KR101834063B1 - Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof - Google Patents

Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof Download PDF

Info

Publication number
KR101834063B1
KR101834063B1 KR1020160019028A KR20160019028A KR101834063B1 KR 101834063 B1 KR101834063 B1 KR 101834063B1 KR 1020160019028 A KR1020160019028 A KR 1020160019028A KR 20160019028 A KR20160019028 A KR 20160019028A KR 101834063 B1 KR101834063 B1 KR 101834063B1
Authority
KR
South Korea
Prior art keywords
inverse
radar image
range
opening surface
denotes
Prior art date
Application number
KR1020160019028A
Other languages
Korean (ko)
Other versions
KR20170097373A (en
Inventor
김경태
강민석
Original Assignee
포항공과대학교 산학협력단
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 포항공과대학교 산학협력단 filed Critical 포항공과대학교 산학협력단
Priority to KR1020160019028A priority Critical patent/KR101834063B1/en
Publication of KR20170097373A publication Critical patent/KR20170097373A/en
Application granted granted Critical
Publication of KR101834063B1 publication Critical patent/KR101834063B1/en

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • 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
    • G01S13/9094—Theoretical aspects

Landscapes

  • Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The present invention relates to an apparatus and method for scaling an inverse synthetic aperture radar image using a principal component analysis technique. According to the present invention, a method for scaling the transverse distance direction of an inverse synthesized opening surface radar image includes a first de-compositing opening surface radar image at a first time point and a second inverse composite point at a second time point later than the first time point Extracting scattered points representing the shape of the object from the first inverse synthetic aperture surface radar image and the second inverse synthetic aperture surface radar image, Calculating principal axes of respective images having maximum dispersion with respect to scattering points through principal component analysis using scattering points of spherical radar images, estimating a rotation speed of the object using the extracted principal axes, And changing the scale in the transverse distance direction of the first and second inverse synthesized opening surface radar images by length units using the estimated rotation speed It includes.
As described above, according to the present invention, even if there is no information about the center of rotation of the object, it is possible to scale the traverse direction of the inverse synthetic aperture surface radar image in units of Hertz in metric units. Can be improved.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and method for scaling an inverse composite aperture radar image using a principal component analysis technique,

The present invention relates to an apparatus and a method for scaling an inverse synthetic aperture radar image using a principal component analysis technique and more particularly to an apparatus and method for scaling an inverse synthetic aperture radar image using a principal component analysis technique, The present invention relates to a device for scaling a transverse distance direction of an inverse synthesized opening surface radar image using a principal component analysis technique for scaling in a distance unit and a method thereof.

Radar technology has made rapid progress since the Second World War, and recently it has reached the stage of producing images of objects using radar. The biggest advantage of radar images is that they can acquire images of objects regardless of day or night.

A typical example of such radar images is a reverse synthetic aperture radar image. The inverse synthetic aperture radar is used to acquire images of moving targets using fixed radar.

The inverse composite aperture radar image formed by the range-Doppler technique is represented by the two-dimensional shape of the scattering points of the object in the distance direction and the transverse direction, and the distance direction of the inverse composite aperture radar image is represented by the radar bandwidth In meters (m), and the transverse distance in hertz (Hz).

In this case, since the unit of distance in the horizontal direction is different from that in the direction of the horizontal distance in the vertical direction, it is necessary to scale the horizontal distance direction in units of meters (m) do.

Since crossing distance scaling proceeds without knowing the start information of the object, it is essential to estimate the rotation center of the object in order to estimate the rotation speed of the object and to perform the crossing distance scaling. Therefore, if the estimation of the center of rotation is inaccurate, a large error occurs in the scaling.

However, in the prior art, there is a problem that the estimation accuracy of the rotation center of the object is lowered and the process of estimating the rotation center is required, so that the calculation speed also falls.

The technology underlying the present invention is disclosed in Japanese Patent Laid-Open No. 10-2014-0008219 (published on Jul. 03, 2013).

SUMMARY OF THE INVENTION The present invention is directed to an apparatus and method for scaling a transverse distance direction of an inverse composite aperture radar image using a principal component analysis technique that scales the transverse distance direction of an inverse composite aperture radar image by distance units without estimating the rotational center of the object. And to provide such a method.

According to an aspect of the present invention, there is provided a method of scaling a traverse direction of an inverse synthesized opening surface radar image, the method comprising: generating a first inverse synthetic aperture surface radar image at a first time point, Acquiring a second inverse synthetic aperture surface radar image at two points in time, extracting scattering points representing the shape of the object from the first inverse synthetic aperture surface radar image and the second inverse synthetic aperture surface radar image, Calculating a principal axis of each image having a maximum variance with respect to scattering points through principal component analysis using scattering points of the first and second inverse synthesized opening surface radar images, Estimating a rotation speed of the object, and estimating a rotation speed of the object based on the estimated rotation speed, And changing the scale of the direction in units of length.

The step of calculating the principal axes of the respective images may include calculating respective average vectors of scattering points of the first and second reverse synthesized opening surface radar images in a range and a cross-range (RC) domain Calculating a difference value between the position vector of the scattering points of the first and second inverse synthesized opening surface radar images in the range and the cross-range domain and the average vector, - computing respective covariance matrices for scattering points of first and second inverse spreading aperture radar images in the range domain, and using the covariance matrices to determine the first and second ranges in the range and cross- And calculating a principal axis of the 2-ary composite aperture-based radar image.

The step of calculating the average vector may calculate the average vector (k 1 , k 2 ) through the following equation.

Figure 112016016138639-pat00001

Here, L means the number of scattering points, and p s 1, i And p s 2, i is a range and cross-representing a position vector of the first and second inverse-synthesis opening surface radar images i-th scattering points in the range domain, and [M x1, N y1] and [M x2, N y2 ] Means an average vector of the scattering points of the first and second inverse synthesized opening surface radar images in the range and cross-range domains.

The calculating of the covariance matrices may calculate the covariance matrices C p1 and C p2 using the following equations.

Figure 112016016138639-pat00002

Here, S denotes a scaling matrix, Q denotes an orthogonal matrix composed of eigenvectors of a matrix A arranged in descending order, A denotes a diagonal matrix having two eigenvalues, R denotes a rotation matrix ,

Figure 112016016138639-pat00003
ego,
Figure 112016016138639-pat00004
Denotes a data set of scattering points of the first inverse synthetic aperture surface radar image, and L denotes the number of scattering points.

The step of calculating the principal axis may calculate the principal axis through the following equation.

Figure 112016016138639-pat00005

E 1 is the principal axis vector of the first inverse composite aperture radar image in the range and cross-range domain, and e 2 is the principal axis vector of the second inverse composite aperture surface radar image in the range and cross- and, (

Figure 112016016138639-pat00006
) 1 denotes an eigenvector corresponding to the maximum eigenvalue.

Wherein the step of estimating the rotational speed of the target object includes the steps of: estimating a rotational angle of the target object by performing an inner product calculation of an eigenvector for the principal axes of the respective images; and estimating a rotational speed of the target object using the estimated rotational angle And a step of estimating.

The step of estimating the rotational speed of the target object using the estimated rotational angle may estimate the rotational angle of the target object through the following equation.

Figure 112016016138639-pat00007

here,

Figure 112016016138639-pat00008
Refers to the estimated angle of rotation, and Tm refers to the sum total of concatenation processing time interval, and f c denotes the carrier frequency of the radar transmitted signal, and w means the actual rotational speed of the target object, and e x1 and e y1 Denotes the distance and transversal distance of the eigenvector with respect to the principal axis of the first de-interrogating aperture radar image in the range-doppler (RD) domain, and e x2 and e y2 denotes a distance and a transverse distance of the eigenvector with respect to the main axis of the second inverse synthetic aperture surface radar image, and B denotes a bandwidth of the transmission signal of the radar.

Wherein the step of changing the scale in the transverse distance direction in units of length comprises the steps of generating a cost function for the rotation speed using the estimated rotation speed, receiving a candidate rotation speed, Detecting a rotational speed value at which the value of the cost function becomes zero using a bisection algorithm, and detecting a rotational speed value of the first and second reverse synthesized opening and closing radar images And changing the scale in the transverse distance direction in units of length.

The step of generating the cost function may generate the cost function E (w) through the following equation.

Figure 112016016138639-pat00009

Here, w denotes a rotation speed,

Figure 112016016138639-pat00010
Figure 112016016138639-pat00011
Means the estimated rotational speed.

The apparatus for scaling a transverse distance direction of an inverse synthesized opening surface radar image according to another embodiment of the present invention includes a first inverse synthesized opening surface radar image at a first time point and a second inverse composite at a second time point later than the first time point, An extracting unit for extracting scattering points representing the shape of the object from the first inverse synthesized opening surface radar image and the second inverse synthetic aperture surface radar image, An arithmetic unit operable to calculate a main axis of each image having a maximum variance with respect to scattering points through analysis of principal components using scattering points of a two-arcs synthetic aperture radar image, a rotation speed of the object using the main axis of each extracted image, And a scale in the transverse direction of the first and second inverse synthesized opening surface radar images using the estimated rotational speed, Scale includes unit for changing a.

As described above, according to the present invention, even if there is no information about the center of rotation of the object, it is possible to scale the traverse direction of the inverse synthetic aperture surface radar image in units of Hertz in metric units. Can be improved.

FIG. 1A is a view for explaining an inverse synthesized opening surface radar system according to an embodiment of the present invention.
FIG. 1B is a view for explaining a transverse distance direction scaling apparatus for an inverse synthesized opening surface radar image according to an embodiment of the present invention.
2 is a flowchart of a method for scaling the traverse direction of an inverse synthesized opening surface radar image according to an embodiment of the present invention.
FIG. 3 is a view for explaining the step S230 according to the embodiment of the present invention.
4 is a flow chart of step S230 according to an embodiment of the present invention.
5 is a flow chart of step S240 according to an embodiment of the present invention.
6 is a flow chart of step S250 according to an embodiment of the present invention.
7 is a diagram for explaining a cost function according to an embodiment of the present invention.
8 is a diagram for explaining a process of calculating a rotational speed value using a dichotomy according to an embodiment of the present invention.
9 is a diagram showing a simulation result according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

Throughout the specification, when an element is referred to as "comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention.

FIG. 1A is a view for explaining an inverse synthesized opening surface radar system according to an embodiment of the present invention.

As shown in FIG. 1A, an inverse synthetic aperture radar system acquires an image of a moving object using a fixed radar. Specifically, the radar 200 transmits a radio wave, analyzes the signal reflected and reflected by the target object, and obtains an inverse synthesized opening and closing radar image.

Then, the radar 200 transmits the acquired inverse synthesized opening surface radar image to the transverse direction direction scaling device 100.

First, the distance from the radar 200 to the scattering point of the object located at (x, y) can be expressed by Equation (1) below.

Figure 112016016138639-pat00012

Here, θ (t m ) can be expressed as θ (t m ) = t m w as the rotation angle of the object at time t m , and w means the rotation speed of the object. And r 0 is the distance from the radar to the center of rotation of the object.

At this time, since it can be assumed that the coherent processing interval (CPI) is sufficiently small to maintain a constant rotation speed, the distance from the radar 200 to the scattering point (x, y) have.

Figure 112016016138639-pat00013

Therefore, the backscattering signal reflected and reflected by each scattering point (x i , y i ) of the object can be expressed by the following Equation (3).

Figure 112016016138639-pat00014

Here, f denotes the frequency of the radar transmission signal, L denotes the number of scattering points, A i denotes the magnitude of the i-th scattering point, f c denotes the carrier frequency of the radar transmission signal, B denotes the bandwidth of the transmission signal of the radar, Tm denotes the total of the joint processing time intervals, and c denotes the speed of light.

Then, an inverse synthesized opening surface radar image I (r, f d ) can be generated using a backscattering signal, which can be expressed by Equation (4).

Figure 112016016138639-pat00015

Here, FT means Fourier transform, and, r means a distance in the distance direction f d denotes the Doppler frequency in the transverse direction of distances.

On the other hand, if the inverse composite aperture radar image generated at time t 1 is I 1 and the inverse composite aperture surface radar image generated at time t 2 is I 2 , then the scattering point at I 1 and I 2 in the distance-Doppler domain of the data set (P 1, P 2) is the distance direction as shown in equation 5 it can be expressed as empty (m, n) of the cross-sectional direction distance.

Figure 112016016138639-pat00016

Here, p 1, i and p 2, i mean the positions of the i-th scattering points in I 1 and I 2 , respectively, and p 1, i = [m 1i , n 1i ] T and p 2, 2i , n 2i ] T.

1A, the airplane shown by the solid line means the object at time t1, and the airplane shown by the dotted line indicates the object at time t2. Therefore, if the object is an airplane as shown in FIG. 1A, P1 is an airplane And P2 represents the data set of the scattering point of the airplane shown by the dotted line.

Then, the data set (P 1 , P 2 ) of the scattering points at I 1 and I 2 is stored in the distance-Doppler domain (expressed as meter-by-hertz, hereinafter referred to as RD domain) (P 1 S , P 2 S ) of a scattering point scaled by a range and a cross-range (RC) domain (represented by a meter-by-meter, Can be expressed by Equation (6) below.

Figure 112016016138639-pat00017

Here, S means a scaling matrix and is expressed as S = diag [eta r , eta c ].

Here, η r denotes a distance direction scaling coefficient, and η c denotes a scaling coefficient in the transverse direction, which is expressed by Equation (7) below.

Figure 112016016138639-pat00018

Here, T m is a time difference (t 2 -t 1 ) between t 1 and t 2 .

On the other hand, the data set (P 1 S , P 2 S ) of the scaled scattering point can be expressed as Equation 8 below

Figure 112016016138639-pat00019

Here, R denotes a rotation matrix, and a rotation matrix at a time interval T m can be expressed by Equation (9) below.

Figure 112016016138639-pat00020

Therefore, if the correct rotation angle w can be computed, the cross-directional direction of the inverse composite aperture radar image can be scaled to a distance-domain in the Hertz-domain.

Next, a description will be made of a configuration of an apparatus 100 for scrolling a transverse distance direction of an inverse synthesized opening surface radar image according to an embodiment of the present invention with reference to FIG. FIG. 1B is a view for explaining a transverse distance direction scaling apparatus for an inverse synthesized opening surface radar image according to an embodiment of the present invention.

1B, the transverse direction direction scaling apparatus 100 includes an image acquisition unit 110, an extraction unit 120, an operation unit 130, an estimation unit 140, and a schedule unit 150.

First, the image acquiring unit 110 acquires a first de-interlacing opening surface radar image at a first time point and a second de-composition opening surface surface radar image at a second time point later than the first time point.

Here, the first and second deinterlaced aperture radar images are obtained by first and second deinterlaced aperture radar images in the RD domain, and the distance-Doppler domain is obtained from the first and second deinterlaced aperture radar images The down-range direction is expressed in length units, and the vertical axis, cross-range direction, is expressed in frequency units.

Next, the extracting unit 120 extracts scattering points representing the shape of the object from the first inverse synthesized opening surface radar image and the second inverse synthesized opening surface radar image. Here, scattering points indicate scattering points that are strong to change and can be extracted using FAST algorithm.

The arithmetic unit 130 calculates principal axes of each image having the maximum variance with respect to the scattering points through principal component analysis using the extracted scattering points of the first and second reverse synthesized aperture surface radar images.

Specifically, the calculation unit 130 calculates the respective average vectors of the scattering points of the first and second de-interlacing opening surface radar images in the RC domain, and calculates the average of the first and second scattering points in the range and cross- The position vector of the scattering points of the inverse synthesized opening surface radar image and the difference value of the average vector are calculated.

The arithmetic unit 130 computes respective covariance matrices for the scattering points of the first and second uncompensated opening surface radar images in the RC domain using the difference value, Of the first and second inverse synthesized aperture radar images.

Then, the estimator 140 estimates the rotation speed of the object using the main axis of each extracted image.

Specifically, the estimator 140 internally computes an eigenvector for the major axes of each image to estimate the rotation angle of the target object, and estimates the rotation velocity of the target object using the estimated rotation angle.

Then, the scale unit 150 changes the scale of the first and second inverse synthesized opening and closing radar images in the transverse direction in units of length by using the estimated rotation speed.

Specifically, the scheduling unit 150 generates a cost function for the rotation speed using the estimated rotation speed, receives the candidate rotation speed, and calculates a cost function using the candidate rotation speed, the cost function, and the bisection algorithm, Is detected as a rotation speed value.

The scale part 150 changes the scale of the first and second reverse synthesized opening and closing radar images in the transverse distance direction by length unit using the detected rotation speed value.

A method of scaling the traverse direction of the inverse synthesized opening surface radar image according to an embodiment of the present invention will now be described with reference to FIGS. 2 to 8. FIG. 2 is a flowchart of a method for scaling the traverse direction of an inverse synthesized opening surface radar image according to an embodiment of the present invention.

First, the transverse direction direction scaling apparatus 100 acquires a first de-interlaced opening surface radar image at a first point of time and a second de-interlaced opening surface radar image at a second point later than the first point in step S210.

For example, the first reverse synthesized opening surface radar image at the first time point represents the P1 image shown by the solid line in FIG. 1A, and the second reverse synthesized opening surface radar image at the second time point is shown by the dotted line in FIG. It represents one P2 image.

At this time, the first and second deinterlaced opening surface radar images are acquired on the RD domain, the x-axis direction as the horizontal axis is represented by a length unit (in meters) in the down-range, The cross-range is expressed in frequency units (in hertz).

Then, the transverse direction direction scaling apparatus 100 extracts scattering points representing the shape of the object from the first de-interlacing opening surface radar image and the second de-interlacing opening surface radar image (S220).

At this time, a large number of scattering points can be extracted. However, according to the embodiment of the present invention, the scattering points extracted are scattering points that are not sensitive to a change in observation angle, that is, robust scattering points.

Meanwhile, according to the embodiment of the present invention, the transverse direction direction scaling apparatus 100 can use the FAST algorithm to extract scattering points of the first and second inverse synthesized opening surface radar images.

Here, the FAST (Features from Accelerated Segment Test) algorithm is an algorithm for extracting feature points of an image, and is an algorithm for extracting corner points of an image. Specifically, the FAST algorithm calculates the number of pixels that are darker than a certain value (p + t) or less than a certain value (pt) than a certain point (p) (P) is determined as a corner. Therefore, the FAST algorithm is used in various versions depending on how the number of pixels (N) is selected.

According to an embodiment of the present invention, the transverse direction direction scaling apparatus 100 may select N = 10 (FAST-10) and t = 3.5 to extract the scattering point stronger to the change, The computation efficiency can be increased.

Also, as a preprocessing step before extracting a strong scattering point by using the FAST algorithm, the transverse direction direction scaling apparatus 100 calculates the side scattering function (Point Spread Functions) of the inverse synthetic aperture surface radar image sidelove) in a windowing operation.

Next, the traverse direction direction scaling apparatus 100 calculates a principal axis of each image having maximum dispersion with respect to scattering points through principal component analysis using scattering points of the extracted first and second inverse synthesized opening surface radar images (S230).

Here, Principal Component Analysis (PCA) is a statistical technique for extracting a principal component that concisely expresses a pattern of a dispersion method of many variables as a linear combination (average point to weight) of an original variable. In the present invention, the main axis of each image is calculated through principal component analysis based on the fact that scattering points of an image have a specific direction having maximum dispersion.

FIG. 3 is a view for explaining the step S230 according to the embodiment of the present invention. As shown in FIG. 3, the scattering point set of the object has a direction having maximum dispersion, that is, a main axis (long arrow in FIG. 3) When the object rotates, the main axis also rotates.

Therefore, the transverse direction direction scaling apparatus 100 can know the rotation angle of the object by comparing the principal axes of the two images. That is, in order to calculate the rotation angle of the object, the present invention computes the principal axis having the maximum variance of the scattering point set of the object through principal component analysis.

In FIG. 3, the short arrow indicates the direction having the second largest variance value.

Then, step S230 will be described in detail with reference to FIG. 4 is a flow chart of step S230 according to an embodiment of the present invention.

First, the transverse direction direction scaling apparatus 100 calculates respective average vectors of scattering points of the first and second uncovered opening surface radar images in the RC domain (S231).

Specifically, the transverse direction direction scaling apparatus 100 calculates the mean vector (k 1 , k 2 ) of the data set (P 1 S , P 2 S ) of the scattering points in the RC domain through Equation .

Figure 112016016138639-pat00021

Here, [M x1 , N y1 ] and [M x2 , N y2 ] are the average vectors of the scattering point positions of the first and second reverse synthesized opening surface radar images in the RC domain, respectively.

Next, the transverse direction direction scaling apparatus 100 calculates the difference value between the position value and the average value of the scattering points of the first and second inverse synthesized opening surface radar images in the RC domain (S232).

First, since the scattering point sets of the first and second inverse synthesized opening surface radar images are in a rotation relationship, the transverse direction direction scaling apparatus 100 calculates the data sets (P 1 S , P 2 S ) of the scattering points in the RC domain, (K 1 , k 2 ) are represented by matrices K 1 and K 2 .

Figure 112016016138639-pat00022

Here, K 1 (

Figure 112016016138639-pat00023
) Is K 1 = [k 1 , k 1 , ..., k 1 ], and K 2
Figure 112016016138639-pat00024
) Is K 2 = [k 2 , k 2 , ..., k 2 ].

The transverse direction direction scaling apparatus 100 calculates the difference value (P 1 S , P 2 S ) of the scattering points (P 1 S , P 2 S ) and the mean vector matrix (K 1 , K 2 )

Figure 112016016138639-pat00025
,
Figure 112016016138639-pat00026
) By the following equation (12).

Figure 112016016138639-pat00027

Then, the transverse direction direction scaling apparatus 100 calculates the respective covariance matrices C P1 and C P2 for the scattering points of the first and second inverse synthesized opening surface radar images in the RC domain using the difference value (S233).

Specifically, the transverse distance direction scaling apparatus 100 calculates a covariance matrix C P1 for scattering points of the first de-interleaved opening surface radar image in the RC domain through Equation (13) and Equation (14) below .

Figure 112016016138639-pat00028

Where A denotes a symmetric matrix of covariances.

Figure 112016016138639-pat00029

Here, A denotes a diagonal matrix having two eigenvalues, and Q denotes an orthogonal matrix composed of eigenvectors of a matrix A arranged in descending order.

The transverse distance direction scaling apparatus 100 computes a covariance matrix C P2 for the scattering points of the second de-interleaved opening surface radar image in the RC domain through the following equations (15) and (16).

Figure 112016016138639-pat00030

Figure 112016016138639-pat00031

Then, the transverse direction direction scaling apparatus 100 calculates principal axes of respective images in the RC domain using the covariance matrices (S234). Specifically, the transverse direction direction scaling apparatus 100 calculates the principal axes of the first and second inverse synthesized opening surface radar images through Equation (17) below.

Figure 112016016138639-pat00032

Where e 1 denotes the principal axis vector of the first de-composite aperture sidewall radar image in the RC domain, e 2 denotes the principal vector of the second de-composition aperture sidewall radar image in the RC domain, and ( 1 ) Quot; means an eigenvector corresponding to a value.

After calculating the main axis in step S230, the transverse direction direction scaling apparatus 100 estimates the rotational speed of the target object by using the main axis of each extracted image (S240). FIG. 5 is a flow chart of step S240 according to an embodiment of the present invention, and step S240 will be described in detail with reference to FIG.

First, the traverse direction direction scaling apparatus 100 internally computes an eigenvector for principal axes of each image to estimate the rotation angle of the object (S241).

As described above, since the scattering points of the two images have a rotation relationship, the rotation relationship between the principal axes can be expressed by Equation (18) below.

Figure 112016016138639-pat00033

The principal axes ((Q) 1 , (RQ) 1 ) of the first and second reverse synthesized opening surface radar images in the RD domain can be calculated by the following equation (19).

Figure 112016016138639-pat00034

Here, [e x1 , e y1 ] T and [e x2 , e y2 ] T are the eigenvectors of the principal axes of the first and second de-interlaced opening surface radar images in the RD domain.

Therefore, the main axes of the first and second inverse synthesized opening surface radar images in the RC domain with respect to the rotation speed w can be expressed as Equation (20) below.

Figure 112016016138639-pat00035

Accordingly, the transverse direction direction scaling apparatus 100 computes the rotation angle of the object by performing an inner product computation of the eigenvectors of the principal axes of the first and second inverse distance-scaled first and second synthesized opening surface radar images by the following expression (21).

Figure 112016016138639-pat00036

here,

Figure 112016016138639-pat00037
Is an estimated rotation angle.

Then, the transverse direction direction scaling apparatus 100 estimates the rotational speed of the object using the estimated rotational angle (S242).

As described above,

Figure 112016016138639-pat00038
The transverse direction direction scaling apparatus 100 can calculate the estimated rotation speed (< RTI ID = 0.0 >
Figure 112016016138639-pat00039
).

Figure 112016016138639-pat00040

In step S240, the rotational speed is estimated and the scale of the first and second reverse synthesized opening and closing radar images in the transverse direction is changed in units of length by using the estimated rotational speed in step S250.

FIG. 6 is a flow chart of step S250 according to an embodiment of the present invention, and step S250 will be described in detail with reference to FIG.

First, the transverse direction direction scaling apparatus 100 generates a cost function for the rotation speed using the estimated rotation speed (S251). At this time, the cost function is expressed by Equation 23 below.

Figure 112016016138639-pat00041

7 is a diagram for explaining a cost function according to an embodiment of the present invention.

As can be seen from the equation (21), when the arbitrary rotation speed w is smaller than the actual rotation speed (w true ) as shown in FIG. 8 (a)

Figure 112016016138639-pat00042
) Becomes larger than the actual rotation speed (w true ) (w true <
Figure 112016016138639-pat00043
). This is because the transverse distance axis expansion induced by the inverse relationship between any of the rotational speed (w) and the transverse distance resolution (Δy = C / (2f c T m w)).

Likewise, when the actual rotational speed w true is smaller than the arbitrary rotational speed w (w true <w), the estimated rotational speed (w true )

Figure 112016016138639-pat00044
) Becomes smaller than the actual rotation speed (w true ) (
Figure 112016016138639-pat00045
<w true ).

That is, when the arbitrary rotation speed w is less than the estimated rotation speed

Figure 112016016138639-pat00046
), The estimated rotational speed (
Figure 112016016138639-pat00047
) Is the actual rotation speed of the object (w true ). Therefore, the cost function in the present invention is generated as shown in Equation (23) using the error between w and w.

Then, the transverse direction direction scaling apparatus 100 receives the candidate rotation speed (S252), and then uses the candidate rotation speed, the cost function, and the bisection algorithm to calculate a rotation speed value at which the value of the cost function becomes zero (S253).

As described above, when the cost function of the present invention is 0, the actual rotational speed of the object is detected. When the cost function is zero, the rotational speed w may be zero at the actual rotational speed w true and in the following equation (24).

Figure 112016016138639-pat00048

Then, the transverse direction direction scaling apparatus 100 detects a rotational speed value at which the cost function becomes zero by using the dichotomy, wherein the dichotomy is repeatedly performed to divide the section into two halves and select the sub- How to find a muscle.

8 is a diagram for explaining a process of calculating a rotational speed value using a dichotomy according to an embodiment of the present invention. As shown in FIG. 8, the input candidate rotation speeds a 1 and b 1 can be input in consideration of a point where the cost function becomes zero. In the embodiment of the present invention, it is exemplified that a = 0.01 deg. / Tm and b = 8.00 deg. Tm are selected to increase the calculation performance, and the input candidate rotation speed can be changed by a person skilled in the art.

The curve shown in FIG. 8 is a graph showing the cost function. In step S252, a cost function value corresponding to w 1 , which is an intermediate value between two input values, is sought. At this time, since the cost function value of w 1 has a negative value, a cost function value corresponding to w 2 , which is an intermediate value between b 1 and w 1 having a positive value, is found.

Since the cost function value corresponding to w 2 is a positive value, in this case, the cost function value corresponding to w 3 , which is an intermediate value between w 1 having a negative value in the previous step, is found. The transverse direction direction scaling apparatus 100 repeats this process to detect a rotational speed value at which the cost function becomes zero.

At this time, since the synchro- nization processing interval CPI is set to be short enough such that the rotation speed is uniform, the transverse direction direction scaling apparatus 100 can detect the rotation speed quickly.

Then, the transverse direction direction scaling apparatus 100 changes the scale of the first and second reverse synthesized opening and closing radar images in the transverse distance direction in units of length by using the detected rotation speed values (S255).

Specifically, the transverse direction direction scaling apparatus 100 can calculate the scaling factor? C in the transverse direction using the detected rotational speed value and Equation (7) Change the scale in the transverse distance direction of the inverse composite aperture radar image from Hertz to length units.

Hereinafter, the simulation results of the prior art and the present invention will be described with reference to FIG.

9 is a diagram showing a simulation result according to an embodiment of the present invention.

FIG. 9A shows a simulation result of comparing the actual rotation speed of the present invention and the conventional rotation speed with the detected rotation speed, and FIG. 9B shows the comparison between the present invention and the conventional calculation speed A simulation result is shown.

9 (a) and 9 (b), a rotational correlation method (RCM) means a case using a conventional rotation center.

In FIG. 9 (a), root mean square errors (RMSE) is an index indicating the degree of error between the actual rotational speed and the detected rotational speed, and varies depending on the signal-to-noise ratio (SNR) The error level of the present invention is about 1.5 * 10 -3 , which is about twice the error improvement effect.

As shown in FIG. 9 (b), the computation time is reduced by about 4 seconds compared with the conventional method using the center of rotation, which indicates that the speed improvement effect is about four times or more.

According to the embodiment of the present invention, even if there is no information about the rotation center of the object, it is possible to scale the transverse distance direction of the inverse synthetic aperture surface radar image in units of Hertz in meters, The accuracy of scaling can be improved.

While the present invention has been described with reference to exemplary embodiments, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

100: transverse direction direction scaling device 110:
120: Extracting unit 130:
140: Chuo government 150: Scale part
200: Radar

Claims (18)

Acquiring a first de-composite aperture sidewall radar image at a first time point and a second de-composite aperture sidewall radar image at a second time point later than the first time point,
Extracting scattering points representing a shape of a target object from the first de-composition aperture grid radar image and the second de-composition aperture grid radar image,
Calculating a principal axis of each image having a maximum variance with respect to scattering points by performing principal component analysis using scattering points of the extracted first and second reverse synthesized opening surface radar images,
Estimating a rotation speed of the object using the main axis of each extracted image, and
And scaling the transverse distance direction of the inverse synthesized opening surface radar image including the step of changing the scale in the transverse distance direction of the first and second inverse synthesized opening surface radar images by length units using the estimated rotation speed . The method according to claim 1,
Wherein the step of calculating the main axis of each image comprises:
Calculating respective average vectors of scattering points of first and second reverse synthesized opening surface radar images in a range and a cross-range (RC) domain,
Calculating a difference value between the position vector of the scattering points of the first and second inverse synthesized opening surface radar images in the range and the cross-range domain and the average vector,
Calculating respective covariance matrices for scattering points of the first and second reverse synthesized opening surface radar images in the range and the cross-range domain using the difference value, and
And computing a principal axis of the first and second inverse synthesized opening surface radar images in the range and cross-range domains using the covariance matrices. 3. The method of claim 2,
Wherein the step of calculating the average vector comprises:
A method for scaling the traverse direction of an inverse composite aperture radar image for calculating the average vector (k 1 , k 2 ) using the following equation:
Figure 112016016138639-pat00049

Here, L means the number of scattering points, and p s 1, i And p s 2, i is a range and cross-representing a position vector of the first and second inverse-synthesis opening surface radar images i-th scattering points in the range domain, and [M x1, N y1] and [M x2, N y2 ] Means an average vector of the scattering points of the first and second inverse synthesized opening surface radar images in the range and cross-range domains. 3. The method of claim 2,
Computing the covariance matrices comprises:
A method for scaling the transverse distance direction of an inverse synthesized opening surface radar image for computing the covariance matrices C p1 and C p2 using the following equation:
Figure 112016016138639-pat00050

Here, S denotes a scaling matrix, Q denotes an orthogonal matrix composed of eigenvectors of a matrix A arranged in descending order, A denotes a diagonal matrix having two eigenvalues, R denotes a rotation matrix ,
Figure 112016016138639-pat00051
ego,
Figure 112016016138639-pat00052
Denotes a data set of scattering points of the first inverse synthetic aperture surface radar image, and L denotes the number of scattering points. 5. The method of claim 4,
Wherein the step of calculating the principal axis comprises:
A method for scaling the transverse distance direction of an inverse synthetic aperture surface radar image for calculating the principal axis through the following equation:
Figure 112016016138639-pat00053

E 1 is the principal axis vector of the first inverse composite aperture radar image in the range and cross-range domain, and e 2 is the principal axis vector of the second inverse composite aperture surface radar image in the range and cross- and, (
Figure 112016016138639-pat00054
) 1 denotes an eigenvector corresponding to the maximum eigenvalue. The method according to claim 1,
Wherein the step of estimating the rotational speed of the object comprises:
Estimating a rotation angle of the target object by performing an inner product operation on an eigenvector for the principal axes of the respective images,
Estimating a rotation speed of the object based on the estimated rotation angle; and scaling the traverse direction of the inverse composite aperture surface radar image. The method according to claim 6,
Estimating a rotation speed of the object using the estimated rotation angle,
A method of scaling a transverse distance direction of an inverse synthesized opening surface radar image for estimating a rotation angle of the object through the following equation:
Figure 112016016138639-pat00055

here,
Figure 112016016138639-pat00056
Refers to the estimated angle of rotation, and Tm refers to the sum total of concatenation processing time interval, and f c denotes the carrier frequency of the radar transmitted signal, and w means the actual rotational speed of the target object, and e x1 and e y1 Denotes the distance and transversal distance of the eigenvector with respect to the principal axis of the first de-interrogating aperture radar image in the range-doppler (RD) domain, and e x2 and e y2 denotes a distance and a transverse distance of the eigenvector with respect to the main axis of the second inverse synthetic aperture surface radar image, and B denotes a bandwidth of the transmission signal of the radar. The method according to claim 1,
Wherein the step of changing the scale in the transverse distance direction in units of length comprises:
Generating a cost function for the rotation speed using the estimated rotation speed,
Receiving the candidate rotation speed,
Detecting a rotational speed value at which the value of the cost function becomes zero using the candidate rotational speed, the cost function, and a bisection algorithm, and
And scaling the transverse distance direction of the inverse synthesized opening surface radar image including the step of changing the scale in the transverse distance direction of the first and second inverse synthesized opening surface radar images by length units using the detected rotation speed values Way. 9. The method of claim 8,
Wherein generating the cost function comprises:
A method for scaling a transverse distance direction of an inverse synthesized opening surface radar image that generates the cost function (E (w)) through the following equation:
Figure 112016016138639-pat00057

Here, w denotes a rotation speed,
Figure 112016016138639-pat00058
Means the estimated rotational speed. An image acquiring unit acquiring a first de-interlaced opening surface radar image at a first point in time and a second de-composition aperture surface radar image at a second point later than the first point in time,
An extracting unit for extracting scattering points representing the shape of the object from the first de-composition aperture grid radar image and the second de-
A calculation unit for calculating a principal axis of each image having a maximum variance with respect to scattering points through principal component analysis using scattering points of the extracted first and second reverse synthesized opening surface radar images,
An estimator for estimating a rotation speed of the object using the extracted principal axes of the respective images, and
And a scale unit for changing, in units of length, a scale in a transverse distance direction of the first and second inverse synthesized opening surface radar images using the estimated rotation speed. 11. The method of claim 10,
The operation unit,
Range and cross-range (RC) domains of the first and second reverse synthesized opening surface radar images, and calculates the respective average vectors for the scattering points in the range and cross-range domain Calculating a difference value between the position vector of the scattering points of the first and second inverse synthesized opening surface radar images and the average vector and calculating a difference between the first and second inverse combinations in the range and cross- An inverse synthesizing unit for computing respective covariance matrices for scattering points of the spherical radar image and calculating the principal axes of the first and second reverse synthesized opening surface radar images in the range and cross-range domains using the covariance matrices An apparatus for scaling a directional distance of a spherical radar image. 12. The method of claim 11,
The operation unit,
A transverse distance direction scaling apparatus for an inverse synthesized opening surface radar image that computes the average vector (k 1 , k 2 ) using the following equation:
Figure 112016016138639-pat00059

Here, L means the number of scattering points, and p s 1, i And p s 2, i is a range and cross-representing a position vector of the first and second inverse-synthesis opening surface radar images i-th scattering points in the range domain, and [M x1, N y1] and [M x2, N y2 ] Means an average vector of the scattering points of the first and second inverse synthesized opening surface radar images in the range and cross-range domains. 12. The method of claim 11,
The operation unit,
A transverse direction direction scaling apparatus for an inverse synthetic aperture surface radar image that computes the covariance matrixes ( Cp1 , Cp2 ) using the following equation:
Figure 112016016138639-pat00060

Here, S denotes a scaling matrix, Q denotes an orthogonal matrix composed of eigenvectors of a matrix A arranged in descending order, A denotes a diagonal matrix having two eigenvalues, R denotes a rotation matrix ,
Figure 112016016138639-pat00061
ego,
Figure 112016016138639-pat00062
Denotes a data set of scattering points of the first inverse synthetic aperture surface radar image, and L denotes the number of scattering points. 14. The method of claim 13,
The operation unit,
An apparatus for scaling an intersection distance direction of an inverse composite aperture radar image for calculating the principal axis through the following equation:
Figure 112016016138639-pat00063

E 1 is the principal axis vector of the first inverse composite aperture radar image in the range and cross-range domain, and e 2 is the principal axis vector of the second inverse composite aperture surface radar image in the range and cross- and, (
Figure 112016016138639-pat00064
) 1 denotes an eigenvector corresponding to the maximum eigenvalue. 11. The method of claim 10,
Wherein the estimating unit comprises:
A transverse distance directional scaling of an inverse synthetic aperture surface radar image for estimating a rotational angle of the object by internally calculating an eigenvector for principal axes of each image and estimating a rotational velocity of the object using the estimated rotational angle; Device. 16. The method of claim 15,
Wherein the estimating unit comprises:
A transverse distance direction scaling device for an inverse synthesized opening surface radar image for estimating a rotation angle of the object through the following equation:
Figure 112016016138639-pat00065

here,
Figure 112016016138639-pat00066
Refers to the estimated angle of rotation, and Tm refers to the sum total of concatenation processing time interval, and f c denotes the carrier frequency of the radar transmitted signal, and w means the actual rotational speed of the target object, and e x1 and e y1 Denotes the distance and transversal distance of the eigenvector with respect to the principal axis of the first de-interrogating aperture radar image in the range-doppler (RD) domain, and e x2 and e y2 denotes a distance and a transverse distance of the eigenvector with respect to the main axis of the second inverse synthetic aperture surface radar image, and B denotes a bandwidth of the transmission signal of the radar. 11. The method of claim 10,
The scale unit includes:
Generating a cost function for the rotation speed using the estimated rotation speed, receiving a candidate rotation speed, calculating a cost function using the candidate rotation speed, the cost function, and a bisection algorithm, Of the first and second reverse synthesized opening and closing radar images is changed in units of length by using the detected rotational speed values, Distance direction scaling device. 18. The method of claim 17,
The scale unit includes:
A transverse direction direction scaling apparatus of an inverse synthesized opening surface radar image generating the cost function (E (w)) by the following equation:
Figure 112016016138639-pat00067

Here, w denotes a rotation speed,
Figure 112016016138639-pat00068
Means the estimated rotational speed.
KR1020160019028A 2016-02-18 2016-02-18 Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof KR101834063B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020160019028A KR101834063B1 (en) 2016-02-18 2016-02-18 Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020160019028A KR101834063B1 (en) 2016-02-18 2016-02-18 Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof

Publications (2)

Publication Number Publication Date
KR20170097373A KR20170097373A (en) 2017-08-28
KR101834063B1 true KR101834063B1 (en) 2018-02-28

Family

ID=59759814

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020160019028A KR101834063B1 (en) 2016-02-18 2016-02-18 Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof

Country Status (1)

Country Link
KR (1) KR101834063B1 (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101854573B1 (en) * 2017-10-13 2018-06-08 포항공과대학교 산학협력단 Apparatus for cross-range scaling of isar image usnig radon transform and projection and method thereof
CN112069651B (en) * 2020-07-23 2024-04-09 西安空间无线电技术研究所 Method for estimating spin-stabilized target rotation axis based on ISAR imaging
CN114966693B (en) * 2022-07-20 2022-11-04 南京信息工程大学 Airborne ship target ISAR refined imaging method based on deep learning

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001264435A (en) 2000-03-21 2001-09-26 Mitsubishi Electric Corp Radar device
JP2006242632A (en) 2005-03-01 2006-09-14 Mitsubishi Electric Corp Radar device
KR101103416B1 (en) 2010-07-21 2012-01-06 영남대학교 산학협력단 Method of cross-range scaling for inverse synthetic aperture radar image and apparatus thereof

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001264435A (en) 2000-03-21 2001-09-26 Mitsubishi Electric Corp Radar device
JP2006242632A (en) 2005-03-01 2006-09-14 Mitsubishi Electric Corp Radar device
KR101103416B1 (en) 2010-07-21 2012-01-06 영남대학교 산학협력단 Method of cross-range scaling for inverse synthetic aperture radar image and apparatus thereof

Also Published As

Publication number Publication date
KR20170097373A (en) 2017-08-28

Similar Documents

Publication Publication Date Title
US11340332B2 (en) Method and apparatus for processing radar data
CN101498788B (en) Target rotation angle estimating and transverse locating method for inverse synthetic aperture radar
CN109490874B (en) Method for determining suitability of radar target as position landmark
US20170356991A1 (en) Radar device and detection method
CN108469608B (en) Method for accurately estimating Doppler centroid of moving platform radar
KR100795497B1 (en) Wave measure method and system using radar
JP5181704B2 (en) Data processing apparatus, posture estimation system, posture estimation method and program
KR101628154B1 (en) Multiple target tracking method using received signal strengths
JP2017067756A (en) Object detection device and object detection method
KR101834063B1 (en) Apparatus of cross-range scaling for inverse synthetic aperture radar image using principal component analysis and method thereof
EP0888560A1 (en) Improved method of moment estimation and feature extraction for devices which measure spectra as a function of range or time
CN115546526B (en) Three-dimensional point cloud clustering method, device and storage medium
JP6440912B2 (en) Radar equipment
Pengcheng et al. Radar group target recognition based on HRRPs and weighted mean shift clustering
CN104155653B (en) SAR back projection imaging method based on feature distance subspace
US12092733B2 (en) Radar anti-spoofing system for identifying ghost objects created by reciprocity-based sensor spoofing
KR101854573B1 (en) Apparatus for cross-range scaling of isar image usnig radon transform and projection and method thereof
WO2020171881A1 (en) Detection system and method for characterizing target height
CN111708028B (en) SAR image secondary imaging method based on multi-fractal spectrum
JP5460399B2 (en) Image radar device
JP2008304329A (en) Measuring device
Wan et al. Azimuth cutoff compensation method for SAR wave observation based on multiview wave spectrum data fusion
JP7188594B2 (en) Image processing device and image processing method
Wang et al. A novel algorithm for ocean wave direction inversion from X-band radar images based on optical flow method
Shen et al. An alternative method for surface current extraction from X-band marine radar images

Legal Events

Date Code Title Description
A201 Request for examination
E701 Decision to grant or registration of patent right
GRNT Written decision to grant