CN106291494A - The SAR cheating interference target identification method and system strengthened based on differential characteristics - Google Patents

Abstract

The invention discloses the SAR cheating interference target identification method and system strengthened based on differential characteristics, method includes: obtain the blended data of true SAR scene echoes and SAR cheating interference, generates data matrix, generate reference signal after correction process；The data matrix one-dimensional picture of second orientation after different disposal step generates the one-dimensional picture of the first orientation comprising full detail and suppression interference characteristic respectively, obtains differential characteristics image array after processing further；Use rectangular window that differential characteristics image array is processed, generate intensity histogram diagram data and reference gray level rectangular histogram；When the grey level histogram matching distance with reference histograms being detected more than the threshold value pre-set, it is labeled as cheating interference target.The present invention is directed to imaging radar initial data Differential Characteristics under different formation methods strengthen, in SAR image, cheating interference target is identified, improves the imaging radar ability to anti-deceptive interference, reduce cost.

Description

The SAR cheating interference target identification method and system strengthened based on differential characteristics

Technical field

The present invention relates to signal processing technology field, particularly relate to the SAR cheating interference target strengthened based on differential characteristics Identification method and system.

Background technology

SAR (Synthetic Aperture Radar, synthetic aperture radar) deception jammer is according to pre-designed False scene carries out time delay and phase-modulation to the SAR signal intercepted and captured, lower than the power needed for traditional compacting interference, and danger Evil is bigger.Jammer produces false target true to nature in real scene, thus upsets acquisition of information and the decision-making of SAR.Therefore Research for SAR cheating interference causes the extensive concern of scholars, the most correspondingly achieves more significant achievement so that SAR deception jammer can produce more false target more life-like under the mode of operation of Various Complex, more efficiently, Thus add the menace of cheating interference.Therefore, in order to promote the survival ability of SAR system, it is necessary to research accordingly should Cheating interference target is screened by method and identifies.

Zhao Shanshan etc. are at document " Discrimination of Deception Targets in Multistatic Radar Based on Clustering Analysis”(IEEE Sensors Journal,Vol.16, No.8, Apr.2016:2500-2508) in system architecture based on many bases radar, in amplitude proportional feature space use layering The method of classification analysis, is analyzed the dispersion characteristic of cheating interference target with real goal, and is classified number by optimization With the classification that the method designing different classes of minimum cost realizes false target and real goal.Although the method can be one Completing in individual pulse recurrence interval, but need multistatic radar system as support, system complexity is with relatively costly.

Lv Gaohuan etc. are at document " Ground Moving Target Indication in SAR Images With Symmetric Doppler Views”(IEEE Transactions on Geoscience and Remote Sensing, Vol.54, No.1, Jan.2016:533-543) in initial data is divided into Doppler's view of symmetry, according to moving target Moving target is identified by the diversity being distributed in different Doppler's views from static scene echoing characteristics.Owing to deception is dry Disturbing target similar to the doppler characterization of moving target, therefore the method can be applied equally in cheating interference target. But the division of symmetrical Doppler's view can cause the loss of effective imaging bandwidth, thus causes the decline of imaging resolution.

Therefore, prior art has yet to be improved and developed.

Summary of the invention

In view of the deficiencies in the prior art, present invention aim at providing a kind of SAR deception strengthened based on differential characteristics to do Disturb target identification method and system, it is intended to solve cheating interference target recognition resolution in prior art and decline, and cheat dry Disturb the defect that goal systems cost is high.

Technical scheme is as follows:

A kind of SAR cheating interference target identification method strengthened based on differential characteristics, wherein, method includes:

A, obtain the blended data of true SAR scene echoes and SAR cheating interference, after being corrected blended data processing Generate the data matrix of image scene, generate the reference signal of deramp processing according to SAR system parameter；

B, data matrix carry out Fourier transformation with reference signal pointwise after being multiplied, generate the one-dimensional picture of first orientation, data Matrix is filtered after carrying out Fourier transformation, and filter result is carried out inverse Fourier transform, after the interference characteristic that is inhibited Distance cell data, will be multiplied with reference signal pointwise apart from cell data and carry out Fourier transformation, generate suppression interference The one-dimensional picture of second orientation after feature, subtracts each other rear pointwise delivery value by one-dimensional for first orientation picture and the one-dimensional picture of second orientation, generates The one-dimensional picture of differential characteristics, is arranged in order one-dimensional for differential characteristics picture along distance dimension, obtains differential characteristics image array；

C, the rectangular window of employing particular size carry out pointwise process to the data in differential characteristics image array, obtain and count The grey level histogram of the view data in calculation window, generates the intensity histogram diagram data around each point；

D, acquisition grey level are averaged less than the calculated rectangular histogram of pixel of the first threshold pre-set, Generate reference gray level rectangular histogram；

E, calculate the matching distance of the grey level histogram of arbitrary element and reference histograms in data matrix, when detecting When joining distance more than the Second Threshold pre-set, it is cheating interference target by target label corresponding for element.

The described SAR cheating interference target identification method strengthened based on differential characteristics, wherein, described A specifically includes:

A1, obtain the blended data of true SAR scene echoes and SAR cheating interference, carry out Range compress, range migration school After just processing, generate the data matrix of K × L dimension image scene, be designated as s (t_r,t_a), in data matrix, column direction represents distance dimension, Line direction represents azimuth dimension,；t_rFor distance to fast time, t_aFor orientation to the slow time；

A2, according to SAR system parameter generate azimuth dimension deramp processing reference signal, be designated as s₀(t_a)；Particularly as follows:

$s_0\left(t_a\right)=\exp (-{\mathrm{j\π\γ}}_a{t}_a^2)$

Wherein γ_aFor doppler frequency rate.

The described SAR cheating interference target identification method strengthened based on differential characteristics, wherein, described B specifically includes:

B1, acquisition data matrix s (t_r,t_aRow k data s in)_k(t_a), with reference signal s₀(t_a) pointwise is multiplied, right Result after being multiplied carries out Fourier transformation, generates the one-dimensional picture of first orientation comprising real scene with cheating interference full detail

WhereinRepresent Fourier transformation；

B2, acquisition data matrix s (t_r,t_aRow k data s in)_k(t_a), carry out Fourier transformation,

Obtain frequency spectrum dataObtain frequency domain band filter H (f_a)

$H\left(f_a\right)=rect\left(\frac{f_a}{B_a}\right)=\left\{\begin{array}{cc}1,& \left|f_a\right|\≤B_a/2\\ 0,& \left|f_a\right|>B_a/2\end{array}\right.$

Wherein B_aFor the doppler bandwidth of signal, rect () is rectangular window function, according to H (f_a) to frequency domain data S_k (f_a) be filtered, filter result is carried out inverse Fourier transform, generates the kth distance cell data after suppression interference characteristic

WhereinRepresent inverse Fourier transform, by the kth distance cell data after suppression interference characteristic and s₀(t_a) Pointwise is multiplied, and result is carried out Fourier transformation, the one-dimensional picture of second orientation after the interference characteristic that is inhibited

B3, by q_1k(x) and q_2kX () subtracts each other, and to subtracting each other the complex data pointwise delivery value obtained, be calculated differential special Levy one-dimensional picture

Δq_k(x)=norm (q_1k(x)-q_2k(x))

Wherein norm () represents the computing of pointwise modulus value；

B4, make k=1,2 ..., K, repeats step B1～B3, by obtained differential characteristics one-dimensional picture Δ q_kX () is along distance Dimension is arranged in order, and obtains differential characteristics image array Δ I.

The described SAR cheating interference target identification method strengthened based on differential characteristics, wherein, described C specifically includes:

C1, the rectangular window of employing W × W carry out pointwise process to the data in difference image, and wherein W is the length of side point of window Number, uses Δ I_wRepresent the view data in window, calculate its grey level histogram

H=hist (vec (Δ I_w))

Wherein vec () represents matrix Δ I_wThe computing of vectorization, hist () represents the fortune calculating grey level histogram Calculate,For grey level histogram vector, G represents histogrammic quantization level, rectangular window pointwise on Δ I is slided, generates The intensity histogram diagram data of each point.

The described SAR cheating interference target identification method strengthened based on differential characteristics, wherein, described D specifically includes:

In D1, acquisition Δ I, grey level is less than first threshold ε_rThe calculated rectangular histogram of pixel be averaged, To reference gray level rectangular histogram

$h_o=\frac{1}{U}\underset{u=1}{\overset{U}{\Σ}}{h}_u$

Wherein h_uRepresent and be positioned at (k_u,l_u) rectangular histogram of point at place, the grey level in Differential Characteristics figure is less than threshold value ε_r, i.e. Δ I (k_u,l_u) ＜ ε_r, U represent less than this threshold value number a little.

The described SAR cheating interference target identification method strengthened based on differential characteristics, wherein, described E specifically includes:

The grey level histogram h of arbitrary element in E1, calculating data matrix_aMatching distance with reference histograms

d_M(h_o,h_a)=‖ h_o-h_a‖₁

Wherein | | | |₁Represent and ask the computing of vectorial 1 norm, the individual element point in image is calculated Histogram Matching away from From, generate Histogram Matching matrix

E2, when detecting that matching distance is more than the Second Threshold pre-set, by target label corresponding for element For cheating interference target, otherwise it is labeled as real scene.

A kind of SAR cheating interference target identification system strengthened based on differential characteristics, wherein, system includes:

Reference signal generation module, for obtaining the blended data of true SAR scene echoes and SAR cheating interference, to mixed Close the data matrix generating image scene after data are corrected processing, generate the ginseng of deramp processing according to SAR system parameter Examine signal；

Differential characteristics image array generation module, carries out Fourier's change with reference signal pointwise after data matrix is multiplied Changing, generate the one-dimensional picture of first orientation, data matrix is filtered after carrying out Fourier transformation, and is carried out by filter result in inverse Fu Leaf transformation, the distance cell data after the interference characteristic that is inhibited, will be multiplied with reference signal pointwise apart from cell data and go forward side by side Row Fourier transformation, generates the one-dimensional picture of second orientation after suppression interference characteristic, by one-dimensional for first orientation picture and second orientation one Dimension, as subtracting each other rear pointwise delivery value, generates the one-dimensional picture of differential characteristics, is arranged in order along distance dimension by one-dimensional for differential characteristics picture, obtains Differential characteristics image array；

Grey level histogram generation module, for using the rectangular window of particular size to the data in differential characteristics image array Carry out pointwise process, the grey level histogram of the view data in acquisition calculation window, generate the grey level histogram around each point Data；

Reference gray level rectangular histogram generation module, is less than the pixel of the first threshold pre-set for obtaining grey level Calculated rectangular histogram is averaged, and generates reference gray level rectangular histogram；

Mark module, for calculate the grey level histogram of arbitrary element and reference histograms in data matrix mate away from From, when detecting that matching distance is more than the Second Threshold pre-set, it is cheating interference mesh by target label corresponding for element Mark.

The described SAR cheating interference target identification system strengthened based on differential characteristics, wherein, described reference signal generates Module specifically includes:

Correction unit, for obtaining the blended data of true SAR scene echoes and SAR cheating interference, carry out Range compress, After range migration correction processes, generate the data matrix of K × L dimension image scene, be designated as s (t_r,t_a), column direction in data matrix Representing distance dimension, line direction represents azimuth dimension,；t_rFor distance to fast time, t_aFor orientation to the slow time；

Computing unit, for generating the reference signal of azimuth dimension deramp processing according to SAR system parameter, is designated as s₀(t_a)； Particularly as follows:

$s_0\left(t_a\right)=\exp (-{\mathrm{j\π\γ}}_a{t}_a^2)$

Wherein γ_aFor doppler frequency rate.

The described SAR cheating interference target identification system strengthened based on differential characteristics, wherein, described differential characteristics image Matrix generation module specifically includes:

First orientation one-dimensional picture signal generating unit, is used for obtaining data matrix s (t_r,t_aRow k data s in)_k(t_a), with Reference signal s₀(t_a) pointwise is multiplied, and the result after being multiplied carries out Fourier transformation, generate and comprise real scene and cheating interference The one-dimensional picture of first orientation of full detail

WhereinRepresent Fourier transformation；

Second orientation one-dimensional picture signal generating unit, is used for obtaining data matrix s (t_r,t_aRow k data s in)_k(t_a), enter Row Fourier transformation, obtains frequency spectrum dataObtain frequency domain band filter H (f_a)

$H\left(f_a\right)=rect\left(\frac{f_a}{B_a}\right)=\left\{\begin{array}{cc}1,& \left|f_a\right|\≤B_a/2\\ 0,& \left|f_a\right|>B_a/2\end{array}\right.$

Wherein B_aFor the doppler bandwidth of signal, rect () is rectangular window function, according to H (f_a) to frequency domain data S_k (f_a) be filtered, filter result is carried out inverse Fourier transform, generates the kth distance cell data after suppression interference characteristic

WhereinRepresent inverse Fourier transform, by the kth distance cell data after suppression interference characteristic and s₀(t_a) Pointwise is multiplied, and result is carried out Fourier transformation, the one-dimensional picture of second orientation after the interference characteristic that is inhibited

Differential characteristics one-dimensional picture signal generating unit, for by q_1k(x) and q_2kX () subtracts each other, and to subtract each other the complex data that obtains by Point delivery value, is calculated the one-dimensional picture of differential characteristics

Δq_k(x)=norm (q_1k(x)-q_2k(x))

Wherein norm () represents the computing of pointwise modulus value；

Differential image array signal generating unit, is used for making k=1,2 ..., K, the one-dimensional picture of double counting differential characteristics, by gained The differential characteristics one-dimensional picture Δ q arrived_kX () is arranged in order along distance dimension, obtain differential characteristics image array Δ I.

The described SAR cheating interference target identification system strengthened based on differential characteristics, wherein,

Data in difference image are carried out by described grey level histogram generation module specifically for using the rectangular window of W × W Pointwise processes, and wherein W is that the length of side of window is counted, and uses Δ I_wRepresent the view data in window, calculate its grey level histogram

H=hist (vec (Δ I_w))

Wherein vec () represents matrix Δ I_wThe computing of vectorization, hist () represents the fortune calculating grey level histogram Calculate,For grey level histogram vector, G represents histogrammic quantization level, rectangular window pointwise on Δ I is slided, generates The intensity histogram diagram data of each point.

The invention provides a kind of SAR cheating interference target identification method and system strengthened based on differential characteristics, this The bright method strengthened by Differential Characteristics, it is possible to do not improving existing single channel radar system complexity and do not reducing imaging resolution On the premise of, strengthen for imaging radar initial data Differential Characteristics under different formation methods, thus in SAR image In cheating interference target is identified, improve the ability of conventional one-channel broadband imaging radar electronic warfare cheating interference, method is kept away Exempt from the high expense of single channel scalable multi channel system upgrade, greatly reduce SAR system anti-deceptive interference capacity upgrade Cost.

Accompanying drawing explanation

Fig. 1 is the preferable enforcement of a kind of SAR cheating interference target identification method strengthened based on differential characteristics of the present invention The flow chart of example.

Fig. 2 is the concrete application of a kind of SAR cheating interference target identification method strengthened based on differential characteristics of the present invention The real scene schematic diagram of embodiment.

Fig. 3 is the concrete application of a kind of SAR cheating interference target identification method strengthened based on differential characteristics of the present invention The cheating interference scene template schematic diagram of embodiment.

Fig. 4 is the concrete application of a kind of SAR cheating interference target identification method strengthened based on differential characteristics of the present invention The SAR imaging results schematic diagram of the existence cheating interference of embodiment.

Fig. 5 is the concrete application of a kind of SAR cheating interference target identification method strengthened based on differential characteristics of the present invention The cheating interference target identification result schematic diagram of embodiment.

Fig. 6 is the preferable enforcement of a kind of SAR cheating interference target identification system strengthened based on differential characteristics of the present invention The functional schematic block diagram of example.

Detailed description of the invention

For making the purpose of the present invention, technical scheme and effect clearer, clear and definite, below to the present invention the most specifically Bright.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

The invention provides the preferred embodiment of a kind of SAR cheating interference target identification method strengthened based on differential characteristics Flow chart, as it is shown in figure 1, wherein, method includes:

Step S100, obtain the blended data of true SAR scene echoes and SAR cheating interference, blended data is carried out school Generate the data matrix of image scene after just processing, generate the reference signal of deramp processing according to SAR system parameter.

When being embodied as, step S100 specifically includes:

Step S101, obtain the blended data of true SAR scene echoes and SAR cheating interference, carry out Range compress, distance After migration correction process, generate the data matrix of K × L dimension image scene, be designated as s (t_r,t_a), in data matrix, column direction represents Distance dimension, line direction represents azimuth dimension,；t_rFor distance to fast time, t_aFor orientation to the slow time；

Step S102, according to SAR system parameter generate azimuth dimension deramp processing reference signal, be designated as s₀(t_a)；Specifically For:

$\exp (-{\mathrm{j\π\γ}}_a{t}_a^2)$

Wherein γ_aFor doppler frequency rate.

Wherein, during range migration refers to synthetic aperture, the oblique distance between radar and target varied more than one away from From resolution cell so that from the echo-signal of same target in distance in being distributed in different distance unit, cause letter Number in orientation to distance to coupling.As it was previously stated, the two dimension of imaging processing is moved change process becoming two one-dimensional shiftings not Change process, need to carry out range migration correction to eliminate distance to orientation to coupling.So-called range migration correction, it is simply that Be corrected to range migration curvilinear path to be parallel to orientation to straight line, its precision to reach a synthetic aperture time In, the change of oblique distance is less than the half of Range resolution unit.In Space-borne SAR Imaging, echo-signal generally entails big distance Migration, thus range migration correction becomes the important step in imaging processing, directly affects the design of imaging algorithm with final Image quality.

Step S200, data matrix carry out Fourier transformation with reference signal pointwise after being multiplied, generate first orientation one-dimensional Picture, data matrix is filtered after carrying out Fourier transformation, and filter result is carried out inverse Fourier transform, and be inhibited interference Distance cell data after feature, will be multiplied with reference signal pointwise apart from cell data and carry out Fourier transformation, and generation presses down The one-dimensional picture of second orientation after interference characteristic processed, subtracts each other rear pointwise delivery by one-dimensional for first orientation picture and the one-dimensional picture of second orientation Value, generates the one-dimensional picture of differential characteristics, is arranged in order along distance dimension by one-dimensional for differential characteristics picture, obtains differential characteristics image array.

When being embodied as, step S200 specifically includes:

Step S201, acquisition data matrix s (t_r,t_aRow k data s in)_k(t_a), with reference signal s₀(t_a) pointwise phase Take advantage of, the result after being multiplied is carried out Fourier transformation, generate the first orientation comprising real scene with cheating interference full detail One-dimensional picture

WhereinRepresent Fourier transformation；

Step S202, acquisition data matrix s (t_r,t_aRow k data s in)_k(t_a), carry out Fourier transformation, obtain frequency Modal dataObtain frequency domain band filter H (f_a)

$H\left(f_a\right)=rect\left(\frac{f_a}{B_a}\right)=\left\{\begin{array}{cc}1,& \left|f_a\right|\≤B_a/2\\ 0,& \left|f_a\right|>B_a/2\end{array}\right.$

Wherein B_aFor the doppler bandwidth of signal, rect () is rectangular window function, according to H (f_a) to frequency domain data S_k (f_a) be filtered, filter result is carried out inverse Fourier transform, generates the kth distance cell data after suppression interference characteristic

WhereinRepresent inverse Fourier transform, by the kth distance cell data after suppression interference characteristic and s₀(t_a) Pointwise is multiplied, and result is carried out Fourier transformation, the one-dimensional picture of second orientation after the interference characteristic that is inhibited

Step S203, by q_1k(x) and q_2kX () subtracts each other, and to subtracting each other the complex data pointwise delivery value obtained, be calculated The one-dimensional picture of differential characteristics

Δq_k(x)=norm (q_1k(x)-q_2k(x))

Wherein norm () represents the computing of pointwise modulus value；

Step S204, make k=1,2 ..., K, repeats step S201～S203, by obtained differential characteristics one-dimensional picture Δ q_kX () is arranged in order along distance dimension, obtain differential characteristics image array Δ I.

Step S300, utilize the rectangular window of particular size that the data in differential characteristics image array are carried out pointwise process, The grey level histogram of the view data in acquisition calculation window, generates the intensity histogram diagram data around each point.

When being embodied as, step S300 specifically includes:

Step S301, utilizing the rectangular window of W × W that the data in difference image carry out pointwise process, wherein W is the limit of window Length is counted.As a example by the pixel of the row k l row in image, it is assumed that the length of side of window is counted W=3, uses Δ I_wRepresent the figure in window As data, then have

${\mathrm{\ΔI}}_w=\left[\begin{array}{ccc}{\mathrm{\ΔI}}_{k-1,l-1}& {\mathrm{\ΔI}}_{k-1,l}& {\mathrm{\ΔI}}_{k-1,l+1}\\ {\mathrm{\ΔI}}_{k,l-1}& {\mathrm{\ΔI}}_{k,l}& {\mathrm{\ΔI}}_{k,l+1}\\ {\mathrm{\ΔI}}_{k+1,l-1}& {\mathrm{\ΔI}}_{k+1,l}& {\mathrm{\ΔI}}_{k+1,l+1}\end{array}\right]$

Wherein, Δ I_k,lRepresent the data at the row k l row of differential characteristics image array Δ I.To Δ I_wCalculate its gray scale Rectangular histogram

H=hist (vec (Δ I_w))

Wherein vec () represents matrix Δ I_wThe computing of vectorization, hist () represents the fortune calculating grey level histogram Calculate,For grey level histogram vector, G represents histogrammic quantization level, rectangular window pointwise on Δ I is slided, generates The intensity histogram diagram data of each point.

Step S400, acquisition grey level enter less than the calculated rectangular histogram of pixel of the first threshold pre-set Row is average, generates reference gray level rectangular histogram.

When being embodied as, step S400 specifically includes:

In step S401, acquisition Δ I, grey level is less than first threshold ε_rThe calculated rectangular histogram of pixel carry out Averagely, reference gray level rectangular histogram is obtained

$h_o=\frac{1}{U}\underset{u=1}{\overset{U}{\Σ}}{h}_u$

First threshold ε is set_rMax{ Δ I} equal to 5%, the maximum of all elements during wherein max{ Δ I} represents Δ I Value.Wherein h_uRepresent and be positioned at (k_u,l_u) rectangular histogram of point at place, the grey level in Differential Characteristics figure is less than threshold epsilon_r, i.e. Δ I(k_u,l_u) ＜ ε_r, U represent less than this threshold value number a little.

Step S500, calculate the matching distance of the grey level histogram of arbitrary element and reference histograms in data matrix.When Detect when matching distance is more than the Second Threshold pre-set, be cheating interference target by target label corresponding for element.With In differential map picture the spacing of position of orientation and jammer less than the grey level histogram of the element of 5% length of synthetic aperture mate away from From meansigma methods as Second Threshold,

When being embodied as, step S500 specifically includes:

Step S501, calculate the matching distance of the grey level histogram ha of arbitrary element and reference histograms in data matrix

d_M(h_o,h_a)=‖ h_o-h_a‖₁

Wherein | | | |₁Represent and ask the computing of vectorial 1 norm, the individual element point in image is calculated Histogram Matching away from From, generate Histogram Matching matrix

Step S502, when detecting that matching distance is more than the Second Threshold pre-set, by mesh corresponding for element It is labeled as cheating interference target, is otherwise labeled as real scene.

Further, the present invention uses MATLAB software to emulate, and the parameter of emulation data is as follows: Fig. 2 show not Having the real scene SAR imaging results of cheating interference, Fig. 3 show cheating interference target, devises many groups in cheating interference Vehicle target (is respectively designated as " I " " III ") so that it is thus generation interlaced with the real vehicles target of depletion region mixes Confuse.Landform in real scene is then covered and has been cheated by false target " II " " IV ", thus to real scene and mesh Mark is protected.Fig. 4 show the result of cheating interference, and real goal is interlaced with false target, it is difficult to distinguishes, reaches The purpose of cheating interference.Fig. 5 show the result utilizing the inventive method to be identified cheating interference, effectively marks The information such as false scene and the shape of target, position, although in the range of the length of synthetic aperture 5% centered by jammer Occur in that the situation of indicating failure, but by this region outside identification information, still can be to the false building mesh in this region Mark is identified.

Present invention also offers the preferable enforcement of a kind of SAR cheating interference target identification system strengthened based on differential characteristics The functional schematic block diagram of example, as shown in Figure 6, system includes:

Reference signal generation module 100, for obtaining the blended data of true SAR scene echoes and SAR cheating interference, right Blended data generates the data matrix of image scene after being corrected processing, generate deramp processing according to SAR system parameter Reference signal；Specifically as described in embodiment of the method.

Differential characteristics image array generation module 200, is carried out in Fu after data matrix is multiplied with reference signal pointwise Leaf transformation, generates the one-dimensional picture of first orientation, and data matrix is filtered after carrying out Fourier transformation, and carries out inverse by filter result Fourier transformation, the distance cell data after the interference characteristic that is inhibited, will be multiplied with reference signal pointwise apart from cell data And carry out Fourier transformation, generate the one-dimensional picture of second orientation after suppression interference characteristic, by one-dimensional for first orientation picture and second party The one-dimensional picture in position subtracts each other rear pointwise delivery value, generates the one-dimensional picture of differential characteristics, is arranged in order along distance dimension by one-dimensional for differential characteristics picture, Obtain differential characteristics image array；Specifically as described in embodiment of the method.

Grey level histogram generation module 300, for using the rectangular window of particular size in differential characteristics image array Data carry out pointwise process, the grey level histogram of the view data in acquisition calculation window, generate the gray scale around each point straight Side's diagram data；Specifically as described in embodiment of the method.

Reference gray level rectangular histogram generation module 400, is less than the picture of the first threshold pre-set for obtaining grey level The calculated rectangular histogram of vegetarian refreshments is averaged, and generates reference gray level rectangular histogram；Specifically as described in embodiment of the method.

Mark module 500, for calculating mating of the grey level histogram of arbitrary element and reference histograms in data matrix Distance, when detecting that matching distance is more than the Second Threshold pre-set, is cheating interference by target label corresponding for element Target；Specifically as described in embodiment of the method.

The described SAR cheating interference target identification system strengthened based on differential characteristics, wherein, described reference signal generates Module specifically includes:

Correction unit, for obtaining the blended data of true SAR scene echoes and SAR cheating interference, carry out Range compress, After range migration correction processes, generate the data matrix of K × L dimension image scene, be designated as s (t_r,t_a), column direction in data matrix Representing distance dimension, line direction represents azimuth dimension,；t_rFor distance to fast time, t_aFor orientation to the slow time；Specifically implement such as method Described in example.

Computing unit, for generating the reference signal of azimuth dimension deramp processing according to SAR system parameter, is designated as s₀(t_a)； Particularly as follows:

$s_0\left(t_a\right)=\exp (-{\mathrm{j\π\γ}}_a{t}_a^2)$

Wherein γ_aFor doppler frequency rate；Specifically as described in embodiment of the method.

The described SAR cheating interference target identification system strengthened based on differential characteristics, wherein, described differential characteristics image Matrix generation module specifically includes:

First orientation one-dimensional picture signal generating unit, is used for obtaining data matrix s (t_r,t_aRow k data s in)_k(t_a), with Reference signal s₀(t_a) pointwise is multiplied, and the result after being multiplied carries out Fourier transformation, generate and comprise real scene and cheating interference The one-dimensional picture of first orientation of full detail

WhereinRepresent Fourier transformation；Specifically as described in embodiment of the method.

Second orientation one-dimensional picture signal generating unit, is used for obtaining data matrix s (t_r,t_aRow k data s in)_k(t_a), enter Row Fourier transformation, obtains frequency spectrum dataObtain frequency domain band filter H (f_a)

$H\left(f_a\right)=rect\left(\frac{f_a}{B_a}\right)=\left\{\begin{array}{cc}1,& \left|f_a\right|\≤B_a/2\\ 0,& \left|f_a\right|>B_a/2\end{array}\right.$

Wherein B_aFor the doppler bandwidth of signal, rect () is rectangular window function, according to H (f_a) to frequency domain data S_k (f_a) be filtered, filter result is carried out inverse Fourier transform, generates the kth distance cell data after suppression interference characteristic

WhereinRepresent inverse Fourier transform, by the kth distance cell data after suppression interference characteristic and s₀(t_a) Pointwise is multiplied, and result is carried out Fourier transformation, the one-dimensional picture of second orientation after the interference characteristic that is inhibited

Specifically as described in embodiment of the method.

Differential characteristics one-dimensional picture signal generating unit, for by q_1k(x) and q_2kX () subtracts each other, and to subtract each other the complex data that obtains by Point delivery value, is calculated the one-dimensional picture of differential characteristics

Δq_k(x)=norm (q_1k(x)-q_2k(x))

Wherein norm () represents the computing of pointwise modulus value；Specifically as described in embodiment of the method.

Differential image array signal generating unit, is used for making k=1,2 ..., K, the one-dimensional picture of double counting differential characteristics, by gained The differential characteristics one-dimensional picture Δ q arrived_kX () is arranged in order along distance dimension, obtain differential characteristics image array Δ I；Concrete such as method Described in embodiment.

The described SAR cheating interference target identification system strengthened based on differential characteristics, wherein,

Data in difference image are carried out by described grey level histogram generation module specifically for using the rectangular window of W × W Pointwise processes, and wherein W is that the length of side of window is counted, and uses Δ I_wRepresent the view data in window, calculate its grey level histogram

H=hist (vec (Δ I_w))

Wherein vec () represents matrix Δ I_wThe computing of vectorization, hist () represents the fortune calculating grey level histogram Calculate,For grey level histogram vector, G represents histogrammic quantization level, rectangular window pointwise on Δ I is slided, generates The intensity histogram diagram data of each point；Specifically as described in embodiment of the method.

Described reference gray level rectangular histogram generation module is less than first threshold ε specifically for obtaining grey level in Δ I_rPicture The calculated rectangular histogram of vegetarian refreshments is averaged, and obtains reference gray level rectangular histogram

$h_o=\frac{1}{U}\underset{u=1}{\overset{U}{\Σ}}{h}_u$

Wherein h_uRepresent and be positioned at (k_u,l_u) rectangular histogram of point at place, the grey level in Differential Characteristics figure is less than threshold value ε_r, i.e. Δ I (k_u,l_u) ＜ ε_r, U represent less than this threshold value number a little；Specifically as described in embodiment of the method.

Described mark module specifically includes:

Matching distance computing unit, for calculating the grey level histogram h of arbitrary element in data matrix_aWith reference histograms Matching distance

d_M(h_o,h_a)=‖ h_o-h_a‖₁

Wherein | | | |₁Represent and ask the computing of vectorial 1 norm, the individual element point in image is calculated Histogram Matching away from From, generate Histogram Matching matrixSpecifically as described in embodiment of the method.

Indexing unit is for when detecting that matching distance is more than the Second Threshold pre-set, corresponding by element Target label be cheating interference target, be otherwise labeled as real scene；Specifically as described in embodiment of the method.

Unit transferred by file, if for comprising timestamp and the keyword of user's input, performing according to the log of user's input Corresponding operation, transfers the journal file after operation；Specifically as described in embodiment of the method.

In sum, the invention provides a kind of based on differential characteristics strengthen SAR cheating interference target identification method and System, method includes: obtain the blended data of true SAR scene echoes and SAR cheating interference, generates data square after correction process Battle array, generates reference signal；Data matrix generates the one-dimensional picture of the first orientation comprising full detail respectively through different disposal step The one-dimensional picture of second orientation with after suppression interference characteristic, obtains differential characteristics image array after processing further；Use rectangular window Differential characteristics image array is processed, generates intensity histogram diagram data and reference gray level rectangular histogram；When detecting that gray scale is straight When the matching distance of side's figure and reference histograms is more than the threshold value pre-set, it is labeled as cheating interference target.The present invention is directed to Imaging radar initial data Differential Characteristics under different formation methods strengthens, to cheating interference target in SAR image It is identified, improves the imaging radar ability to anti-deceptive interference, reduce cost.

It should be appreciated that the application of the present invention is not limited to above-mentioned citing, for those of ordinary skills, can To be improved according to the above description or to convert, all these modifications and variations all should belong to the guarantor of claims of the present invention Protect scope.

Claims (10)

1. the SAR cheating interference target identification method strengthened based on differential characteristics, it is characterised in that described method includes:

A, obtain the blended data of true SAR scene echoes and SAR cheating interference, generate after being corrected blended data processing The data matrix of image scene, generates the reference signal of deramp processing according to SAR system parameter；

B, data matrix carry out Fourier transformation with reference signal pointwise after being multiplied, generate the one-dimensional picture of first orientation, data matrix Be filtered after carrying out Fourier transformation, and filter result carried out inverse Fourier transform, after the interference characteristic that is inhibited away from From cell data, will be multiplied with reference signal pointwise apart from cell data and carry out Fourier transformation, generate suppression interference characteristic After the one-dimensional picture of second orientation, one-dimensional for first orientation picture and the one-dimensional picture of second orientation are subtracted each other rear pointwise delivery value, generate differential The one-dimensional picture of feature, is arranged in order one-dimensional for differential characteristics picture along distance dimension, obtains differential characteristics image array；

C, the rectangular window of employing particular size carry out pointwise process to the data in differential characteristics image array, obtain and calculate window The grey level histogram of the view data in Kou, generates the intensity histogram diagram data around each point；

D, acquisition grey level are averaged less than the calculated rectangular histogram of pixel of the first threshold pre-set, and generate Reference gray level rectangular histogram；

E, calculate the matching distance of the grey level histogram of arbitrary element and reference histograms in data matrix, when detect coupling away from From during more than the Second Threshold pre-set, it is cheating interference target by target label corresponding for element.

The SAR cheating interference target identification method strengthened based on differential characteristics the most according to claim 1, its feature exists In, described A specifically includes:

A1, obtain the blended data of true SAR scene echoes and SAR cheating interference, carry out at Range compress, range migration correction After reason, generate the data matrix of K × L dimension image scene, be designated as s (t_r,t_a), in data matrix, column direction represents distance dimension, row side To representing azimuth dimension,；t_rFor distance to fast time, t_aFor orientation to the slow time；

A2, according to SAR system parameter generate azimuth dimension deramp processing reference signal, be designated as s₀(t_a)；Particularly as follows:

$s_0\left(t_a\right)=\exp (-{\mathrm{j\π\γ}}_a{t}_a^2)$

Wherein γ_aFor doppler frequency rate.

The SAR cheating interference target identification method strengthened based on differential characteristics the most according to claim 2, its feature exists In, described B specifically includes:

B1, acquisition data matrix s (t_r,t_aRow k data s in)_k(t_a), with reference signal s₀(t_a) pointwise is multiplied, to being multiplied After result carry out Fourier transformation, generate the one-dimensional picture of first orientation comprising real scene with cheating interference full detail

WhereinRepresent Fourier transformation；

B2, acquisition data matrix s (t_r,t_aRow k data s in)_k(t_a), carry out Fourier transformation, obtain frequency spectrum dataObtain frequency domain band filter H (f_a)

$H\left(f_a\right)=rect\left(\frac{f_a}{B_a}\right)=\left\{\begin{array}{cc}1,& \left|f_a\right|\≤B_a/2\\ 0,& \left|f_a\right|>B_a/2\end{array}\right.$

Wherein B_aFor the doppler bandwidth of signal, rect () is rectangular window function, according to H (f_a) to frequency domain data S_k(f_a) carry out Filtering, carries out inverse Fourier transform by filter result, generates the kth distance cell data after suppression interference characteristic

WhereinRepresent inverse Fourier transform, by the kth distance cell data after suppression interference characteristic and s₀(t_a) pointwise It is multiplied, and result is carried out Fourier transformation, the one-dimensional picture of second orientation after the interference characteristic that is inhibited

B3, by q_1k(x) and q_2kX () subtracts each other, and to subtracting each other the complex data pointwise delivery value obtained, be calculated differential characteristics one-dimensional Picture

Δq_k(x)=norm (q_1k(x)-q_2k(x))

Wherein norm () represents the computing of pointwise modulus value；

B4, make k=1,2 ..., K, repeats step B1～B3, by obtained differential characteristics one-dimensional picture Δ q_kX () depends on along distance dimension Secondary arrangement, obtains differential characteristics image array Δ I.

The SAR cheating interference target identification method strengthened based on differential characteristics the most according to claim 3, its feature exists In, described C specifically includes:

C1, using the rectangular window of W × W that the data in difference image carry out pointwise process, wherein W is that the length of side of window is counted, and uses ΔI_wRepresent the view data in window, calculate its grey level histogram

H=hist (vec (Δ I_w))

Wherein vec () represents matrix Δ I_wThe computing of vectorization, hist () represents the computing calculating grey level histogram,For grey level histogram vector, G represents histogrammic quantization level, rectangular window pointwise on Δ I is slided, generates each point Intensity histogram diagram data.

The SAR cheating interference target identification method strengthened based on differential characteristics the most according to claim 4, its feature exists In, described D specifically includes:

In D1, acquisition Δ I, grey level is less than first threshold ε_rThe calculated rectangular histogram of pixel be averaged, joined Examine grey level histogram

$h_o=\frac{1}{U}\underset{u=1}{\overset{U}{\Σ}}{h}_u$

Wherein h_uRepresent and be positioned at (k_u,l_u) rectangular histogram of point at place, the grey level in Differential Characteristics figure is less than threshold epsilon_r, i.e. Δ I(k_u,l_u) ＜ ε_r, U represent less than this threshold value number a little.

The SAR cheating interference target identification method strengthened based on differential characteristics the most according to claim 5, its feature exists In, described E specifically includes:

The grey level histogram h of arbitrary element in E1, calculating data matrix_aMatching distance with reference histograms

d_M(h_o,h_a)=‖ h_o-h_a‖₁

Wherein | | | |₁Represent the computing seeking vectorial 1 norm, the individual element point in image is calculated Histogram Matching distance, raw Become Histogram Matching matrix

E2, when detecting that matching distance is more than the Second Threshold pre-set, by target label corresponding for element for taking advantage of Deceive jamming target, be otherwise labeled as real scene.

7. the SAR cheating interference target identification system strengthened based on differential characteristics, it is characterised in that system includes:

Reference signal generation module, for obtaining the blended data of true SAR scene echoes and SAR cheating interference, to mixed number According to being corrected generating after processing the data matrix of image scene, generate the reference letter of deramp processing according to SAR system parameter Number；

Differential characteristics image array generation module, carries out Fourier transformation with reference signal pointwise after data matrix is multiplied, Generating the one-dimensional picture of first orientation, data matrix is filtered after carrying out Fourier transformation, and filter result carries out inverse Fourier Conversion, the distance cell data after the interference characteristic that is inhibited, will be multiplied with reference signal pointwise apart from cell data and carry out Fourier transformation, generates the one-dimensional picture of second orientation after suppression interference characteristic, by one-dimensional with second orientation for one-dimensional for first orientation picture As subtracting each other rear pointwise delivery value, generate the one-dimensional picture of differential characteristics, one-dimensional for differential characteristics picture is arranged in order along distance dimension, obtains micro- Dtex levies image array；

Grey level histogram generation module, for using the rectangular window of particular size to carry out the data in differential characteristics image array Pointwise processes, the grey level histogram of the view data in acquisition calculation window, generates the intensity histogram diagram data around each point；

Reference gray level rectangular histogram generation module, calculates less than the pixel of the first threshold pre-set for obtaining grey level The rectangular histogram obtained is averaged, and generates reference gray level rectangular histogram；

Mark module, for calculating the grey level histogram of arbitrary element and the matching distance of reference histograms in data matrix, when Detect when matching distance is more than the Second Threshold pre-set, be cheating interference target by target label corresponding for element.

The SAR cheating interference target identification system strengthened based on differential characteristics the most according to claim 7, its feature exists In, described reference signal generation module specifically includes:

Correction unit, for obtaining the blended data of true SAR scene echoes and SAR cheating interference, carries out Range compress, distance After migration correction process, generate the data matrix of K × L dimension image scene, be designated as s (t_r,t_a), in data matrix, column direction represents Distance dimension, line direction represents azimuth dimension,；t_rFor distance to fast time, t_aFor orientation to the slow time；

Computing unit, for generating the reference signal of azimuth dimension deramp processing according to SAR system parameter, is designated as s₀(t_a)；Specifically For:

$s_0\left(t_a\right)=\exp (-{\mathrm{j\π\γ}}_a{t}_a^2)$

Wherein γ_aFor doppler frequency rate.

The SAR cheating interference target identification system strengthened based on differential characteristics the most according to claim 8, its feature exists In, described differential characteristics image array generation module specifically includes:

First orientation one-dimensional picture signal generating unit, is used for obtaining data matrix s (t_r,t_aRow k data s in)_k(t_a), with reference letter Number s₀(t_a) pointwise is multiplied, and the result after being multiplied carries out Fourier transformation, generate and comprise real scene and all believe with cheating interference The one-dimensional picture of first orientation of breath

WhereinRepresent Fourier transformation；

Second orientation one-dimensional picture signal generating unit, is used for obtaining data matrix s (t_r,t_aRow k data s in)_k(t_a), carry out in Fu Leaf transformation, obtains frequency spectrum dataObtain frequency domain band filter H (f_a)

$H\left(f_a\right)=rect\left(\frac{f_a}{B_a}\right)=\left\{\begin{array}{cc}1,& \left|f_a\right|\≤B_a/2\\ 0,& \left|f_a\right|>B_a/2\end{array}\right.$

Wherein B_aFor the doppler bandwidth of signal, rect () is rectangular window function, according to H (f_a) to frequency domain data S_k(f_a) carry out Filtering, carries out inverse Fourier transform by filter result, generates the kth distance cell data after suppression interference characteristic

WhereinRepresent inverse Fourier transform, by the kth distance cell data after suppression interference characteristic and s₀(t_a) pointwise It is multiplied, and result is carried out Fourier transformation, the one-dimensional picture of second orientation after the interference characteristic that is inhibited

Differential characteristics one-dimensional picture signal generating unit, for by q_1k(x) and q_2kX () subtracts each other, and take subtracting each other the complex data pointwise obtained Modulus value, is calculated the one-dimensional picture of differential characteristics

Δq_k(x)=norm (q_1k(x)-q_2k(x))

Wherein norm () represents the computing of pointwise modulus value；

Differential image array signal generating unit, is used for making k=1,2 ..., K, the one-dimensional picture of double counting differential characteristics, by obtained Differential characteristics one-dimensional picture Δ q_kX () is arranged in order along distance dimension, obtain differential characteristics image array Δ I.

The SAR cheating interference target identification system strengthened based on differential characteristics the most according to claim 9, its feature exists In,

Described grey level histogram generation module carries out pointwise specifically for using the rectangular window of W × W to the data in difference image Processing, wherein W is that the length of side of window is counted, and uses Δ I_wRepresent the view data in window, calculate its grey level histogram

H=hist (vec (Δ I_w))

Wherein vec () represents matrix Δ I_wThe computing of vectorization, hist () represents the computing calculating grey level histogram,For grey level histogram vector, G represents histogrammic quantization level, rectangular window pointwise on Δ I is slided, generates each point Intensity histogram diagram data.