CN106772379A - Two-parameter CFAR Ship Detections based on Fourier transformation - Google Patents
Two-parameter CFAR Ship Detections based on Fourier transformation Download PDFInfo
- Publication number
- CN106772379A CN106772379A CN201710173725.4A CN201710173725A CN106772379A CN 106772379 A CN106772379 A CN 106772379A CN 201710173725 A CN201710173725 A CN 201710173725A CN 106772379 A CN106772379 A CN 106772379A
- Authority
- CN
- China
- Prior art keywords
- matrix
- clutter
- background window
- imaging data
- fourier transform
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 78
- 230000009466 transformation Effects 0.000 title abstract 2
- 239000011159 matrix material Substances 0.000 claims abstract description 341
- 238000003384 imaging method Methods 0.000 claims abstract description 92
- 238000000034 method Methods 0.000 claims abstract description 32
- 230000014509 gene expression Effects 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 9
- 101100173586 Schizosaccharomyces pombe (strain 972 / ATCC 24843) fft2 gene Proteins 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000005315 distribution function Methods 0.000 claims description 3
- 102000002274 Matrix Metalloproteinases Human genes 0.000 claims 1
- 108010000684 Matrix Metalloproteinases Proteins 0.000 claims 1
- 238000000926 separation method Methods 0.000 abstract description 6
- 238000004422 calculation algorithm Methods 0.000 description 28
- 230000006870 function Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 6
- 230000010354 integration Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000003556 assay Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
- G01S13/9021—SAR image post-processing techniques
- G01S13/9029—SAR image post-processing techniques specially adapted for moving target detection within a single SAR image or within multiple SAR images taken at the same time
-
- 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
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
-
- 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
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
- G01S13/9021—SAR image post-processing techniques
- G01S13/9027—Pattern recognition for feature extraction
-
- 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
- G01S7/415—Identification of targets based on measurements of movement associated with the target
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a kind of two-parameter CFAR Ship Detections based on Fourier transformation, its thinking is：SAR radar return data are obtained, and imaging is carried out to SAR radar returns data, obtain SAR radar imagery data matrix I, and then obtain the bianry image matrix I after threshold processbw；IbwBe be marked with comprising naval vessel classification and pseudo- target bianry image；All 1's matrix is setIt is designated as backdrop window；And calculate the clutter number statistical matrix N in backdrop window successively and take down the clutter number statistical matrix in operation rear backdrop windowAnd then calculate the clutter average statistical matrix M in background window；Then the clutter variance statistic matrix V in calculating background window, and calculating naval vessel class target trip current F；Finally according to I and F, detection obtains multiple Ship Targets, and the ships target that each is detected is respectively Ship Target interested, so far completes the separation of Ship Target interested and pseudo- target, and then completes Ship Target Detection interested.
Description
Technical Field
The invention belongs to the technical field of radar signal processing, and particularly relates to a Fourier transform-based double-parameter CFAR ship detection method, namely a Fourier transform-based double-parameter Constant False Alarm Rate (CFAR) ship detection method, which is a target detection method in a synthetic aperture radar and is suitable for radar moving target detection in the synthetic aperture radar or inverse synthetic aperture radar under a Gaussian background.
Background
In large-scene high-resolution Synthetic Aperture Radar (SAR) image detection, a Constant False Alarm Rate (CFAR) algorithm is the most widely applied algorithm in the SAR image target detection field, and comprises a traditional CFAR algorithm and a double-parameter CFAR detection algorithm; the specific implementation process of the traditional CFAR algorithm is as follows: according to a classical statistical detection theory, under a given false alarm probability condition, firstly, a detection threshold value is obtained in a self-adaptive mode according to the statistical characteristics of the background clutter around the target, then each pixel to be detected in the SAR image is compared with the self-adaptive threshold value respectively, and whether the pixel is the target point or not is judged; the double-parameter CFAR detection algorithm is based on the assumption that background clutter is Gaussian distribution, is suitable for SAR image detection with complex local background change, and slides in the SAR image in a certain step length after a target window, a protection window and a background window are set; the protection window is used for preventing the ship part in the target window from leaking into the background window, and then the detection threshold of the local window is obtained by calculating the mean value and the variance of all pixels in the background window, so that the part, higher than the threshold, in the target window is judged as the ship.
The sizes of a target window, a protection window and a background window of the double-parameter CFAR detection algorithm are respectively obtained through empirical learning, and the sliding step length is also obtained through experience; for ship targets with a short distance, the protection window cannot ensure that the ship part cannot leak into the background window, so that the mean value and the variance of the clutter in the background window cannot represent the actual value of the clutter through statistics, and false alarm and missing detection can occur in the detection process.
The CFAR detection speed is an important index for determining the practicability of a target detection algorithm, the double-parameter CFAR detection algorithm is essentially a local sliding window detection method, each pixel in the SAR image participates in sliding window operation for multiple times respectively due to local dynamics, so that the CFAR algorithm is large in operation amount and low in calculation efficiency, and the target detection speed is low; because the detection face is the whole SAR image scene, the data volume of the operation is large, if the calculation efficiency is low, the performance of the whole system where the SAR image target detection is located is influenced, the real-time performance of the system is reduced, and the requirement of a hardware implementation system is increased.
Disclosure of Invention
The invention aims to provide a double-parameter CFAR ship detection method based on Fourier transform, which can be used for rapidly detecting ship targets in SAR images and has small calculated amount and high detection efficiency.
In order to achieve the technical purpose, the invention is realized by adopting the following technical scheme.
A double-parameter CFAR ship detection method based on Fourier transform comprises the following steps:
step 1, SAR radar echo data are obtained, imaging processing is carried out on the SAR radar echo data, and an SAR radar imaging data matrix I is obtained; then carrying out threshold processing on the SAR radar imaging data matrix I to obtain a binary image matrix I after threshold processingbw(ii) a The above-mentionedBinary image matrix I after threshold processingbwThe method comprises the steps of obtaining a binary image containing a ship target and a pseudo target;
step 2, setting a full 1 matrixRecording as a background window; and processing the binary image matrix I according to the threshold valuebwCalculating to obtain a clutter number statistical matrix N in the background window, and further calculating to obtain a clutter number statistical matrix in the background window after the inverting operation
Step 3, processing the binary image matrix I according to the threshold valuebwAn SAR radar imaging data matrix I and a clutter number statistical matrix in a background window after inverting operationCalculating to obtain a clutter mean statistical matrix M in the background window;
step 4, according to the clutter number statistical matrix in the background window after the inverting operationCalculating to obtain a clutter variance statistical matrix V in the background window;
step 5, calculating to obtain a ship target judgment matrix F according to the clutter mean statistical matrix M in the background window and the clutter variance statistical matrix V in the background window;
and 6, detecting to obtain a plurality of ship targets according to the SAR radar imaging data matrix I and the ship target judgment matrix F.
Compared with the prior art, the method has the following advantages:
the method can improve the detection effect of the ship target, reduce the operation amount of the double-parameter CFAR detection algorithm, improve the operation efficiency and quickly realize the ship target detection in the SAR image under the accurate detection of the ship target.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
FIG. 1 is a flow chart of a Fourier transform-based dual-parameter CFAR ship detection method of the present invention;
figure 2(a) is a schematic diagram of synthetic aperture radar imaging in the C-band,
FIG. 2(b) is a schematic diagram of the results obtained after the detection of FIG. 2(a) using the conventional two-parameter CFAR algorithm;
FIG. 2(c) is a schematic diagram of the result obtained after the detection of FIG. 2(a) is performed by using an improved SAR image dual-parameter CFAR ship detection algorithm;
FIG. 2(d) is a graph showing the results obtained after the assay of FIG. 2(a) is performed using the method of the present invention;
fig. 3 is a comparison graph of time taken for respectively performing target detection on SAR images with different pixel numbers by using a conventional two-parameter CFAR algorithm, an improved two-parameter CFAR algorithm, and the method of the present invention.
Detailed Description
Referring to fig. 1, it is a flow chart of a dual-parameter CFAR ship detection method based on fourier transform of the present invention; the following Fourier transform-based double-parameter CFAR ship detection method comprises the following steps:
step 1, SAR radar echo data are obtained and are subjected to imaging processing to obtain an SAR radar imaging data matrix I, wherein the SAR radar imaging data matrix I comprises a sea surface target and a non-sea surface target; then achieving SAR radarCarrying out threshold processing on the image data matrix I, namely separating a sea surface target and a non-sea surface target in the SAR radar imaging data matrix I, and further obtaining a binary image matrix I after threshold processingbw(ii) a The thresholded binary image matrix IbwThe method is a binary image containing a ship target and a pseudo target.
Specifically, SAR radar imaging data may obey different distribution models, the corresponding various distribution models are different along with the change of the resolution and imaging parameters of SAR radar imaging, Gaussian distribution, negative exponential distribution, lognormal distribution and K distribution can be used for fitting a target and detecting the background clutter of the target, and the Gaussian distribution is adopted for fitting the target and detecting the background clutter of the target in the embodiment of the invention.
Firstly, SAR radar echo data are obtained, imaging processing is carried out on the SAR radar echo data, and an SAR radar imaging data matrix I is obtained; the SAR radar imaging data matrix I comprises a sea surface target and a non-sea surface target; the sea surface target and the non-sea surface target in the SAR radar imaging data matrix I are separated through threshold processing, and a binary image matrix I after threshold processing is obtainedbw(ii) a The threshold processing uses a maximum between-class variance algorithm (namely an Otsu threshold method), which is a very popular threshold selection algorithm at present, and the core idea is as follows: determining an optimal threshold, and after the SAR radar imaging data matrix I is subjected to binarization processing, maximizing the inter-class variance between the background clutter class pixels and the target class pixels of the SAR radar imaging data matrix I to achieve the purpose of distinguishing the background clutter from the target in the SAR radar imaging data matrix I; wherein, the background clutter is a sea surface target, and the target is a non-sea surface target; and further, separation of sea surface targets and non-sea surface targets in the SAR radar imaging data matrix I is achieved.
1a) Setting the gray level of an SAR radar imaging data matrix I as {0: K } distribution, wherein K is 255 in the embodiment; wherein the gray scale isHas a number of pixels ofThe total number of pixels in the SAR radar imaging data matrix I isA gray scale ofHas a probability of occurrence of a pixel ofAnd isFurther calculating to obtain the total gray average value of the SAR radar imaging data matrix I
1b) The gray level in the SAR radar imaging data matrix I is set asIs recorded as Is composed of
K, and dividing the pixel gray level of the SAR radar imaging data matrix I into two types of pixels which are recorded asClass pixel andclass pixels, which are respectively expressed as:whereinThe class pixels represent the background clutter and,the class pixels represent targets, and the targets comprise ship targets and pseudo targets generated by strong sea clutter; k is a positive integer greater than 1.
Then separately calculateProbability of class pixel generationProbability of class pixel generationMean of class pixelsAndmean of class pixelsThe expressions are respectively:
then calculating to obtain the inter-class variance of the SAR radar imaging data matrix I asThe expression is as follows:
wherein,and (4) the average value of the total gray scale of the SAR radar imaging data matrix I.
1c) In order to obtain the maximum inter-class variance of the SAR radar imaging data matrix I, the following optimization problem needs to be carried out:
calculating to obtain a threshold-processed binary image matrix IbwOptimum threshold value T ofopt(ii) a Arg representsThe gray level in the SAR radar imaging data matrix I corresponding to the maximum value isPixel threshold ofMax represents the maximum value operation.
1d) The binary image matrix I processed according to the threshold valuebwOptimum threshold value T ofoptCarrying out threshold processing on the SAR radar imaging data matrix I to obtain a binary image matrix I after threshold processingbw(ii) a The thresholded binary image matrix IbwIs a binary image containing a ship target and a pseudo target generated by strong sea clutter, and the binary image matrix I after threshold processingbwIs m '× n' dimensional matrix, and the binary image matrix I after threshold processingbwThe value of any one pixel point is marked as Ibw(i ', j') expressed as:
the SAR radar imaging data matrix I is an m × n-dimensional matrix, (I, j) is a coordinate of any pixel point in the SAR radar imaging data matrix I, I ∈ {1,2, …, m }, j ∈ {1,2, …, n }, and (I ', j') is a binary image matrix I subjected to threshold processingbwAny one pixel point coordinate, I ' ∈ {1,2, …, m ' }, j ' ∈ {1,2, …, n ' }, m represents the row number of the SAR radar imaging data matrix I, n represents the column number of the SAR radar imaging data matrix I, and m ' represents the binary image matrix I after threshold processingbwN' represents the thresholded binary image matrix IbwThe number of columns; binary image matrix I after threshold processingbwThe number of lines of the binary image matrix I is equal to the value of the number of lines of the SAR radar imaging data matrix I, and the binary image matrix I after threshold processingbwThe column number of the binary image matrix I is equal to the column number of the SAR radar imaging data matrix I in value, and the binary image matrix I is subjected to threshold processingbwThe pixels in the SAR radar imaging data matrix I correspond to the pixels in the SAR radar imaging data matrix I one by one; m, n, m 'and n' are positive integers greater than 0.
Step 2, setting a m '× n' dimension full 1 matrixRecording as a background window; and processing the binary image matrix I according to the threshold valuebwCalculating to obtain a clutter number statistical matrix N in the background window, and further calculating to obtain a clutter number statistical matrix in the background window after the inverting operationWherein m 'is more than or equal to 0 and less than or equal to m', n 'is more than or equal to 0 and less than or equal to n', and m 'and n' are positive integers respectively.
Specifically, after separation of a sea surface target and a non-sea surface target in an SAR radar imaging data matrix I, a binary image containing a ship target and a pseudo target generated by a strong sea clutter is obtained; in order to achieve the separation of ship targets and pseudo targets generated by strong sea clutter, a background window is arranged, the separation is realized through the accurate statistics of the clutter in the background window, and the process is as follows:
setting an m "× n" dimensional full 1 matrixRecording as a background window; performing two-dimensional Fourier transform on the background window to obtain a two-dimensional frequency domain filter matrix H; thresholding binary image matrix IbwPerforming two-dimensional Fourier transform operation to obtain a binary image matrix subjected to the two-dimensional Fourier transform operation; performing dot multiplication on the binary image matrix subjected to the two-dimensional Fourier transform operation and the two-dimensional frequency domain filter matrix H to obtain a dot multiplied matrix S; then, performing two-dimensional inverse Fourier transform on the dot-multiplied matrix S to obtain a clutter number statistical matrix N in the background window; performing inverting operation on each element in the clutter number statistical matrix N in the background window to obtain a clutter number statistical matrix in the background window after inverting operationClutter number statistical matrix in background window after inverting operationThe method is used for counting the mean value and the variance of the clutter in the background window; wherein m 'is more than or equal to 0 and less than or equal to m', n 'is more than or equal to 0 and less than or equal to n', and m 'and n' are positive integers respectively.
2a) Setting an m "× n" dimensional full 1 matrixMarking as a background window, wherein m 'is more than or equal to 0 and less than or equal to m', n 'is more than or equal to 0 and less than or equal to n', and m 'and n' are positive integers respectively; performing two-dimensional Fourier transform on the background window to obtain a two-dimensional frequency domain filter matrix H; thresholding binary image matrix IbwPerforming two-dimensional Fourier transform operation to obtain a binary image matrix subjected to the two-dimensional Fourier transform operation; performing point multiplication on the binary image matrix subjected to the two-dimensional Fourier transform operation and the two-dimensional frequency domain filter matrix H to obtain a moment subjected to the point multiplicationArray S; then, performing two-dimensional inverse Fourier transform on the dot-multiplied matrix S to obtain a clutter number statistical matrix N in the background window, wherein the expression is as follows:
N＝IFFT2(H×FFT2(Ibw))
wherein, the clutter number statistical matrix N in the background window isThe dimension matrix is a matrix of dimensions,the line number of the statistic matrix N representing the clutter number in the background window,representing the column number of the clutter number statistical matrix N in the background window, the row number of the clutter number statistical matrix N in the background window and the binary image matrix I after threshold processingbwThe row number values of the two-value image matrix I are equal, and the column number of the clutter number statistical matrix N in the background window and the threshold value are processedbwThe number of columns is equal; FFT2 (-) is a two-dimensional fourier transform function and IFFT2 (-) is a two-dimensional inverse fourier transform function.
2b) Performing inverting operation on each element in the clutter number statistical matrix N in the background window to obtain a clutter number statistical matrix in the background window after inverting operationClutter number statistical matrix in background window after inverting operationIs composed ofA dimension matrix; statistic matrix for clutter number in background window after inverting operationIs recorded as any one pixel point valueThe expression is as follows:
wherein, the clutter number statistical matrix N in the background window isThe dimension matrix is a matrix of dimensions,the line number of the statistic matrix N representing the clutter number in the background window,the column number of the clutter number statistical matrix N in the background window is represented,any pixel point value in the statistical matrix N of the clutter number in the background window is calculated, statistical matrix for representing clutter number in background window after inverting operationThe number of rows of (a) to (b),statistical matrix for representing clutter number in background window after inverting operationThe number of columns; line number of clutter number statistical matrix N in background window and clutter number statistical matrix in background window after inverting operationThe row number values of the background window are equal, the column number of the clutter number statistical matrix N in the background window and the clutter number statistical matrix in the background window after the inverting operation are carried outThe column number of the background window is equal in value, and the pixel in the clutter number statistical matrix N in the background window and the clutter number statistical matrix in the background window after the inverting operationThe pixels in (1) correspond one to one; respectively positive integers greater than 0.
Step 3, processing the binary image matrix I according to the threshold valuebwAn SAR radar imaging data matrix I and a clutter number statistical matrix in a background window after inverting operationAnd calculating to obtain a clutter mean statistical matrix M in the background window.
Specifically, the thresholded binary image matrix IbwPerforming point multiplication operation with the SAR imaging data matrix I to obtain a first result matrix of the point multiplication operation; performing two-dimensional Fourier transform on the first result matrix of the dot product operation to obtain a first result matrix S' after the two-dimensional Fourier transform; performing dot multiplication operation on the first result matrix S' subjected to two-dimensional Fourier transform and the two-dimensional frequency domain filter matrix H to obtain a second result matrix subjected to dot multiplication operation; second after the multiplication operationPerforming two-dimensional inverse Fourier transform on the result matrix to obtain a second result matrix after the two-dimensional inverse Fourier transform; the second result matrix after two-dimensional inverse Fourier transform and the clutter number statistical matrix in the background window after inverting operation are carried outPerforming dot multiplication to obtain a clutter mean statistical matrix M in the background window, wherein the calculation expression is as follows:
the FFT2 (-) is a two-dimensional fourier transform function, and the IFFT2 (-) is a two-dimensional inverse fourier transform function.
Step 4, according to the clutter number statistical matrix in the background window after the inverting operationAnd calculating to obtain a clutter variance statistical matrix V in the background window.
Specifically, performing a square operation on a first result matrix of the dot product operation to obtain a result matrix after the square operation; performing two-dimensional Fourier transform on the result matrix after the squaring operation to obtain a result matrix S' subjected to the two-dimensional Fourier transform after the squaring operation; performing dot multiplication on the result matrix S' subjected to the two-dimensional Fourier transform after the squaring operation and the two-dimensional frequency domain filter matrix H to obtain a third result matrix subjected to the dot multiplication operation; performing two-dimensional inverse Fourier transform on the third result matrix after the dot multiplication operation to obtain a third result matrix after the two-dimensional inverse Fourier transform; then, a third result matrix after two-dimensional inverse Fourier transform and a clutter number statistical matrix after inverting operation are carried outPerforming dot multiplication operation to obtain a fourth result matrix after the dot multiplication operation; the first after dot product operationSubtracting the square of the clutter mean statistical matrix M in the background window from the four result matrixes to obtain a fifth result matrix after the dot multiplication operation; and finally, performing square root solving operation on the fifth result matrix after the point multiplication operation to obtain a clutter variance statistical matrix V in the background window, wherein the calculation expression is as follows:
the FFT2 (-) is a two-dimensional fourier transform function, and the IFFT2 (-) is a two-dimensional inverse fourier transform function.
And 5, calculating to obtain a ship target judgment matrix F according to the clutter mean statistical matrix M in the background window and the clutter variance statistical matrix V in the background window.
5a) The CFAR method based on Gaussian distribution assumes that clutter in a background window obeys Gaussian distribution, wherein the probability density function of the Gaussian distribution is f (x), and the expression is as follows:
wherein u is a clutter mean value in the background window, sigma is a clutter standard deviation in the background window, exp represents an exponential function, and x is a function variable; and further obtaining a clutter accumulation integral function F (x) in the background window, wherein the expression is as follows:
(t) represents an integration function, t represents an integration variable, the integration interval of the integration variable t is { - ∞, x }, - ∞ is minus infinity, and dt represents the differentiation of t; as variable substitutionThen obtain the backClutter optimization accumulation integral function F' (x) in the scene window is expressed as:
where Φ (·) represents a standard normal distribution function.
Setting false alarm rate Pfa，0≤Pfa1 or less, normal false alarm rate PfaHas a value range of [10 ]-8,10-2](ii) a Then, the target detection threshold T is obtained according to the following formulacfar：
The simplification can be obtained:
and then calculating to obtain a target detection threshold value TcfarThe calculation expression is as follows:
Tcfar＝Φ-1(1-Pfa)×σ+u
wherein phi-1(. cndot.) represents the inverse of the standard normal distribution function.
5b) According to clutter mean statistical matrix M in the background window, clutter variance statistical matrix V in the background window, and target detection threshold TcfarCalculating to obtain a ship target judgment matrix F, wherein the ship target judgment matrix F isA dimension matrix, whose expression is:
F＝Φ-1(1-Pfa)×V+M
and 6, detecting to obtain a plurality of ship targets according to the SAR radar imaging data matrix I and the ship target judgment matrix F.
Specifically, each pixel point value in the SAR radar imaging data matrix I is compared with a corresponding pixel point value in the ship target determination matrix F one by one, and the specific determination process is as follows:
if the pixel point value I (I, j) corresponding to any pixel point coordinate (I, j) in the SAR imaging data matrix I is larger than any pixel point coordinate in the target judgment matrixCorresponding pixel point valueTaking the pixel point coordinates (I, j) corresponding to the pixel point value I (I, j) as the detected ship target pixel; on the contrary, if the pixel point value I (I, j) corresponding to any pixel point coordinate (I, j) in the SAR imaging data matrix I is smaller than any pixel point coordinate in the target judgment matrixCorresponding pixel point valueAnd taking the pixel point coordinate (I, j) corresponding to the pixel point value I (I, j) as a pseudo target pixel generated by the strong sea clutter.
The SAR imaging data matrix I is an m × n-dimensional matrix, (I, j) is any pixel point coordinate in the SAR imaging data matrix I, I (I, j) is a pixel point value corresponding to any pixel point coordinate (I, j) in the SAR imaging data matrix I, I ∈ {1,2, …, m }, j ∈ {1,2, …, n };the coordinates of any one pixel point in the matrix are determined for the target,determining the coordinates of any pixel point in the matrix for the targetThe value of the corresponding pixel point is, the number of rows of the matrix F is determined for the ship-like target,the column number of the ship target judgment matrix F is equal to the row number value of the SAR radar imaging data matrix I, the column number of the ship target judgment matrix F is equal to the column number value of the SAR radar imaging data matrix I, and pixel points in the ship target judgment matrix F correspond to pixel points in the SAR radar imaging data matrix I one by one.
Then, traversing each pixel point in the SAR radar imaging data matrix I and the ship target judgment matrix F respectively, and recording the pixel point coordinates of each ship target in the detected SAR radar imaging data matrix I in sequence to obtain the respective pixels of a plurality of ship targets, wherein the plurality of ship targets are detected at the moment, and each detected ship target is an interested ship target respectively, so that the separation of a sea target and a non-sea target (namely the interested ship target and a pseudo target generated by strong sea clutter) is completed, and the detection of the interested ship target is further completed.
The effect of the present invention is further verified and explained below with the simulation experiment.
Measured data acquisition system parameter
In a ship target detection experiment, the method respectively compares the method with a traditional two-parameter CFAR algorithm and an improved two-parameter CFAR detection method, the simulation adopts the synthetic aperture radar imaging under the stable sea condition to detect the image in the C wave band, and the clutter of the synthetic aperture radar imaging respectively obeys Gaussian distribution; and (5) intercepting the SAR image with 720 x 869 point pixels.
In order to remove the interference of other targets on detection pixels and consider the size of a ship target and the resolution of a scene in an SAR image, the size of a protection window is 21 × 21 and the size of a background data window is 51 × 51 in an experiment, and the false alarm rate Pfa value set in the detection is 10-3。
(II) actually measured data target detection processing content and result
Fig. 2(a) is a schematic diagram of a synthetic aperture radar imaging in a C-band, fig. 2(b) is a schematic diagram of a result obtained after the detection of fig. 2(a) is performed by using a conventional two-parameter CFAR algorithm, fig. 2(C) is a schematic diagram of a result obtained after the detection of fig. 2(a) is performed by using an improved SAR image two-parameter CFAR ship detection algorithm, and fig. 2(d) is a schematic diagram of a result obtained after the detection of fig. 2(a) is performed by using the method of the present invention.
Wherein, no part of the ship leaks into the background window in the protection window of fig. 2(c), and the detection process is realized by using a sliding window; in fig. 2(b) -2 (d), the white square frame is the correctly detected ship target, the black circular frame is the ship target missed in the detection, and the black rectangular frame is the false ship target in the detection.
In order to measure the detection effect, a quality factor FoM of the ship target detection is defined,Nttfor the correct number of vessel targets contained after detection, NfaNumber of false alarm vessels, NgtThe actual number of the ships is the target number; table 1 lists the index comparisons of the synthetic aperture radar imaging on quiet sea using 3 detection methods on the C-band image; synthetic aperture radar imaging in C wave bandThe number of actual ship targets in the image is 18.
TABLE 1
FIG. 3 is a comparison graph of time taken to perform target detection on SAR images with different pixel numbers respectively using a conventional two-parameter CFAR algorithm, an improved two-parameter CFAR algorithm and the method of the present invention; the SAR images respectively comprising 256 × 256, 512 × 512, 1024 × 1024 and 2048 × 2048 pixels use a traditional double-parameter CFAR algorithm, an improved double-parameter CFAR algorithm and the time comparison effect graph used for detecting the ship target by the method of the invention.
(III) analysis of the results of detection
As can be seen from fig. 2(b), fig. 2(c), fig. 2(d), and fig. 3, the method of the present invention compares with the conventional two-parameter algorithm: the method of the invention can not leak detection for ship targets which are very close to each other, because the ship part in the background window is extracted during clutter statistics, no ship part in the protection window leaks into the background window, and the detection effect is greatly improved.
Compared with the improved two-parameter algorithm, the method has the common points that the clutter statistics is carried out by extracting the ship part in the background window, and the method has the same detection effect; however, the method realizes the process of sliding the window in the CFAR algorithm through two-dimensional frequency domain operation, can greatly reduce the time complexity, effectively reduce the operation amount of the CFAR detection algorithm and reduce the detection time; therefore, the method can rapidly realize the detection of the ship target in the SAR image while accurately obtaining the statistical parameters of the distribution model of the background clutter, and has good practicability for the detection of ships with complex sea conditions.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention; thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (7)
1. A double-parameter CFAR ship detection method based on Fourier transform is characterized by comprising the following steps:
step 1, SAR radar echo data are obtained, imaging processing is carried out on the SAR radar echo data, and an SAR radar imaging data matrix I is obtained; then carrying out threshold processing on the SAR radar imaging data matrix I to obtain a binary image matrix I after threshold processingbw(ii) a The thresholded binary image matrix IbwThe method comprises the steps of obtaining a binary image containing a ship target and a pseudo target;
step 2, setting upAll 1 matrixRecording as a background window; and processing the binary image matrix I according to the threshold valuebwCalculating to obtain a clutter number statistical matrix N in the background window, and further calculating to obtain a clutter number statistical matrix in the background window after the inverting operation
Step 3, processing the binary image matrix I according to the threshold valuebwAn SAR radar imaging data matrix I and a clutter number statistical matrix in a background window after inverting operationCalculating to obtain a clutter mean statistical matrix M in the background window;
step 4, according to the clutter number statistical matrix in the background window after the inverting operationCalculating to obtain a clutter variance statistical matrix V in the background window;
step 5, calculating to obtain a ship target judgment matrix F according to the clutter mean statistical matrix M in the background window and the clutter variance statistical matrix V in the background window;
and 6, detecting to obtain a plurality of ship targets according to the SAR radar imaging data matrix I and the ship target judgment matrix F.
2. The Fourier transform-based two-parameter CFAR ship detection method of claim 1, wherein in step 1, the thresholded binary image matrix I is obtainedbwThe method comprises the following substeps:
1a) setting the gray level of an SAR radar imaging data matrix I as {0: K } distribution, wherein the gray level isHas a number of pixels of The total number of pixels in the SAR radar imaging data matrix I is A gray scale ofHas a probability of occurrence of a pixel of And is Further calculating to obtain the total gray average value of the SAR radar imaging data matrix I
1b) The gray level in the SAR radar imaging data matrix I is set asIs recorded as Is any one of {0: K } gray levels, and divides the pixel gray level of the SAR radar imaging data matrix I into two types of pixels, which are marked asClass pixel andclass pixels, which are respectively expressed as:whereinThe class pixels represent the background clutter and,the class pixels represent targets, and the targets comprise ship targets and pseudo targets generated by strong sea clutter; k is a positive integer greater than 1;
then separately calculateProbability of class pixel generation Probability of class pixel generation Mean of class pixelsAndmean of class pixelsThe expressions are respectively:
then calculating to obtain the inter-class variance of the SAR radar imaging data matrix I asThe expression is as follows:
1c) calculating to obtain a threshold-processed binary image matrix IbwOptimum threshold value T ofoptThe expression is as follows:
arg representsThe gray level in the SAR radar imaging data matrix I corresponding to the maximum value isPixel threshold ofMax represents the operation of solving the maximum value;
1d) the binary image matrix I processed according to the threshold valuebwOptimum threshold value T ofoptCarrying out threshold processing on the SAR radar imaging data matrix I to obtain a binary image matrix I after threshold processingbw(ii) a The thresholded binary image matrix IbwIs a binary image containing a ship target and a pseudo target generated by strong sea clutter, and the binary image matrix I after threshold processingbwIs m '× n' dimensional matrix, and the binary image matrix I after threshold processingbwThe value of any one pixel point is marked as Ibw(i ', j') expressed as:
the SAR radar imaging data matrix I is an m × n-dimensional matrix, (I, j) is a coordinate of any pixel point in the SAR radar imaging data matrix I, I ∈ {1,2, …, m }, j ∈ {1,2, …, n }, and (I ', j') is a binary image matrix I subjected to threshold processingbwAny one pixel point coordinate, I ' ∈ {1,2, …, m ' }, j ' ∈ {1,2, …, n ' }, m represents the row number of the SAR radar imaging data matrix I, n represents the column number of the SAR radar imaging data matrix I, and m ' represents the binary image matrix I after threshold processingbwN' represents the thresholded binary image matrix IbwThe number of columns; binary image matrix I after threshold processingbwThe number of lines of the binary image matrix I is equal to the value of the number of lines of the SAR radar imaging data matrix I, and the binary image matrix I after threshold processingbwThe column number of the binary image matrix I is equal to the column number of the SAR radar imaging data matrix I in value, and the binary image matrix I is subjected to threshold processingbwThe pixels in the SAR radar imaging data matrix I correspond to the pixels in the SAR radar imaging data matrix I one by one; m, n, m 'and n' are positive integers greater than 0.
3. The Fourier transform-based two-parameter CFAR ship detection method of claim 2, wherein in step 2, the full 1 matrixIn particular to an m '× n' dimensional all-1 matrix
A clutter number statistical matrix N in the background window and a clutter number statistical matrix in the background window after the inverting operationThe obtaining process comprises the following steps:
2a) setting an m '× n' dimensional all 1 matrixMarking as a background window, wherein m ' is more than or equal to 0 and less than or equal to m ', n ' is more than or equal to 0 and less than or equal to n, and m ' and n ' are respectively positive integers; performing two-dimensional Fourier transform on the background window to obtain a two-dimensional frequency domain filter matrix H; thresholding binary image matrix IbwPerforming two-dimensional Fourier transform operation to obtain a binary image matrix subjected to the two-dimensional Fourier transform operation; performing dot multiplication on the binary image matrix subjected to the two-dimensional Fourier transform operation and the two-dimensional frequency domain filter matrix H to obtain a dot multiplied matrix S; then, performing two-dimensional inverse Fourier transform on the dot-multiplied matrix S to obtain a clutter number statistical matrix N in the background window, wherein the expression is as follows:
N＝IFFT2(H×FFT2(Ibw))
wherein, the clutter number statistical matrix N in the background window isThe dimension matrix is a matrix of dimensions,the line number of the statistic matrix N representing the clutter number in the background window,representing the column number of the clutter number statistical matrix N in the background window, the row number of the clutter number statistical matrix N in the background window and the binary image matrix I after threshold processingbwThe row number values of the two-value image matrix I are equal, and the column number of the clutter number statistical matrix N in the background window and the threshold value are processedbwThe number of columns is equal; FFT2 (-) is a two-dimensional fourier transform function, and IFFT2 (-) is a two-dimensional inverse fourier transform function;
2b) performing inverting operation on each element in the clutter number statistical matrix N in the background window to obtain a clutter number statistical matrix in the background window after inverting operationClutter number statistical matrix in background window after inverting operationIs composed ofA dimension matrix; statistic matrix for clutter number in background window after inverting operationIs recorded as any one pixel point valueThe expression is as follows:
wherein, the clutter number statistical matrix N in the background window isThe dimension matrix is a matrix of dimensions,the line number of the statistic matrix N representing the clutter number in the background window,the column number of the clutter number statistical matrix N in the background window is represented,any pixel point value in the statistical matrix N of the clutter number in the background window is calculated, statistical matrix for representing clutter number in background window after inverting operationThe number of rows of (a) to (b),statistical matrix for representing clutter number in background window after inverting operationThe number of columns; line number of clutter number statistical matrix N in background window and clutter number statistical matrix in background window after inverting operationThe row number values of the background window are equal, the column number of the clutter number statistical matrix N in the background window and the clutter number statistical matrix in the background window after the inverting operation are carried outThe column number of the background window is equal in value, and the pixel in the clutter number statistical matrix N in the background window and the clutter number statistical matrix in the background window after the inverting operationThe pixels in (1) correspond one to one; respectively positive integers greater than 0.
4. The Fourier transform-based two-parameter CFAR ship detection method of claim 3, wherein in step 3, the clutter mean statistical matrix M in the background window is obtained by:
thresholding the binary image matrix IbwPerforming point multiplication operation with the SAR imaging data matrix I to obtain a first result matrix of the point multiplication operation; performing two-dimensional Fourier transform on the first result matrix of the dot product operation to obtain a first result matrix S' after the two-dimensional Fourier transform; performing dot multiplication operation on the first result matrix S' subjected to two-dimensional Fourier transform and the two-dimensional frequency domain filter matrix H to obtain a second result matrix subjected to dot multiplication operation; performing two-dimensional inverse Fourier transform on the second result matrix after the dot product operation to obtain a second result matrix after the two-dimensional inverse Fourier transform; the second result matrix after two-dimensional inverse Fourier transform and the clutter number statistical matrix in the background window after inverting operation are carried outPerforming dot multiplication to obtain a clutter mean statistical matrix M in the background window, wherein the calculation expression is as follows:
the FFT2 (-) is a two-dimensional fourier transform function, and the IFFT2 (-) is a two-dimensional inverse fourier transform function.
5. The Fourier transform-based two-parameter CFAR ship detection method of claim 4, wherein in step 4, the clutter variance statistical matrix V in the background window is obtained by:
performing square operation on the first result matrix of the dot product operation to obtain a result matrix after the square operation; performing two-dimensional Fourier transform on the result matrix after the squaring operation to obtain a result matrix S' subjected to the two-dimensional Fourier transform after the squaring operation; performing dot multiplication on the result matrix S' subjected to the two-dimensional Fourier transform after the squaring operation and the two-dimensional frequency domain filter matrix H to obtain a third result matrix subjected to the dot multiplication operation; performing two-dimensional inverse Fourier transform on the third result matrix after the dot multiplication operation to obtain a third result matrix after the two-dimensional inverse Fourier transform; then, a third result matrix after two-dimensional inverse Fourier transform and a clutter number statistical matrix after inverting operation are carried outPerforming dot multiplication operation to obtain a fourth result matrix after the dot multiplication operation; subtracting the square of the clutter mean statistical matrix M in the background window from the fourth result matrix after the dot multiplication operation to obtain a fifth result matrix after the dot multiplication operation; and finally, performing square root solving operation on the fifth result matrix after the point multiplication operation to obtain a clutter variance statistical matrix V in the background window, wherein the calculation expression is as follows:
the FFT2 (-) is a two-dimensional fourier transform function, and the IFFT2 (-) is a two-dimensional inverse fourier transform function.
6. The Fourier transform-based dual-parameter CFAR ship detection method of claim 5, wherein in step 5, the ship class target decision matrix F has an expression:
F＝Φ-1(1-Pfa)×V+M
wherein phi-1(. represents the inverse of a standard normal distribution function, PfaFor a set false alarm rate, P is greater than or equal to 0fa≤1。
7. The Fourier transform-based dual-parameter CFAR ship detection method of claim 6, wherein in step 6, the detection obtains a plurality of ship targets by the process of:
comparing pixel values of each pixel point value in the SAR radar imaging data matrix I with corresponding pixel point values in the ship target determination matrix F one by one, wherein the specific determination process comprises the following steps:
if the pixel point value I (I, j) corresponding to any pixel point coordinate (I, j) in the SAR imaging data matrix I is larger than any pixel point coordinate in the target judgment matrixCorresponding pixel point valueTaking the pixel point coordinates (I, j) corresponding to the pixel point value I (I, j) as the detected ship target pixel; on the contrary, if the pixel point value I (I, j) corresponding to any pixel point coordinate (I, j) in the SAR imaging data matrix I is smaller than any pixel point coordinate in the target judgment matrixCorresponding pixel point valueThe pixel point coordinate (I, j) corresponding to the pixel point value I (I, j) is determined) As a dummy target pixel;
the SAR radar imaging data matrix I is an m × n-dimensional matrix, I (I, j) is a pixel point value corresponding to any pixel point coordinate (I, j) in the SAR radar imaging data matrix I, I ∈ {1,2, …, m }, and j ∈ {1,2, …, n };the coordinates of any one pixel point in the matrix are determined for the target,determining the coordinates of any pixel point in the matrix for the targetThe value of the corresponding pixel point is, the number of rows of the matrix F is determined for the ship-like target,the column number of a ship target judgment matrix F is determined, the row number of the ship target judgment matrix F is equal to the row number value of an SAR radar imaging data matrix I, the column number of the ship target judgment matrix F is equal to the column number value of the SAR radar imaging data matrix I, and pixel points in the ship target judgment matrix F correspond to pixel points in the SAR radar imaging data matrix I one by one;
then, traversing each pixel point in the SAR radar imaging data matrix I and the ship target judgment matrix F respectively, and recording the coordinates of each ship target pixel point in the detected SAR radar imaging data matrix I in sequence to obtain respective pixels of a plurality of ship targets, wherein the plurality of ship targets are detected at the moment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710173725.4A CN106772379B (en) | 2017-03-22 | 2017-03-22 | Two-parameter CFAR Ship Detection based on Fourier transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710173725.4A CN106772379B (en) | 2017-03-22 | 2017-03-22 | Two-parameter CFAR Ship Detection based on Fourier transformation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106772379A true CN106772379A (en) | 2017-05-31 |
| CN106772379B CN106772379B (en) | 2019-08-02 |
Family
ID=58966489
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710173725.4A Active CN106772379B (en) | 2017-03-22 | 2017-03-22 | Two-parameter CFAR Ship Detection based on Fourier transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106772379B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107942329A (en) * | 2017-11-17 | 2018-04-20 | 西安电子科技大学 | Motor platform single-channel SAR is to surface vessel object detection method |
| CN110992387A (en) * | 2019-11-08 | 2020-04-10 | 浪潮电子信息产业股份有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN111458711A (en) * | 2020-04-24 | 2020-07-28 | 北京卫星信息工程研究所 | Satellite-borne dual-band SAR system and detection method of ship target |
| CN112649791A (en) * | 2020-12-24 | 2021-04-13 | 北京海兰信数据科技股份有限公司 | Radar echo processing method and device |
| CN113256667A (en) * | 2021-04-14 | 2021-08-13 | 大连海事大学 | SAR image ship target detection method based on Otsu and joint distribution |
| CN113674308A (en) * | 2021-05-06 | 2021-11-19 | 西安电子科技大学 | SAR image ship target rapid detection method based on image enhancement and multiple detection |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4062012A (en) * | 1974-07-11 | 1977-12-06 | The United States Of America As Represented By The Secretary Of The Navy | Digital radar signal processor |
| KR100888674B1 (en) * | 2007-07-02 | 2009-03-16 | (주)인펙비전 | Method for measuring similarity using frenquency range |
| CN102768356A (en) * | 2012-08-03 | 2012-11-07 | 南京大学 | Method for detecting ocean stationary targets by multi-temporal radar data |
| CN102938150A (en) * | 2012-11-20 | 2013-02-20 | 西安电子科技大学 | Synthetic aperture radar (SAR) image ship detection method based on self-adaptation sea clutter statistics |
| CN103996166A (en) * | 2014-06-05 | 2014-08-20 | 上海理工大学 | Holographic anti-counterfeiting algorithm based on shape of screen dot |
| CN105974410A (en) * | 2016-04-28 | 2016-09-28 | 西安电子科技大学 | Multi-ship target SAR and ISAR hybrid imaging method of airborne radar |
-
2017
- 2017-03-22 CN CN201710173725.4A patent/CN106772379B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4062012A (en) * | 1974-07-11 | 1977-12-06 | The United States Of America As Represented By The Secretary Of The Navy | Digital radar signal processor |
| KR100888674B1 (en) * | 2007-07-02 | 2009-03-16 | (주)인펙비전 | Method for measuring similarity using frenquency range |
| CN102768356A (en) * | 2012-08-03 | 2012-11-07 | 南京大学 | Method for detecting ocean stationary targets by multi-temporal radar data |
| CN102938150A (en) * | 2012-11-20 | 2013-02-20 | 西安电子科技大学 | Synthetic aperture radar (SAR) image ship detection method based on self-adaptation sea clutter statistics |
| CN103996166A (en) * | 2014-06-05 | 2014-08-20 | 上海理工大学 | Holographic anti-counterfeiting algorithm based on shape of screen dot |
| CN105974410A (en) * | 2016-04-28 | 2016-09-28 | 西安电子科技大学 | Multi-ship target SAR and ISAR hybrid imaging method of airborne radar |
Non-Patent Citations (3)
| Title |
|---|
| XING X 等: ""A fast ship detection algorithm in SAR imagery for wide area ocean surveillance"", 《IEEE RADAR CONFERENCE》 * |
| YANG X L 等: ""CFAR detection of moving range-spread target in white Gaussian noisessing waveform contrast"", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 * |
| 宋文青 等: ""基于幂次变换的SAR图像Otsu分割法"", 《系统工程与电子技术》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107942329A (en) * | 2017-11-17 | 2018-04-20 | 西安电子科技大学 | Motor platform single-channel SAR is to surface vessel object detection method |
| CN107942329B (en) * | 2017-11-17 | 2021-04-06 | 西安电子科技大学 | Method for detecting sea surface ship target by maneuvering platform single-channel SAR |
| CN110992387A (en) * | 2019-11-08 | 2020-04-10 | 浪潮电子信息产业股份有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN111458711A (en) * | 2020-04-24 | 2020-07-28 | 北京卫星信息工程研究所 | Satellite-borne dual-band SAR system and detection method of ship target |
| CN111458711B (en) * | 2020-04-24 | 2022-03-04 | 北京卫星信息工程研究所 | Satellite-borne dual-band SAR system and detection method of ship target |
| CN112649791A (en) * | 2020-12-24 | 2021-04-13 | 北京海兰信数据科技股份有限公司 | Radar echo processing method and device |
| CN113256667A (en) * | 2021-04-14 | 2021-08-13 | 大连海事大学 | SAR image ship target detection method based on Otsu and joint distribution |
| CN113256667B (en) * | 2021-04-14 | 2023-07-21 | 大连海事大学 | SAR image ship target detection method based on Otsu and joint distribution |
| CN113674308A (en) * | 2021-05-06 | 2021-11-19 | 西安电子科技大学 | SAR image ship target rapid detection method based on image enhancement and multiple detection |
| CN113674308B (en) * | 2021-05-06 | 2024-02-13 | 西安电子科技大学 | SAR image ship target rapid detection method based on image enhancement and multiple detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106772379B (en) | 2019-08-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106772379B (en) | Two-parameter CFAR Ship Detection based on Fourier transformation | |
| CN107301661B (en) | High-resolution remote sensing image registration method based on edge point features | |
| CN105354541B (en) | The SAR image object detection method of view-based access control model attention model and constant false alarm rate | |
| CN107229918B (en) | SAR image target detection method based on full convolution neural network | |
| CN112837303A (en) | Defect detection method, device, equipment and medium for mold monitoring | |
| CN102096824B (en) | Multi-spectral image ship detection method based on selective visual attention mechanism | |
| CN107808383B (en) | Rapid detection method for SAR image target under strong sea clutter | |
| CN116664559B (en) | Machine vision-based memory bank damage rapid detection method | |
| CN107038416B (en) | Pedestrian detection method based on binary image improved HOG characteristics | |
| CN112560803A (en) | Radar signal modulation identification method based on time-frequency analysis and machine learning | |
| CN106815819B (en) | More strategy grain worm visible detection methods | |
| CN113256667B (en) | SAR image ship target detection method based on Otsu and joint distribution | |
| CN113674308B (en) | SAR image ship target rapid detection method based on image enhancement and multiple detection | |
| CN109726649B (en) | Remote sensing image cloud detection method and system and electronic equipment | |
| CN108038856B (en) | Infrared small target detection method based on improved multi-scale fractal enhancement | |
| Yang et al. | A curvature-based saliency method for ship detection in SAR images | |
| CN110889843A (en) | SAR image ship target detection method based on maximum stable extremal region | |
| CN111707998B (en) | Sea surface floating small target detection method based on connected region characteristics | |
| CN106842143A (en) | The quick CFAR detection methods of radar target based on filtering | |
| CN107369163A (en) | A kind of quick SAR image object detection method based on best entropy Double Thresholding Segmentation | |
| CN113899349A (en) | Sea wave parameter detection method, equipment and storage medium | |
| CN106254723A (en) | A kind of method of real-time monitoring video noise interference | |
| CN117496109A (en) | Image comparison and analysis method and device, electronic equipment and storage medium | |
| CN107729903A (en) | SAR image object detection method based on area probability statistics and significance analysis | |
| CN111767856A (en) | Infrared small target detection algorithm based on gray value statistical distribution model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |