KR101029175B1 - METHOD OF GENERATING LONG RANGE INVERSE SYNTHETIC APERTURE RADAR IMAGE using polynomial and Gaussian basis function AND APPARATUS THEREOF - Google Patents

Abstract

The present invention relates to a method and apparatus for generating a long range inverse synthetic aperture radar image using a polynomial and a Gaussian function. According to the present invention, a method for generating a reverse synthetic aperture radar image of a target in flight, comprising: receiving a reflected wave scattered and reflected from a target, and generating an estimated trajectory of the target using a distance side view extracted from the reflected wave Step, modeling the estimated trajectory as an orbital function combined with a polynomial and a Gaussian function, obtaining the center of gravity trajectory of the distance profile, and setting the parameter values constituting the trajectory function to minimize the error between the estimated trajectory and the center of gravity trajectory. Adjusting the parameter values to maximize the cost function of the distance profile, compensating for the phase error of the distance-aligned distance profile, and generating a reverse synthetic aperture radar image using the distance profile It includes. As described above, according to the present invention, the orbital function combined with the polynomial and the Gaussian function is primarily modeled so that the difference between the center of gravity and the estimated trajectory is minimized, and is modeled secondarily through the PSO algorithm, thereby more accurately and precisely. Distance alignment can be performed.

ISAR, radar, distance profile, distance component, PSO, cost function

Description

FIELD OF GENERATING LONG RANGE INVERSE SYNTHETIC APERTURE RADAR IMAGE using polynomial and Gaussian basis function AND APPARATUS THEREOF}

The present invention relates to a method and apparatus for generating a long range reverse synthetic aperture radar image. More particularly, the present invention relates to a long range reverse synthetic aperture radar image using a polynomial and a Gaussian basis function that can improve the accuracy of distance alignment within a short time. It relates to a method and an apparatus for generating.

Recently, a technology for an inverse synthetic aperture radar (hereinafter referred to as "ISAR") imaging system for obtaining a two-dimensional high resolution image of a moving target has been widely used. The down-range resolution of the ISAR image is determined by the bandwidth of the radar signal, and the cross-range resolution is determined by the relative rotational movement of the target and the radar. Due to the two-dimensional nature of these ISAR images, they are applied for many military purposes.

Radar images can be used to identify targets over long distances, a technique known as non-cooperative target recognition (NCTR) or automatic target recognition (ATR). For this NCTR or ATR, a high resolution ISAR image must be obtained, but when the target moves, a phase error occurs due to the blurring of the radar image. The most commonly used method for obtaining ISAR images is the range-Doppler algorithm, and the most important step in the range-Doppler algorithm is the translational movement between pulses for chirp radar systems and bursts for stepped-frequency radar systems. compensation for motion).

The fluctuation of the target object generally consists of translational and rotational motions, especially when the compensation for the translational motion is inadequate, the same scatterers range. in different positions on the profile, which causes the ISAR image to become very blurred.

Compensation for translational motion is largely divided into two stages: range alignment and phase adjustment. Distance alignment is a method of aligning signals reflected from the same scattering source so as to be located at the same position in different distance side views, and phase adjustment is a method of compensating a phase error due to a forced distance alignment.

In the case of the distance alignment algorithm according to the prior art, since most targets are measured at a short distance, since the number of distance components (hereinafter referred to as "range bin") is small due to the short flight trajectory, the alignment time is short and the flight trajectory is short. This one can be modeled as a polynomial, so the alignment result is good.

However, during the actual image formation, the target is detected at a long distance, so the flight trajectory of the target is long, and it is very difficult to express the flight trajectory in one polynomial since the flight trajectory is significantly affected by the target's initial position and motion variables. . Therefore, when the distance alignment algorithm according to the prior art is used, a problem arises in that the distance alignment accuracy is degraded or the distance alignment calculation time is greatly increased.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method and apparatus for generating a long range reverse synthetic aperture radar image capable of rapid and accurate distance alignment with respect to a target at a long distance.

According to an embodiment of the present invention for solving this problem, in the method for generating a reverse synthetic aperture radar image for a target in flight, receiving a reflected wave scattered and reflected from the target, and extracted from the reflected wave Generating an estimated trajectory of the target using a distance side view, modeling the estimated trajectory as an orbital function combined with a polynomial and a Gaussian function, obtaining a center of gravity trajectory of the distance side view, the estimated trajectory and the weight Setting a parameter value constituting the trajectory function so as to minimize an error between center trajectories, adjusting the parameter value so that the cost function of the distance side view is maximum, and distance-aligning the distance-aligned distance side view Compensate for the phase error and inversely synthesize the distance profile Generating an aperture radar image.

The distance side view includes a distance component according to an observation angle of the radar.

The orbital function is represented by the following equation.

Where P (x) is the estimated orbital function, x is the respective distance profile, p _i is the polynomial coefficient, a _i is the Gaussian coefficient, b _i is the mean value of the distance profile, c _i is the standard deviation, L is the order of the polynomial, and G + 1 represents the number of Gaussian functions.

The center of gravity of the distance side view is represented by the following equation.

Where COM _m is the center of gravity of the m th distance side view, G _m (n) represents the m th distance side view, and N represents the number of range bins.

In the step of adjusting the parameter values to maximize the cost function of the distance profile, the parameter values are adjusted through a particle swarm optimization (PSO) algorithm, and the PSO algorithm is represented by the following equation.

Here, t is the number of generations, φ is an inertia weight, and r _i is a random variable having a uniform distribution between 0 and 1.

The cost function is represented by the following equation.

C is the cost function, G _k is the kth distance profile, and M and N are the number of distance profiles and the number of range bins of each distance profile, respectively.

According to another embodiment of the present invention, in the apparatus for generating a reverse synthetic aperture radar image for a target in flight, receiving a reflected wave scattered and reflected from the target, and from the distance side view extracted from the reflected wave of the target A distance side view generator for generating an estimated trajectory, and modeling the estimated trajectory as an orbital function combining a polynomial and a Gaussian function, and configuring the trajectory function to minimize an error between the estimated trajectory and the center of gravity trajectory of the distance side view. After setting a parameter value, a distance alignment unit for adjusting the parameter value to maximize the cost function of the distance side view, a phase adjuster for compensating a phase error of the distance-aligned distance side view, and an inverse using the distance side view And an image generator for generating a composite aperture radar image.

As described above, according to the present invention, the orbital function combined with the polynomial and the Gaussian function is primarily modeled so that the difference between the center of gravity and the estimated trajectory is minimized, and is modeled secondarily through the PSO algorithm, thereby more accurately and precisely. Distance alignment can be performed. In particular, more accurate ISAR images can be generated for trajectories in flight over long distances.

DETAILED DESCRIPTION Embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention.

First, an ISAR image will be described with reference to FIG. 1. FIG. 1 is an exemplary diagram illustrating movement of a target for generating an ISAR image, and FIG. 2 is a diagram illustrating a scattering source generated from a target.

As shown in FIG. 1, the change in the observation angle caused by the relative movement of the target 100 away from the radar 200 is used to form an ISAR image. That is, the radar signal according to the change in the observation angle between the radar 200 and the target 100 is transmitted to the image generating apparatus 300, and the image generating apparatus 300 generates an ISAR image based on the received radar signal. . Here, when the radar 200 emits a radar beam toward the target 100, the target 100 captured by the radar beam scatters the radar signal into a plurality of reflected waves. In this case, the target 100 captured by the radar beam is represented as an isotropic point scatterer on the 3D coordinate side as shown in FIG. 2.

Before describing the present invention, a Doppler algorithm for functionalizing the radar signal will be described. First, in the radar signal, the monostatic chirp waveform is most widely used because it has a high distance resolution. The transmitted chirp signal may be expressed as Equation 1 below.

일 경우 1이고, 그 밖의 범위에서는 0을 갖는 함수이다. Where r (t) is the signal transmitted at time t, A ₀ is the amplitude of the signal, and f ₀ Is the starting frequency and B is the bandwidth. τ is the pulse width, rect is

If 1, it is a function with 0 in other ranges.

Here, the received signal reflected from the scattering source may be represented by Equation 2 below.

A is the amplitude of the reflected signal and d ₀ is the delay time between the radar and the scattering source. In the case of a plurality of targets, the received signal g (t) can be expressed by Equation 3 below.

Where d _{k , n} represents the time delay between the radar and the nth scattering source of the target, N is the number of scattering circles, and K is the number of targets. The value of d _{k , n} can be calculated using the planar waveform approximation. In the Doppler (Range-Doppler) algorithm, the reflected signal is represented as a distance profile at a particular lateral angle, undergoes translational compensation, including distance alignment and phase adjustment, and then each distance component (hereinafter referred to as "range bin"). Fast Fourier transform (FFT) is applied, a detailed description thereof will be described with reference to FIG.

3 is a flowchart of a Doppler algorithm for generating an ISAR image. The Doppler (Range-Doppler) algorithm assumes that the relative angle of rotation between the target and radar is small and that scattering circles do not move to adjacent range bins for small angles of rotation. When the rotation angle is large, the interpolation is uniformly performed in the (frequency, angle) region by using a polar format, and a technique for obtaining a high resolution image is one of two-dimensional spectral estimation techniques. Multiple signal classification (MUSIC) techniques and the like can be used.

As shown in FIG. 3, the Doppler algorithm is composed of three stages: a range compression step S310, a translational motion compensation step S320, and an azimuth compression step S350.

The distance compression step S310 is a process of forming a distance side view representing the distribution of scattering circles at each observation angle. In this step, matched filtering using a wideband chirp signal or an inverse discrete Fourier transform (IDFT) using a stepped frequency signal is used. Distance resolution is determined by the bandwidth of the radar signal, expressed as c / 2B to be. Where c is the speed of light and B is the bandwidth. In the distance side view shown in the right end of FIG. 3, the reflected waves reflected from the scattering sources at a particular viewing angle may be represented by a plurality of range bins.

The translational motion compensation step S320 is a process of compensating motion between pulses in the case of a chirp radar and bursts in the case of a stepped radar, and is divided into a distance alignment step S330 and a phase adjustment step S340.

Finally, the vertical compression (S350) is a process of arranging the scattering circles in the vertical direction by using the difference in the Doppler frequencies of the scattering sources. Vertical resolution is affected by the viewing angle and is expressed as λ / 2Δθ. Where Δθ is the observation angle and λ is the wavelength. In the absence of translational compensation, the ISAR image is just a two-dimensional Fourier transform, and if the movement of the distance profiles due to the target's motion is not compensated for, the ISAR image will be seriously blurred. It demonstrates in more detail.

As shown in Figure 3, the translational motion compensation step (S320) is divided into a distance alignment step (S330) and a phase adjustment step (S340), the range alignment step (S330) is a step of aligning the distance side view , Phase adjustment step S340 is a step for compensating for the relative phase error of the distance-sided views that are forcibly aligned.

If the target is stationary and only rotates, the signals reflected from the same scattering source will be at a fixed distance on the distance profile. However, because the target can move several range bins between pulses, the reflected signals will generally be in different locations at different distance profiles. Therefore, the distance side views must be aligned so that the signals reflected from the same scattering source are located at the same distance in all distance side views. However, when the distance side views are moved without any phase compensation, a phase error is generated. Therefore, the phase error should be compensated for each distance side view.

If the target's kinematic parameters are known correctly, translational compensation can be performed directly. However, since it is impossible to know these variables accurately, the relative travel distance and phase error for distance alignment cannot be accurately calculated. For this reason, techniques for distance alignment and phase adjustment are widely used without using motion variables.

In the distance sorting step S330, a cost function indicating similarity of distance side views, such as correlation and entropy, is widely used. Among many cost functions, the one-dimensional entropy cost function is known to be very efficient and robust to noise. This cost function (H _{Gm , Gm +1} ) is given by Equation 4 below.

In Equation 4, G _m (n) and G _{m +1} (n) represent m-th and (m + 1) -th distance profiles in the n-th range bin, N _p represents the total number of range bins, and σ Indicates the number of movements for distance alignment. In general, G _m (n) to minimize errors in distance alignment Instead, the mean of the first to mth distance profiles is used. According to Equation 1, alignment is performed by moving the (m + 1) th distance side view by the number of shifts σ minimizing the one-dimensional entropy cost functions H _{Gm and Gm +1} .

In the phase adjusting step S340, methods such as a maximum contrast method and a minimum entropy method are used. These techniques, like the distance alignment technique, are also applied without information on the motion of the target, and can be easily implemented by those skilled in the art, and thus detailed descriptions are omitted.

On the other hand, in the distance alignment step (S330), since the number of movements σ to minimize the one-dimensional cost function shown in Equation 4 should be an integer, the accuracy of the distance alignment may be lowered and an error may occur. Also, as mentioned above, the distance alignment techniques shown in Equations 1 to 4 are performed under the assumption that the distance between the radar and the target is close, so that when the target is located at a long distance, the same vertical resolution as the distance resolution is achieved. The length of flight trajectory to obtain

Hereinafter, an apparatus and method for generating an ISAR image according to an exemplary embodiment of the present invention, which can perform distance alignment with respect to a long distance target, will be described with reference to FIGS. 4 and 5.

4 illustrates a configuration of an ISAR image generating apparatus according to an embodiment of the present invention. As shown in FIG. 4, an ISAR image generating apparatus 300 according to an exemplary embodiment of the present invention includes a distance side view generator 310, a translational motion compensator 320, and an image generator 330.

The distance side view generator 310 receives the reflected waves scattered and reflected from the target, and generates a distance side view indicating the distribution of scattering circles. The distance side view includes a range bin for the target according to the viewing angle of the radar. In addition, the distance side view generation unit 310 generates an estimated trajectory of the target in flight from the distance side view.

The translation compensation unit 320 compensates for the translation of the target, and includes a distance alignment unit 322 and a phase adjuster 324.

The distance alignment unit 322 models the estimated trajectory as an orbital function combining a polynomial and a Gaussian function, and sets parameter values constituting the trajectory function to minimize an error between the estimated trajectory and the center of gravity trajectory of the distance side view. . Then, the parameter value is adjusted to maximize the cost function of the distance profile.

The phase adjuster 324 compensates for the phase error of the distance-aligned distance side view. The image generator 330 arranges scattering circles in a vertical direction, and generates an ISAR image using a distance side view.

Hereinafter, a distance alignment method according to an embodiment of the present invention will be described. 5 is a flowchart illustrating a distance alignment method for a moving target according to an exemplary embodiment of the present invention, and FIG. 6 illustrates a distance side view trajectory generated for a moving target. 7 is a graph showing the center of gravity trajectory of the first ordered orbital function and the distance side view according to the embodiment of the present invention.

First, the distance alignment unit 322 generates a distance side view indicating a distribution of scattering circles at each observation angle with respect to a target in flight at a long distance. That is, the distance alignment unit 322 receives the reflected wave scattered and reflected from the target, and generates an estimated trajectory for the target as shown in FIG. 6 using the distance side view extracted from the reflected wave (S510).

The distance side view trajectory shown in FIG. 6 shows the magnitude of the reflected wave reflected back after the radar fires a radar pulse (or laser beam) to the target. In FIG. Shows the number of radar pulses fired toward the target (assuming 128 pulses in FIG. 6). In addition, the horizontal axis Down-Range Bin in Figure 6 represents the time it takes for the radar to receive the reflected wave after firing the radar pulse, which corresponds to the distance between the radar and the target. Therefore, the distance side view trajectory shown in FIG. 6 corresponds to the flight trajectory of the target.

In addition, the distance alignment unit 322 models the flight trajectory through the orbital function shown in Equation 5 to track the flight trajectory of the target (S520). Equation 5 is an equation for an orbital function for estimating a flight trajectory, which is estimated through M distance profiles having N range bins. As shown in Equation 5, the orbital function P (x) is represented by a combination of a polynomial and a Gaussian function.

In Equation 5, P (x) is an estimated orbital function, and x is each of the distance profile values in the range of 0 to M-1 and increases by one. In addition, p _i is a polynomial coefficient, a _i , b _i , c _i are Gaussian related parameters, L is the order of the polynomial, and G + 1 represents the number of Gaussian functions. In particular, a _i is a Gaussian coefficient, b _i is the mean value of the distance profile, and c _i represents the standard deviation.

Here, according to the prior art, the right end of Equation 5 is made of a polynomial corresponding to the first term, but the polynomial alone is incorrect to apply to a target in flight at a long distance. Therefore, according to the embodiment of the present invention, by adding a Gaussian function corresponding to the second term to Equation 5, the error between the actual flight trajectory and the polynomial is reduced.

If the target in flight is far away, several calculations are required to optimize the estimated orbital function shown in Equation 5, so it is important to determine the initial value for each parameter.

The distance alignment unit 322 obtains the center of gravity of each distance side view as shown in Equation 6, and generates a center of gravity trajectory (S530). Here, the center of gravity of the distance side view represents the center of gravity of the plurality of range bins shown in the right end of FIG. 3, and in general, the range bin of the center portion having the largest value is relatively It is close to the center of gravity. By connecting such a center of gravity, it is possible to generate the center of gravity trajectory as shown in FIG.

The center of gravity (COM _m ) of the m-th distance profile can be expressed by Equation 6 below.

Here, G _m (n) represents the m-th distance side view, and N represents the number of range bins.

Since most values except for the target portion on the distance side view are near zero, the center of gravity of Equation 6 is located at the target portion. Therefore, according to the embodiment of the present invention it is possible to effectively track the flight trajectory by using the center of gravity of each distance side view.

The distance alignment unit 322 sets an initial parameter to minimize an error between the center of gravity COM _m of the generated distance side view and the estimated orbital function P (x) (S540).

Here, if the value of the orbital function (P (x)) for each distance side view x is y and the center of gravity function of the distance side view is f (x, V), the center of gravity (COM _m ) of the distance side view is estimated and An error between the orbital functions P (x) may be expressed as in Equation 7 below.

Here, E is an error function, and V represents a vector of the parameters p _i , a _i , b _i , c _i shown in equation (5).

In order to obtain a parameter value for minimizing the error function E, Equation 7 can be expressed as Equation 8 below.

Here, V _j is the j-th component of the parameter vector (V), and the parameter is repeated to obtain a parameter that minimizes the error function (E) by repeating the following equation (9).

ΔV _j is a motion vector derived from a Gauss-Newton algorithm shown in equation (10).

Here, e is an error function vector, and J _e represents a Jacobian matrix of _e for each parameter (V).

Equations 8 to 10 are repeated until the error function E becomes minimum, and the parameter values when the error function E becomes minimum are initial parameters p _i , a _i , b _i , c _i Set to). A process of obtaining an initial parameter value through Equations 8 to 10 may be easily performed by those skilled in the art, and thus a detailed description thereof will be omitted.

In this way, if the initial parameter is set to minimize the error between the center of gravity (Com _m ) and the estimated orbital function (P (x)) of the distance side view, the orbital function (P (x)) as shown in Figure 7 It is set to match the center of gravity (COM _m ) trajectory as much as possible. In other words, the trajectory function P (x) is first ordered so as to substantially match the center of gravity COM _m trajectory of the distance profile by setting the initial parameter.

In addition, the distance alignment unit 322 adjusts the parameter to maximize the cost function (S550). Here, the distance alignment unit 322 may be solved through a genetic algorithm (GA) or particle swarm optimization (PSO). In the present invention, the PSO has been proven to be relatively easy to implement and to be effective in various engineering problems.

PSO is a population-based probabilistic optimization technique that initializes entities called particles to random values and then velocity towards the local best and global best of each particle. Change to find the minimum and maximum cost functions and the entity providing them. The method of updating each particle is as shown in Equation 11 below.

)는 i번째 particle(x _i (t))에 더해져서 이 particle을 이동시킨다.Where t is the number of generations, φ is the inertia weight, ρ ₁ = r ₁ c ₁ , ρ ₂ = r ₂ c ₂ , c ₁ , c ₂ > 1, c ₁ + c ₂ <4, r _i Is a random variable with a uniform distribution between 0 and 1. The velocity vector obtained from Equation 6 is the velocity vector

) Is added to the i th particle ( x _i (t) ) to move this particle.

In an embodiment of the present invention, after setting all element values of each particle to a random value between -1 and 1 using the PSO, particles that minimize the cost function are found. The cost function C is defined as in Equation 12 below, and represents the energy of the sum of the aligned total distance profiles. In the embodiment of the present invention, the coefficients of the polynomial maximizing the cost function can be found through the PSO.

G _k is the kth distance side view, M and N are the numbers of the distance side views and the number of range bins of each distance side view, respectively. It is proved that when the distance profiles are arranged such that Equation 12 is maximized, the contrast of the aligned distance profiles is maximized. In addition, when the energy is maximized, the cost function shown in the following equation (13) obtained by Fourier transform of each distance profile has a maximum value.

A _m is the Fourier transform of the m th distance profiles, and n _m is the shift required to align the th distance profiles. M and N are the number of distance profiles and the number of range bins for each distance profile, respectively.

As described above, according to the exemplary embodiment of the present invention, by adjusting the values of the parameters p _i , a _i , b _i , c _i from the initial values so that the cost function is maximized using the PSO algorithm, the estimated orbital function P (x The distance alignment can be performed more accurately so that)) more closely matches the actual flight trajectory of the target during long distance flight.

8 shows a cost function according to the number of households according to an embodiment of the present invention. As shown in FIG. 8, the cost function remains constant even if the number of households increases, and the value at this time becomes the maximum value. Therefore, the values of the parameters p _i , a _i , b _i , and c _i that maximize the cost function are obtained through the PSO algorithm.

9A is a graph illustrating a quadratic ordered orbital function in accordance with an embodiment of the present invention. That is, when the first ordered orbital function P (x) is second ordered through the PSO algorithm using the center of gravity of the distance side view, it shows the trajectory as shown in FIG. 9A.

9B is a graph showing the difference between the quadratic aligned trajectory function and the center of gravity trajectory. That is, according to the embodiment of the present invention, since the orbital function P (x) is first-order aligned by tracking the center of gravity of the distance side view, the graph shown in FIG. 9B is the first ordered orbital function P (x). It can be regarded as the difference of the quadratic aligned orbital function P (x).

As described above, according to the exemplary embodiment of the present invention, the initial value of the parameter is approximately obtained by using the center of gravity of the distance side view, and the optimized parameter value is generated through the PSO algorithm. By calculating the parameter values through the two steps as described above, an orbital function P (x) closer to the actual flight trajectory can be obtained even in a long distance flight.

FIG. 10A illustrates a result of performing distance alignment according to the prior art, and FIG. 10B is an exemplary diagram illustrating an ISAR image generated using FIG. 10A. FIG. 11A illustrates a result of performing distance alignment according to an embodiment of the present invention, and FIG. 11B is an exemplary view showing an ISAR image generated using FIG. 11A.

According to the related art of adjusting a parameter using only a polynomial, as shown in FIG. 10A, distance alignment is not performed smoothly, and therefore, as shown in FIG. 10B, the ISAR image is not clearly generated. In particular, long-range targets do not produce accurate ISAR images because distance alignment is not performed correctly.

However, according to the present invention, which adjusts parameters using a polynomial and a Gaussian function, distance alignment is performed smoothly as shown in FIG. 11A, and thus, the ISAR image is clearly displayed as shown in FIG. 11B. In particular, even if the target is located over a long distance, distance alignment can be precisely and precisely performed, so that the ISAR image is also sharp and well focused.

As described above, according to the exemplary embodiment of the present invention, the orbital function combined with the polynomial and the Gaussian function is primarily modeled so that the difference between the center of gravity and the estimated trajectory is minimized, and modeled secondarily through the PSO algorithm, thereby making it more accurate. Distance alignment can be performed in detail. In particular, more accurate ISAR images can be generated for trajectories in flight over long distances.

Although the preferred embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

1 is an exemplary diagram illustrating movement of a target for generating an ISAR image.

2 is a diagram showing a scattering source generated from a target.

3 is a flowchart of a Doppler algorithm for generating an ISAR image.

4 illustrates a configuration of an ISAR image generating apparatus according to an embodiment of the present invention.

5 is a flowchart illustrating a distance alignment method for a moving target according to an embodiment of the present invention.

Figure 6 shows the distance side view trajectory generated for the moving target.

7 is a graph showing the center of gravity trajectory of the first ordered orbital function and the distance side view according to an embodiment of the present invention.

8 shows a cost function according to the number of households according to an embodiment of the present invention.

9A is a graph illustrating a quadratic ordered orbital function in accordance with an embodiment of the present invention.

9B is a graph showing the difference between the quadratic aligned trajectory function and the center of gravity trajectory.

Figure 10a shows the result of performing the distance alignment according to the prior art.

FIG. 10B is an exemplary diagram illustrating an ISAR image generated using FIG. 10A.

11A illustrates a result of performing distance alignment according to an embodiment of the present invention.

FIG. 11B is an exemplary diagram illustrating an ISAR image generated using FIG. 11A.

Claims (11)

A method for generating a reverse synthetic aperture radar image for a flying target, Receiving a reflected wave scattered and reflected from the target, and generating an estimated trajectory of the target using a distance side view extracted from the reflected wave, Modeling the estimated trajectory as an orbital function combined with a polynomial and a Gaussian function, Obtaining a center of gravity trajectory of the distance side view, Setting a parameter value constituting the trajectory function to minimize the error between the estimated trajectory and the center of gravity trajectory, Adjusting the parameter values so that the cost function of the distance profile is maximized, and Compensating the phase error of the distance-aligned distance side view, and generating a reverse synthetic aperture radar image using the distance side view. The method of claim 1, And said distance side view comprises a distance component according to a viewing angle of said radar. The method of claim 1, The orbital function generates a reverse synthetic aperture radar image represented by the following equation:

Where P (x) is the estimated orbital function, x is the respective distance profile, p _i is the polynomial coefficient, a _i is the Gaussian coefficient, b _i is the mean value of the distance profile, c _i is the standard deviation, L is the order of the polynomial, and G + 1 represents the number of Gaussian functions. The method of claim 3, A method for generating a reverse synthetic aperture radar image in which the center of gravity of the distance side view is expressed by the following equation:

Where COM _m is the center of gravity of the m th distance side view, G _m (n) represents the m th distance side view, and N represents the number of range bins. The method of claim 4, wherein The step of adjusting distance by adjusting the parameter value such that the cost function of the distance side view is maximum, A method of adjusting the parameter value through a particle swarm optimization (PSO) algorithm, wherein the PSO algorithm generates a reverse synthetic aperture radar image represented by the following equation:

Here, t is the number of generations, φ is an inertia weight, and r _i is a random variable having a uniform distribution between 0 and 1. The method of claim 5, The cost function is To generate a reverse synthetic aperture radar image represented by the following equation:

C is the cost function, G _k is the kth distance profile, and M and N are the number of distance profiles and the number of range bins of each distance profile, respectively. An apparatus for generating a reverse synthetic aperture radar image for a flying target, A distance side view generation unit for receiving a reflected wave scattered and reflected from the target and generating an estimated trajectory of the target from the distance side view extracted from the reflected wave, The estimated trajectory is modeled as an orbital function combining a polynomial and a Gaussian function, a parameter value constituting the orbital function is set to minimize the error between the estimated trajectory and the center of gravity trajectory of the distance side view, and then the distance side view Distance alignment unit for adjusting the parameter value so that the cost function of the maximum, A phase adjuster for compensating for the phase error of the distance-aligned distance profile, and And an image generator for generating a reverse synthetic aperture radar image using the distance side view. The method of claim 7, wherein The orbital function is a device for generating a reverse synthetic aperture radar image represented by the following equation:

Where P (x) is the estimated orbital function, x is the respective distance profile, p _i is the polynomial coefficient, a _i is the Gaussian coefficient, b _i is the mean value of the distance profile, c _i is the standard deviation, L is the order of the polynomial, and G + 1 represents the number of Gaussian functions. The method of claim 8, An apparatus for generating a reverse synthetic aperture radar image, wherein the center of gravity of the distance side view is represented by the following equation:

Where COM _m is the center of gravity of the m th distance side view, G _m (n) represents the m th distance side view, and N represents the number of range bins. 10. The method of claim 9, The distance alignment unit, An apparatus for generating a reverse synthetic aperture radar image, wherein the parameter value is adjusted to maximize the cost function of the distance profile through a particle swarm optimization (PSO) algorithm.

Here, t is the number of generations, φ is an inertia weight, and r _i is a random variable having a uniform distribution between 0 and 1. The method of claim 10, The cost function is Apparatus for generating a reverse synthetic aperture radar image represented by the following equation:

C is the cost function, G _k is the kth distance profile, and M and N are the number of distance profiles and the number of range bins of each distance profile, respectively.

Images