) | SAR rapid intelligent sparse self-focusing technology based on distance partition | |
CN106791405A (en) | A kind of method and system for obtaining Displacement sequence image | |
CN114895305B (en) | L-based 1 Norm regularized sparse SAR self-focusing imaging method and device | |
CN112965064B (en) | PCSBL-GAMP-Net-based block sparse aperture ISAR imaging method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200605 |