CN113238201B - Super-resolution radar positioning method, system, equipment and storage medium - Google Patents
Super-resolution radar positioning method, system, equipment and storage medium Download PDFInfo
- Publication number
- CN113238201B CN113238201B CN202110580036.1A CN202110580036A CN113238201B CN 113238201 B CN113238201 B CN 113238201B CN 202110580036 A CN202110580036 A CN 202110580036A CN 113238201 B CN113238201 B CN 113238201B
- Authority
- CN
- China
- Prior art keywords
- radar
- image
- resolution
- detection area
- selecting
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 88
- 238000001514 detection method Methods 0.000 claims description 70
- 238000003384 imaging method Methods 0.000 claims description 60
- 239000011159 matrix material Substances 0.000 claims description 52
- 238000007781 pre-processing Methods 0.000 claims description 20
- 238000005070 sampling Methods 0.000 claims description 20
- 238000011084 recovery Methods 0.000 claims description 19
- 238000004364 calculation method Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 10
- 238000004891 communication Methods 0.000 claims description 9
- 238000013507 mapping Methods 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 8
- 230000009467 reduction Effects 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000013461 design Methods 0.000 claims description 3
- 238000004806 packaging method and process Methods 0.000 claims description 3
- 238000000638 solvent extraction Methods 0.000 claims 3
- 238000004422 calculation algorithm Methods 0.000 description 14
- 230000008569 process Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 12
- 238000013528 artificial neural network Methods 0.000 description 10
- 238000012549 training Methods 0.000 description 9
- 230000006855 networking Effects 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000010287 polarization Effects 0.000 description 2
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Abstract
The embodiment of the invention provides a super-resolution radar positioning method, a system, equipment and a storage medium. Thus, compared with the prior art, the method can obtain the radar image with high resolution without a strong assumption, and has strong self-adaption capability for different environments. In addition, in the embodiment of the invention, the radar image is processed in super-resolution based on grid division, and a high-resolution radar image is obtained by utilizing the low-resolution echo intensity image of each radar, so that the problem of uneven resolution of the radar image is solved.
Description
Technical Field
The invention relates to the field of radar imaging positioning, in particular to a super-resolution radar positioning method, a super-resolution radar positioning system, super-resolution radar positioning equipment and a storage medium.
Background
Existing radar super-resolution techniques include: a deconvolution-based radar angle imaging method, a maximum posterior-based scanning radar angle resolution imaging method, and the like.
The deconvolution-based radar angle imaging method can break through the limitation of antenna system parameters on the resolution of radar images, and radar angle super-resolution imaging is achieved. But it is based on a strong assumption that the angular resolution is poor in adaptation under different circumstances by solving an inversion problem in the complex domain.
The scanning radar angle resolution imaging method based on the maximum posterior probability (MAP) is used for realizing radar angle super resolution imaging. However, the method of maximum posterior probability is proposed based on bayesian theory, and related statistics are required to be assumed to be independently distributed, but radar measurement data cannot be guaranteed to meet the requirement. In addition, the method is also poor in self-adaptive capacity and cannot solve the problem of uneven radar resolution.
Disclosure of Invention
In order to solve the technical problems, the invention provides a super-resolution radar positioning method and a super-resolution radar positioning system, which solve the problems that the existing radar angle resolution imaging technology is poor in self-adaptive capability in different environments and cannot solve the problem of uneven radar resolution.
In order to achieve the above object, the embodiment of the present invention provides the following technical solutions:
In a first aspect, the present invention provides a method for positioning a super-resolution radar, which is characterized by comprising: detecting the same detection area through at least one radar to obtain echo signal spectrograms of all the radars; respectively establishing a global coordinate system and a radar echo reference coordinate system in a plane where the detection area is located based on the detection area and the position information of each radar; in the global coordinate system, configuring a division rule according to preset parameters of each radar, and dividing the detection area into a plurality of imaging surface elements by using the division rule; mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, wherein the resolution of the echo intensity image is a first preset resolution; preprocessing the echo intensity image to generate an image sequence data packet; performing image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution; wherein the first preset resolution is less than the second preset resolution.
Further, detecting the same detection area by at least one radar to obtain echo signal spectrograms of the radars, including: configuring preset parameters of a plurality of radars; detecting the detection areas through each radar respectively to obtain echo signals; sampling and fast Fourier transform processing are carried out on each radar echo signal, and echo spectrograms of each radar are output; wherein the plurality of radars are arranged at two sides of the detection area, and the arrangement method of the plurality of radars comprises the following steps: respectively selecting a first boundary point and a second boundary point on a first side and a second side which are opposite to each other in the detection area; a first side radar arrangement line and a second side radar arrangement line are respectively obtained based on the first boundary point and the second boundary point, wherein the first side radar arrangement line and the second side radar arrangement line are boundary lines of the detection area, or only one intersection point exists among the first side radar arrangement line, the second side radar arrangement line and the detection area, and the first side radar arrangement line and the second side radar arrangement line are parallel to each other; at least one radar is respectively arranged on the first side radar arrangement line and the second side radar arrangement line; the service coverage of each radar completely covers the detection area.
Further, establishing the global coordinate system includes: acquiring position information of the detection area and position information of each radar, and numbering each radar on the first side radar arrangement line or the second side radar arrangement line in sequence; selecting the first side radar arrangement line or the second side radar arrangement line as an x-axis; selecting an arrangement direction of radar numbers from small to large as a positive direction of an x-axis; selecting the minimum vertical projection point of each point of the detection area on the x axis as an origin; and selecting a vertical line of the x axis as a y axis based on the origin, and selecting a direction pointing to the other side as a positive direction of the y axis.
Further, establishing the radar echo reference coordinate system includes: acquiring position information of each radar; selecting the position of the radar as a pole; and selecting a main lobe of the radar as a polar axis.
Further, in the global coordinate system, a division rule is configured according to preset parameters of each radar, and the detection area is divided into a plurality of imaging surface elements by using the division rule, including: selecting the distance resolution as a first dividing step length in the polar axis direction; selecting the angular resolution as a second dividing step length of the angular direction; taking a pole as a starting point, and performing grid division in the polar axis direction by utilizing the first division step length; taking a polar angle equal to 0 DEG as a starting point, and performing grid division in the angular direction by utilizing the second division step length; selecting nodes in the range covered by the detection area from the divided grids, and determining a plurality of imaging surface elements; wherein the total number of meshes of the imaging bin is a product of the first number of nodes in the polar axis direction and the second number of nodes in the angular direction.
Further, mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, wherein the resolution of the echo intensity image is a first preset resolution, and the method comprises the following steps: selecting the geometric center of the imaging surface element as a pixel point, wherein the coordinates of the pixel point are the coordinates of the geometric center in the global coordinate system; selecting the intensity of echo signals in a single imaging surface element as the brightness of the pixel point; and acquiring pixel point coordinates and brightness of pixel points corresponding to each imaging surface element in the detection area, and generating an echo intensity image with the first preset resolution.
Further, preprocessing the echo intensity image to generate an image sequence data packet, including: time alignment is carried out on echo intensity images corresponding to all the radars, and a time tag is generated; performing initial sorting according to coordinates of pixel points in each echo intensity image to generate pixel point cataloging information; processing the pixel point data into an image sequence according to the pixel point cataloging information; packaging all image sequences through an algorithm such as a maximum entropy spectrum estimation algorithm or an optimal linear filtering algorithm and the like to obtain an image sequence data packet; wherein, the design parameters of the data packet include: radar number, time tag, pixel cataloging information and brightness of corresponding pixels.
Further eliminating abnormal values which deviate from the target true value seriously in the image sequence; image sequences at preset absolute times of all radars are acquired.
Further, performing image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution, and the method comprises the following steps: acquiring first image matrixes corresponding to all radars from the image sequence data packet, wherein the number of rows and the number of columns of the first image matrixes are respectively preset number of rows W and preset number of columns L, and the number of output channels of the first image matrixes is preset number of channels N; extracting local features of each first image matrix to obtain a second image matrix, wherein the number of rows and the number of columns of the second image matrix are respectivelyAndA is a pooling dimension reduction parameter; upsampling each second image matrix to obtain a third image matrix, wherein the number of rows and columns of the third image matrix are respectivelyAndB is an up-sampling dimension-increasing parameter; fusing all the third image matrixes to output a single fourth image matrix, wherein the number of rows and the number of columns of the fourth image matrix are respectivelyAndB is an up-sampling dimension-increasing parameter, and the number of output channels of the fourth image matrix is 1; the fourth image matrix is sampled at equal intervals to obtain a fifth image matrix, and the number of rows and columns of the fifth image matrix are respectivelyAndC is an equal sampling interval; the radar image is generated based on the fifth image matrix.
In a second aspect, a super-resolution radar positioning system includes: the communication module is used for detecting the same detection area through at least one radar and acquiring echo signal spectrograms of the radars; the coordinate system establishing module is used for respectively establishing a global coordinate system and a radar echo reference coordinate system in a plane where the detection area is located based on the detection area and the position information of each radar; the imaging bin dividing module is used for configuring dividing rules according to preset parameters of each radar in the global coordinate system, and dividing the detection area into a plurality of imaging bins by using the dividing rules; the preprocessing module is used for mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, and the resolution of the echo intensity image is a first preset resolution; preprocessing the echo intensity image to generate an image sequence data packet; the image recovery calculation module is used for carrying out image super-resolution recovery calculation by utilizing the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution; wherein the first preset resolution is less than the second preset resolution.
In a third aspect, a super-resolution radar positioning apparatus includes: a memory for storing a computer program; a processor for implementing the steps of any one of the above super-resolution radar positioning methods when executing the computer program.
In a fourth aspect, a computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of a super-resolution radar positioning method according to any of the preceding claims.
The embodiment of the invention provides a super-resolution radar positioning method, a system, equipment and a storage medium. Thus, compared with the prior art, the method can obtain the radar image with high resolution without a strong assumption, and has strong self-adaption capability for different environments. In addition, in the embodiment of the invention, the radar image is processed in super-resolution based on grid division, and a high-resolution radar image is obtained by utilizing the low-resolution echo intensity image of each radar, so that the problem of uneven resolution of the radar image is solved.
Drawings
FIG. 1 is a schematic diagram of a super-resolution radar positioning system according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a super-resolution radar positioning method according to an embodiment of the present invention;
FIG. 3 is a flowchart of a method for obtaining a spectrum of each radar according to an embodiment of the present invention;
fig. 4 is a flowchart illustrating a process for establishing a global coordinate system of a super-resolution radar positioning method according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a binning flow of a super-resolution radar positioning method according to an embodiment of the present invention;
FIG. 6 is a schematic flow chart of a method for establishing a radar echo reference coordinate system according to an embodiment of the present invention;
Fig. 7 is a schematic flow chart of generating a first preset resolution of a super-resolution radar positioning method according to an embodiment of the present invention;
fig. 8 is a schematic diagram of a preprocessing flow of a super-resolution radar positioning method according to an embodiment of the present invention;
fig. 9 is a schematic diagram of a denoising process in a super-resolution radar positioning method according to an embodiment of the present invention;
Fig. 10 is a schematic flow chart of image super-resolution recovery processing of a super-resolution radar positioning method according to an embodiment of the present invention;
FIG. 11 is a schematic diagram of radar deployment of a super-resolution radar positioning method according to an embodiment of the present invention;
FIG. 12 is a schematic diagram of binning according to one embodiment of the present invention;
FIG. 13 is a schematic flow chart of a super resolution algorithm according to an embodiment of the present invention.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It will be apparent to those skilled in the art from this disclosure that the drawings described below are merely exemplary and that other embodiments may be derived from the drawings provided without undue effort.
The structures, proportions, sizes, etc. shown in the present specification are shown only for the purposes of illustration and description, and are not intended to limit the scope of the invention, which is defined by the claims, so that any structural modifications, changes in proportions, or adjustments of sizes, which do not affect the efficacy or the achievement of the present invention, should fall within the ambit of the technical disclosure.
The abbreviations and key term definitions in the examples of the present invention are explained below:
And (3) radar networking: the radar networking is to properly and reasonably distribute a plurality of radars with different frequency bands, different polarization modes and different constitutions, collect and transmit information (original signals, points, track data and the like) of each radar in the network in a 'network' mode, and comprehensively process, control and manage the information by a central station, thereby forming a unified, organic and integral radar system. The coverage capability of a time domain, a frequency domain and a space domain of the system is enlarged through radar networking, the advantages of each radar are brought into play, information sharing is realized, the speed of finding a target is effectively improved, false alarms and missed alarms are reduced, the four-antibody capability of the radar which is leased and hung in a severe electronic warfare environment is comprehensively improved, and the survivability of the radar is obviously enhanced.
Super resolution technology: the resolution of the original image is improved through software and hardware, and the super-resolution reconstruction is the process of obtaining a high-resolution image through a series of low-resolution images. The core idea of super-resolution reconstruction is to use time bandwidth (to acquire multi-frame image sequences of the same scene) to exchange spatial resolution, so as to realize conversion from the temporal resolution to the spatial resolution.
FFT: the fast Fourier transform is a fast algorithm of the discrete Fourier transform, and is obtained by improving the algorithm of the discrete Fourier transform according to the characteristics of the discrete Fourier transform, such as odd, even, virtual, real and the like.
Radar echo intensity image: spatial distribution of echo signal intensity amplitude in the radar plane.
Referring to fig. 1, one embodiment of the present invention provides a super-resolution radar positioning system including: a communication module 01, a coordinate establishment module 02, an imaging face element division module 03, a preprocessing module 04 and an image restoration calculation module 05. Further, the communication module 01 is configured to detect the same detection area through at least one radar, and obtain an echo signal spectrogram of each radar; a coordinate system establishing module 02, configured to establish a global coordinate system and a radar echo reference coordinate system in a plane where the detection area is located based on the detection area and position information of each radar; the imaging bin dividing module 03 is configured to configure a dividing rule according to preset parameters of each radar in the global coordinate system, and divide the detection area into a plurality of imaging bins by using the dividing rule; the preprocessing module 04 is used for mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, wherein the resolution of the echo intensity image is a first preset resolution; preprocessing the echo intensity image to generate an image sequence data packet; the image recovery calculation module 05 is configured to perform image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, where the resolution of the radar image is a second preset resolution; wherein the first preset resolution is less than the second preset resolution.
Specifically, the communication module 01 is connected with the coordinate establishment module 02, and sends the detection area position and the position of each radar to the coordinate establishment module 02; the communication module 01 is connected with the preprocessing module to send the echo signal spectrogram to the preprocessing module 04; the coordinate establishing module 02 processes the received detection area position and the position of each radar and divides the detection area position and the position into a global coordinate system and a radar echo reference coordinate system; the coordinate establishing module 02 is connected with the imaging surface element dividing module 03, and sends the information of the global coordinate system and the information of the radar echo reference coordinate system to the imaging surface element dividing module 03, and the imaging surface element dividing module 03 configures dividing rules in the global coordinate system according to preset parameters of each radar and divides the imaging surface element into a plurality of imaging surface elements by utilizing a detection area; the preprocessing module 04 receives the echo signal spectrogram from the communication module 01 and imaging bin information of the imaging bin dividing module 03, maps the signal spectrogram into the imaging bin to obtain a discretized echo intensity image with a first preset resolution, and then preprocesses the echo intensity image to generate an image sequence data packet and sends the image sequence packet to the image recovery calculation module 05; the image recovery calculation module 05 receives the image sequence data packet from the preprocessing module 04, and performs image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image with a second preset resolution.
According to the embodiment of the invention, after a plurality of radars are networked, a target area is detected to obtain a spectrogram, then the spectrogram is mapped into an imaging surface element, and then image super-resolution recovery calculation is carried out to obtain a radar image. Thus, compared with the prior art, the method can obtain the radar image with high resolution without a strong assumption, and has strong self-adaption capability for different environments. In addition, in the embodiment of the invention, the radar image is processed in super-resolution based on grid division, and a high-resolution radar image is obtained by utilizing the low-resolution echo intensity image of each radar, so that the problem of uneven resolution of the radar image is solved.
Corresponding to the above disclosed super-resolution radar positioning system, the embodiment of the invention also discloses a super-resolution radar positioning method. The following describes a super-resolution radar positioning method disclosed in the embodiment of the present invention in detail in conjunction with the above-described system for positioning a super-resolution radar.
Referring to fig. 2, a flow chart of a super-resolution radar positioning method according to an embodiment of the present invention is shown; detecting the same detection area through at least one radar to obtain echo signal spectrograms of all the radars; respectively establishing a global coordinate system and a radar echo reference coordinate system in a plane where the detection area is located based on the detection area and the position information of each radar; in the global coordinate system, configuring a division rule according to preset parameters of each radar, and dividing the detection area into a plurality of imaging surface elements by using the division rule; mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, wherein the resolution of the echo intensity image is a first preset resolution; preprocessing the echo intensity image to generate an image sequence data packet; performing image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution; wherein the first preset resolution is less than the second preset resolution.
Preferably, referring to fig. 3, a flowchart of a radar spectrogram acquired by a super-resolution radar positioning method according to an embodiment of the present invention is shown; detecting the same detection area through at least one radar to obtain echo signal spectrograms of the radars, wherein the method comprises the following steps: configuring preset parameters of a plurality of radars; detecting the detection areas through each radar respectively to obtain echo signals; sampling and fast Fourier transform processing are carried out on each radar echo signal, and echo spectrograms of each radar are output; wherein the plurality of radars are arranged at two sides of the detection area, and the arrangement method of the plurality of radars comprises the following steps: respectively selecting a first boundary point and a second boundary point on a first side and a second side which are opposite to each other in the detection area; a first side radar arrangement line and a second side radar arrangement line are respectively obtained based on the first boundary point and the second boundary point, wherein the first side radar arrangement line and the second side radar arrangement line are boundary lines of the detection area, or only one intersection point exists among the first side radar arrangement line, the second side radar arrangement line and the detection area, and the first side radar arrangement line and the second side radar arrangement line are parallel to each other; at least one radar is respectively arranged on the first side radar arrangement line and the second side radar arrangement line; the service coverage of each radar completely covers the detection area.
Specifically, referring to fig. 11, a radar deployment schematic diagram of a super-resolution radar positioning method according to an embodiment of the present invention is illustrated below, where the arrangement method of each radar is assumed to be a rectangular area ABCD in a plane. Firstly, deployment and installation of networking radars are carried out, N radars with the same distance resolution are sequentially arranged on two sides of a detection area, each radar respectively detects the same coverage area, each radar respectively acquires echo signals of the radars, samples and FFT (fast Fourier transform) are carried out on the echo signals of the radars, and echo signal spectrograms of the radars are output.
Further, the preset parameters of the radar include: distance resolution and angle resolution; the range resolution of each radar is the same; the distance between any two radars on the same side is an integer multiple of the distance resolution, and the distance in the radar arrangement direction between any two radars on different sides is an integer multiple of the distance resolution.
According to the embodiment of the invention, the coverage capability of a time domain, a frequency domain and a space domain of a system is enlarged based on proper optimized layout of multiple radars with different frequency bands, different polarization modes and different constitutions, the advantages of the radars are brought into play, and the capability of acquiring information is enhanced; meanwhile, a high-resolution image is obtained through a series of low-resolution images by means of super-resolution technology, and the spatial resolution is replaced by the time bandwidth, so that the conversion of the time resolution phase spatial resolution is realized.
In addition, according to the embodiment of the invention, based on radar deployment of the target area, the radar deployment method fundamentally ensures uniform resolution and can obtain the original data which is convenient for post-processing.
Referring to fig. 4, a flow chart of a global coordinate system establishment method for a super-resolution radar positioning method according to an embodiment of the present invention is provided; establishing the global coordinate system comprises the following steps: acquiring position information of the detection area and position information of each radar, and numbering each radar on the first side radar arrangement line or the second side radar arrangement line in sequence; selecting the first side radar arrangement line or the second side radar arrangement line as an x-axis; selecting an arrangement direction of radar numbers from small to large as a positive direction of an x-axis; selecting the minimum vertical projection point of each point of the detection area on the x axis as an origin; and selecting a vertical line of the x axis as a y axis based on the origin, and selecting a direction pointing to the other side as a positive direction of the y axis.
Referring to fig. 6, a flow chart of a method for establishing a radar echo reference coordinate system for a super-resolution radar positioning method according to an embodiment of the present invention is shown; establishing the radar echo reference coordinate system comprises the following steps: acquiring position information of each radar; selecting the position of the radar as a pole; and selecting a main lobe of the radar as a polar axis.
Specifically, referring to fig. 11, a radar deployment schematic diagram of a super-resolution radar positioning method according to an embodiment of the present invention is provided; taking a plane coordinate system as an example, taking a point A as an origin o, and taking an AB direction as an x-axis positive direction to establish a plane rectangular coordinate system x-o-y; then, a plane polar coordinate system O Ly (R, θ) is established by taking a position R x where a radar L y (y is a radar number and the range of values is 1 to n) is located as a pole, and the polar axis is a main lobe direction corresponding to the radar L y.
According to the method for establishing the global coordinate system, the detection area is only in one quadrant, so that symbol bits can be recorded when a network is divided and network node coordinates are marked in the global coordinate system, and time expenditure of a subsequent algorithm is greatly reduced.
Referring to fig. 5, a bin-dividing flow diagram of a super-resolution radar positioning method according to an embodiment of the present invention is provided; in the global coordinate system, configuring a division rule according to preset parameters of each radar, dividing the detection area into a plurality of imaging surface elements by using the division rule, wherein the method comprises the following steps: selecting the distance resolution as a first dividing step length in the polar axis direction; selecting the angular resolution as a second dividing step length of the angular direction; taking a pole as a starting point, and performing grid division in the polar axis direction by utilizing the first division step length; taking a polar angle equal to 0 DEG as a starting point, and performing grid division in the angular direction by utilizing the second division step length; selecting nodes in the range covered by the detection area from the divided grids, and determining a plurality of imaging surface elements; wherein the total number of meshes of the imaging bin is a product of the first number of nodes in the polar axis direction and the second number of nodes in the angular direction.
In the embodiment of the invention, the detection area is divided into a plurality of imaging surface elements, the radar image to be processed is sampled, a discrete radar echo image is obtained, and the sampling interval is regulated by a first dividing step length and a second dividing step length; the sampling interval can be customized according to the actual application scene.
Referring to fig. 7, a flow chart of generating a first preset resolution of a super-resolution radar positioning method according to an embodiment of the present invention is shown; mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, wherein the resolution of the echo intensity image is a first preset resolution, and the method comprises the following steps of: selecting the geometric center of the imaging surface element as a pixel point, wherein the coordinates of the pixel point are the coordinates of the geometric center in the global coordinate system; selecting the intensity of echo signals in a single imaging surface element as the brightness of the pixel point; and acquiring pixel point coordinates and brightness of pixel points corresponding to each imaging surface element in the detection area, and generating an echo intensity image with the first preset resolution.
Referring to FIG. 12, a schematic diagram of binning according to one embodiment of the present invention is provided; the relation between the radar echo reference coordinate system and the image pixels is that each imaged pixel unit corresponds to a divided grid node, the initial position of the grid node divided on the polar axis is the origin of the radar echo reference coordinate system, the initial position of the grid node divided on the theta direction is the polar axis theta=0°, and the step length and the node number determining method are as described above, and the corresponding node in the service coverage area is selected from the divided grids for determining radar imaging surface elements. The radar measurement data are radar echo images, the pixel number of the radar echo images is the number of grid nodes of an area, the pixel division of the radar imaging area is shown in a fourth drawing, the imaging pixels correspond to the divided grid areas, the coordinates of the pixel points are the positions (x, y) of the geometric center of the imaging unit in a rectangular coordinate system x-o-y, the brightness of the pixel points represents the intensity of echo signals in a single pixel, the brightness of the pixel points exceeding the radar maximum detection range is specified, and the brightness of the pixel points without echo signals is 0.
The embodiment of the invention solves the problem that a radar image cannot be converted from polar coordinates to Cartesian coordinates in the prior art based on the division of the surface elements. And creating necessary conditions for the subsequent image quantization processing.
Referring to fig. 8, a schematic diagram of a preprocessing flow of a super-resolution radar positioning method according to an embodiment of the present invention is provided; preprocessing the echo intensity image to generate an image sequence data packet, including: time alignment is carried out on echo intensity images corresponding to all the radars, and a time tag is generated; performing initial sorting according to coordinates of pixel points in each echo intensity image to generate pixel point cataloging information; processing the pixel point data into an image sequence according to the pixel point cataloging information; packaging all image sequences through an algorithm such as a maximum entropy spectrum estimation algorithm or an optimal linear filtering algorithm and the like to obtain an image sequence data packet; wherein, the design parameters of the data packet include: radar number, time tag, pixel cataloging information and brightness of corresponding pixels.
Preferably, referring to fig. 9, a schematic diagram of a denoising process of a super-resolution radar positioning method according to an embodiment of the present invention is provided; removing abnormal values which deviate from a target true value seriously in the image sequence; image sequences at preset absolute times of all radars are acquired.
Specifically, given coordinates (x, y) of echo image pixel points corresponding to x-o-y coordinate systems, the pixel coordinates x and y are each the smallest and cataloged as (1, 1), when the pixel point x=x i (i=1, 2, …, weith, weith is the first node number in the polar axis direction), y is ordered from small to large, and when the pixel point y=y j (j=1, 2, …, length is the second node number in the angular direction), x is ordered from small to large.
The method provided by the embodiment of the invention maps the signal spectrogram into the imaging bin to obtain the discretized echo intensity image, which is a process of digitizing the signal, wherein the discretized echo intensity image consists of brightness and coordinates in a global coordinate system, is favorable for computing and processing a plurality of images in the same area and at the same time but with different dimensions, and improves a plurality of low-resolution images to a high-resolution image.
Referring to fig. 10, a flowchart of an image super-resolution recovery process of a super-resolution radar positioning method according to an embodiment of the present invention is shown; performing image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution, and the method comprises the following steps: acquiring first image matrixes corresponding to all radars from the image sequence data packet, wherein the number of rows and the number of columns of the first image matrixes are respectively preset number of rows W and preset number of columns L, and the number of output channels of the first image matrixes is preset number of channels N; extracting local features of each first image matrix to obtain a second image matrix, wherein the number of rows and the number of columns of the second image matrix are respectivelyAndA is a pooling dimension reduction parameter; upsampling each second image matrix to obtain a third image matrix, wherein the number of rows and columns of the third image matrix are respectivelyAndB is an up-sampling dimension-increasing parameter; fusing all the third image matrixes to output a single fourth image matrix, wherein the number of rows and the number of columns of the fourth image matrix are respectivelyAndB is an up-sampling dimension-increasing parameter, and the number of output channels of the fourth image matrix is 1; the fourth image matrix is sampled at equal intervals to obtain a fifth image matrix, and the number of rows and columns of the fifth image matrix are respectivelyAndC is an equal sampling interval; the radar image is generated based on the fifth image matrix.
According to the embodiment of the invention, the characteristic images are formed by the characteristic extracted by convolution of the neural network algorithm for a plurality of times, and then the high-resolution images required by the user are obtained by restoring the plurality of low-resolution characteristic images.
Referring to fig. 13, a flow chart of a super resolution algorithm according to an embodiment of the present invention is shown; the following illustrates the specific implementation steps of the super-division algorithm;
Step one: for the input N first image matrixes I 1,I2,...,IN, N different convolution kernels of 3×3 are used, filling with the length of 1 is carried out on the input N first image matrixes respectively, convolution is carried out with the step length of 1, local features in the 3×3 area in the original picture are extracted once, the size of the output picture is W.L, W is the preset number of rows of the first image matrixes, L is the preset number of columns of the first image matrixes, and the number of output channels is N.
Step two: filling the pictures with the length of 2, convoluting the pictures with the step length of 1 through a convolution kernel of 5 multiplied by 5, adding nonlinear factors to the network through an active layer ReLU layer, reducing the number of parameters, preventing the model from being over fitted, performing 1/2 downsampling through a 2 multiplied by 2 maximum pooling layer, wherein the maximum pooling layer is mainly used for feature dimension reduction, compressing the number of data and parameters, reducing the over fitting, improving the fault tolerance of the model, repeating the process twice, and outputting the pictures with the size ofThe number of channels output is N.
Step three: then convolving with step length of 1 and filling of 2 by convolution kernel of 5×5 twice to obtain second image matrix according to the size of the picture output by the second image matrixThe number of channels output is N.
Step four: up-sampling is performed by a 4 x 16 convolution kernel with a step size of 2. The convolution is then performed with a step size of 1 by a convolution kernel of 5 x 5. And repeating the process through the ReLU layer twice to obtain a third image matrix, wherein the size of the picture output according to the third image matrix is 4W.4L, and the number of the output channels is N.
Step five: finally, the multichannel is changed into single-channel output through a convolution kernel of 1 multiplied by 1. The fourth image matrix corresponding to the high-resolution image with the resolution of 4W.4L is obtained, and in practice, the resolution of the fourth image matrix can be adjusted by changing parameters such as the size of a network convolution kernel, the step size, pooling and the like.
Step six: and (3) sampling a fourth image matrix of 4 W.4L at equal intervals from the first row and the first column according to an equal sampling interval c to obtain a fifth image matrix, and obtaining the super-divided high-resolution radar image I k,Ik according to the fifth image matrix, wherein the resolution is kW.kL, and the resolution is a second preset resolution. This step simulates a mask brought about by the high resolution radar fence effect.
And outputting the processed radar image in real time, outputting radar echo intensity image data with the size of kW.kL after super-resolution recovery, specifically, visually displaying the processed digital image I k, normalizing the value of each pixel point to a section [0,255], and displaying through a gray scale map under a Cartesian coordinate system x-o-y.
The neural network parameter θ in the super-division algorithm in the above description is obtained through neural network training. The following describes the process of neural network parameter training:
The whole training process requires input: training data sets T and X 0 should contain multiple sets of training data. One group of t and x 0 are taken out for each training, t represents N low-resolution images, particularly N W.L matrixes, and x 0 represents high-resolution radar images corresponding to t, particularly kW.kL matrixes.
Specifically, let n=2, w=4, l=4;
step one: firstly, randomly initializing a weight parameter theta of the neural network, wherein the theta comprises values of all convolution kernels and transpose matrixes in the neural network.
Step two: for the N matrices t1, t2, t N of the input, the N matrices of the input are respectively filled with a length of 1 using N different convolution kernels of 3×3, convoluting with a step of 1, extracting local features in a 3×3 region in the original picture once, the size of the output picture is w·l, and the number of channels of the output is N.
Step three: filling the pictures with the length of 2, convoluting the pictures with the step length of 1 through a convolution kernel of 5 multiplied by 5, adding nonlinear factors to the network through an active layer ReLU layer, reducing the number of parameters, preventing the model from being over fitted, performing 1/2 downsampling through a2 multiplied by 2 maximum pooling layer, wherein the maximum pooling layer is mainly used for feature dimension reduction, compressing the number of data and parameters, reducing the over fitting, improving the fault tolerance of the model, repeating the process twice, and outputting the pictures with the size ofThe number of channels output is N.
Step four: then convolving with step length of 1 and filling of 2 by convolution kernel of 5×5 twice, and outputting picture with size of 1The number of channels output is N.
Step five: up-sampling is performed by a 4 x 16 convolution kernel with a step size of 2. The convolution is then performed with a step size of 1 by a convolution kernel of 5 x 5. And repeating the process twice through the ReLU layer, wherein the output picture size is 4W.4L, and the output channel number is N.
Step six: finally, the multichannel is changed into single-channel output through a convolution kernel of 1 multiplied by 1. A matrix t 0 corresponding to a high-resolution image with a resolution of 4w.4l is obtained.
Step seven: for a matrix t 0 of 4w.4l, equidistant sampling is performed at equidistant sampling intervals c starting from the first column of the first row. And obtaining the super-divided high-resolution radar image t sample,tsample, wherein the resolution is kW.kL, and the resolution is a second preset resolution. The method simulates a mask brought by a high-resolution radar fence effect, and enables the obtained matrix resolution to be consistent with a second preset resolution, so that parameters of the neural network can be updated later.
Step eight: and adding Gaussian noise to t samp to obtain a final expected high-definition image t k.tk＝tsamp+G.g～N(0,Pnoise), wherein G epsilon G is Gaussian distribution with the element obeying mean value of 0 and the variance of P noise in G. G is a matrix of k Weith*kLenth and P noise is radar noise power. The gaussian distribution belongs to mathematical general knowledge, and is not described herein, and how to generate random numbers according to the gaussian distribution is also a programming general technique, and is not described herein.
Step nine: and updating the neural network parameter theta according to the difference between the expected high-definition image t k obtained in the step eight and x 0 in the actual training data. Network parameters are trained using a deep learning framework, typically pytorch or TensorFlow can be selected.
Specifically, the neural network parameter θ specific update procedure in step nine is encapsulated in a deep learning framework and updated using a gradient descent method. As an example, a parameter update specific procedure is described: let t be a number, θ being the 2-1 row vector, z being the 1-2 column vector, a very simple network is assumed, t=θ -z. Assume training data asWhen the expected output is x 0 =2, θ is initialized to [ 12 ], when inputThe resulting output is 3, and θ is updated to [ 0.8.1.6 ], at which time the inputThe resulting output is 2.4, which is closer to the expected output than before the parameters were updated.
Step ten: and (3) repeating the second step to the ninth step until convergence by using a plurality of groups of training data to obtain a final neural network parameter theta.
The embodiment of the invention provides super-resolution radar positioning equipment, which comprises: a memory for storing a computer program; a processor for implementing the steps of a super resolution radar positioning method as described in any one of the above when executing the computer program.
An embodiment of the present invention provides a computer readable storage medium, where a computer program is stored on the computer readable storage medium, where the computer program when executed by a processor implements the steps of a super-resolution radar positioning method according to any one of the above.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus, device and unit described above may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus, device and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, e.g., the division of units is merely a logical function division, and there may be additional divisions in actual implementation, e.g., multiple units or components may be combined or integrated into another device, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interface, indirect coupling or communication connection of devices or units, electrical, mechanical, or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in whole or in part in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a function calling device, or a network device, etc.) to perform all or part of the steps of the methods of the embodiments of the present application. While the application has been described in detail in the foregoing general description and specific examples, it will be apparent to those skilled in the art that modifications and improvements can be made thereto. Accordingly, such modifications or improvements may be made without departing from the spirit of the application and are intended to be within the scope of the application as claimed.
Claims (10)
1. A super-resolution radar positioning method, comprising:
detecting the same detection area through at least one radar to obtain echo signal spectrograms of all the radars;
Respectively establishing a global coordinate system and a radar echo reference coordinate system in a plane where the detection area is located based on the detection area and the position information of each radar;
In the global coordinate system, configuring a division rule according to preset parameters of each radar, and dividing the detection area into a plurality of imaging surface elements by using the division rule;
mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, wherein the resolution of the echo intensity image is a first preset resolution;
Preprocessing the echo intensity image to generate an image sequence data packet;
Performing image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution;
wherein the first preset resolution is less than the second preset resolution;
Wherein said establishing said global coordinate system comprises:
acquiring position information of the detection area and position information of each radar, and numbering each radar on a first side radar arrangement line or a second side radar arrangement line in sequence;
selecting a first side radar arrangement line or a second side radar arrangement line as an x-axis;
selecting an arrangement direction of radar numbers from small to large as a positive direction of an x-axis;
Selecting the minimum vertical projection point of each point of the detection area on the x axis as an origin;
selecting a vertical line of the x-axis as a y-axis based on the origin, and selecting a direction pointing to the other side as a positive direction of the y-axis;
wherein the establishing the radar echo reference coordinate system includes:
acquiring position information of each radar;
selecting the position of the radar as a pole;
and selecting a main lobe of the radar as a polar axis.
2. The method for positioning a super-resolution radar according to claim 1, wherein the step of detecting the same detection area by at least one radar to obtain an echo signal spectrogram of each radar comprises:
configuring preset parameters of a plurality of radars;
detecting the detection areas through each radar respectively to obtain echo signals;
sampling and fast Fourier transform processing are carried out on each radar echo signal, and echo spectrograms of each radar are output;
Wherein the plurality of radars are arranged at two sides of the detection area, and the arrangement method of the plurality of radars comprises the following steps: respectively selecting a first boundary point and a second boundary point on a first side and a second side which are opposite to each other in the detection area; a first side radar arrangement line and a second side radar arrangement line are respectively obtained based on the first boundary point and the second boundary point, wherein the first side radar arrangement line and the second side radar arrangement line are boundary lines of the detection area, or only one intersection point exists among the first side radar arrangement line, the second side radar arrangement line and the detection area, and the first side radar arrangement line and the second side radar arrangement line are parallel to each other; at least one radar is respectively arranged on the first side radar arrangement line and the second side radar arrangement line; the service coverage of each radar completely covers the detection area.
3. The method according to claim 1, wherein a partitioning rule is configured according to preset parameters of each radar in the global coordinate system, wherein the configuration parameters include a distance resolution and an angle resolution, and the partitioning of the detection area into a plurality of imaging bins by using the partitioning rule includes:
selecting the distance resolution as a first dividing step length in the polar axis direction;
selecting the angular resolution as a second dividing step length of the angular direction;
Taking a pole as a starting point, and performing grid division in the polar axis direction by utilizing the first division step length;
Taking a polar angle equal to 0 DEG as a starting point, and performing grid division in the angular direction by utilizing the second division step length;
Selecting nodes in the range covered by the detection area from the divided grids, and determining a plurality of imaging surface elements;
wherein the total number of meshes of the imaging bin is a product of the first number of nodes in the polar axis direction and the second number of nodes in the angular direction.
4. The method of claim 1, wherein mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, the resolution of the echo intensity image being a first preset resolution, comprises:
selecting the geometric center of the imaging surface element as a pixel point, wherein the coordinates of the pixel point are the coordinates of the geometric center in the global coordinate system;
selecting the intensity of echo signals in a single imaging surface element as the brightness of the pixel point;
And acquiring pixel point coordinates and brightness of pixel points corresponding to each imaging surface element in the detection area, and generating an echo intensity image with the first preset resolution.
5. The method of claim 1, wherein preprocessing the echo intensity image to generate image sequence data packets comprises:
Time alignment is carried out on echo intensity images corresponding to all the radars, and a time tag is generated;
Performing initial sorting according to coordinates of pixel points in each echo intensity image to generate pixel point cataloging information;
Processing the pixel point data into an image sequence according to the pixel point cataloging information;
Packaging all the image sequences to obtain an image sequence data packet;
wherein, the design parameters of the data packet include: radar number, time tag, pixel cataloging information and brightness of corresponding pixels.
6. The method of claim 5, wherein preprocessing the echo intensity image to generate image sequence data packets, further comprises:
removing abnormal values which deviate from a target true value seriously in the image sequence;
image sequences at preset absolute times of all radars are acquired.
7. The method for positioning a super-resolution radar according to claim 1, wherein performing image super-resolution recovery calculation by using the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution, and the method comprises:
Acquiring first image matrixes corresponding to all radars from the image sequence data packet, wherein the number of rows and the number of columns of the first image matrixes are respectively preset number of rows W and preset number of columns L, and the number of output channels of the first image matrixes is preset number of channels N;
Extracting local features of each first image matrix to obtain a second image matrix, wherein the number of rows and the number of columns of the second image matrix are respectively
AndA is a pooling dimension reduction parameter;
upsampling each second image matrix to obtain a third image matrix, wherein the number of rows and columns of the third image matrix are respectively
AndB is an up-sampling dimension-increasing parameter;
Fusing all the third image matrixes to output a single fourth image matrix, wherein the number of rows and the number of columns of the fourth image matrix are respectively
AndB is the up-sampling up-dimension parameter,
The number of output channels of the fourth image matrix is 1;
The fourth image matrix is sampled at equal intervals to obtain a fifth image matrix, and the number of rows and columns of the fifth image matrix are respectively
AndC is an equal sampling interval;
the radar image is generated based on the fifth image matrix.
8. A super-resolution radar positioning system, comprising:
the communication module is used for detecting the same detection area through at least one radar and acquiring echo signal spectrograms of the radars;
the coordinate system establishing module is used for respectively establishing a global coordinate system and a radar echo reference coordinate system in a plane where the detection area is located based on the detection area and the position information of each radar;
The imaging bin dividing module is used for configuring dividing rules according to preset parameters of each radar in the global coordinate system, and dividing the detection area into a plurality of imaging bins by using the dividing rules;
the preprocessing module is used for mapping the signal spectrogram into the imaging bin to obtain a discretized echo intensity image, and the resolution of the echo intensity image is a first preset resolution; preprocessing the echo intensity image to generate an image sequence data packet;
the image recovery calculation module is used for carrying out image super-resolution recovery calculation by utilizing the image sequence data packet to obtain a radar image, wherein the resolution of the radar image is a second preset resolution;
wherein the first preset resolution is less than the second preset resolution;
Wherein said establishing said global coordinate system comprises:
acquiring position information of the detection area and position information of each radar, and numbering each radar on a first side radar arrangement line or a second side radar arrangement line in sequence;
selecting a first side radar arrangement line or a second side radar arrangement line as an x-axis;
selecting an arrangement direction of radar numbers from small to large as a positive direction of an x-axis;
Selecting the minimum vertical projection point of each point of the detection area on the x axis as an origin;
selecting a vertical line of the x-axis as a y-axis based on the origin, and selecting a direction pointing to the other side as a positive direction of the y-axis;
wherein the establishing the radar echo reference coordinate system includes:
acquiring position information of each radar;
selecting the position of the radar as a pole;
and selecting a main lobe of the radar as a polar axis.
9. A super-resolution radar positioning apparatus, characterized by comprising:
A memory for storing a computer program;
processor for implementing the steps of a super resolution radar positioning method according to any one of claims 1 to 7 when executing said computer program.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of a super resolution radar positioning method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110580036.1A CN113238201B (en) | 2021-05-26 | Super-resolution radar positioning method, system, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110580036.1A CN113238201B (en) | 2021-05-26 | Super-resolution radar positioning method, system, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113238201A CN113238201A (en) | 2021-08-10 |
| CN113238201B true CN113238201B (en) | 2024-06-28 |
Family
ID=
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050716A (en) * | 2014-06-25 | 2014-09-17 | 北京航空航天大学 | Marine multi-target SAR image visual modeling method |
| CN104535999A (en) * | 2015-01-02 | 2015-04-22 | 中国人民解放军国防科学技术大学 | Radar imaging data preprocessing method for correcting antenna directional pattern influences |
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050716A (en) * | 2014-06-25 | 2014-09-17 | 北京航空航天大学 | Marine multi-target SAR image visual modeling method |
| CN104535999A (en) * | 2015-01-02 | 2015-04-22 | 中国人民解放军国防科学技术大学 | Radar imaging data preprocessing method for correcting antenna directional pattern influences |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3120322B1 (en) | Method for processing input low-resolution (lr) image to output high-resolution (hr) image | |
| US10019642B1 (en) | Image upsampling system, training method thereof and image upsampling method | |
| Wei | Image super‐resolution reconstruction using the high‐order derivative interpolation associated with fractional filter functions | |
| US20190005619A1 (en) | Image upscaling system, training method thereof, and image upscaling method | |
| CN109360148B (en) | Remote sensing image super-resolution reconstruction method and device based on mixed random downsampling | |
| CN107392975B (en) | Image self-adaptive multi-scale block compression sensing method and electronic equipment | |
| CN108957448B (en) | Radar correlation imaging method based on generalized total variation regularization | |
| CN204948182U (en) | A kind of image raising frequency system and display unit | |
| CN111738954B (en) | Single-frame turbulence degradation image distortion removal method based on double-layer cavity U-Net model | |
| CN112488931A (en) | Super-resolution enhancement method and system for power transmission line in satellite remote sensing image | |
| Zhang et al. | Super-resolution of large volumes of sentinel-2 images with high performance distributed deep learning | |
| CN113238201B (en) | Super-resolution radar positioning method, system, equipment and storage medium | |
| CN102903104B (en) | Subtractive clustering based rapid image segmentation method | |
| CN114972041B (en) | Polarization radar image super-resolution reconstruction method and device based on residual error network | |
| CN113238201A (en) | Super-resolution radar positioning method, system, equipment and storage medium | |
| Shivagunde et al. | Single image super-resolution based on modified interpolation method using MLP and DWT | |
| Haq et al. | Block-based compressed sensing of MR images using multi-rate deep learning approach | |
| JP2012164147A (en) | Image reduction device, image enlargement device and programs for these | |
| US20220044042A1 (en) | Transferring data between different frequency bands | |
| RU2728949C1 (en) | Method of constructing and processing images and system for implementing thereof | |
| CN112419146B (en) | Image processing method and device and terminal equipment | |
| Li et al. | Polarimetric ISAR super-resolution based on group residual attention network | |
| CN104504654B (en) | High-resolution image reconstruction method based on directivity gradient | |
| Liu et al. | Energy and spectrum efficient radio frequency fingerprint intelligent blind identification | |
| Kumar | A novel wavelet based super resolution reconstruction of low resolution images using adaptive interpolation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |