JP6391054B2 - Synthetic Aperture Radar System for Widening Observation Area and High Resolution Using Identifiers - Google Patents

Description

  The present invention relates to a synthetic aperture radar apparatus that identifies a stationary target or a moving target on the ground or sea using an observation signal from a synthetic aperture radar apparatus.

  There are already papers, patents, and other techniques for identifying a moving target using an observation signal from a synthetic aperture radar apparatus, and a plurality of proposals have been made for the apparatus and method.

  Non-Patent Document 1 proposes a PN sequence (Pseudorandom Noise) in order to identify a moving target using a satellite-borne synthetic aperture radar device. Intra-pulse modulation is alternately performed with two types of different PN sequences, and a pulse wave transmitted from a satellite is irradiated toward a plurality of point targets with a predetermined PRF (Pulse Repeat Frequency), and the received signal is a backscattered wave. Is demodulated with the same PN sequence as the transmission pulse wave. This demodulated component is aperture synthesized. The target is approximated as a point target, and the difference between the back and forth pulse wave demodulated signal components of the backscattered wave from the target shown in Equation (1) is wavelet transformed to identify the moving target and specify the radial speed of the moving target. Is possible.

ここで、（ｘ，ｙ，ｚ）＝（０，０，Ｈ）（Ｈ：衛星の高度）を起点として，送信開始から偶数個の受信パルス波の隣接異種ＰＮ系列による復調信号成分は，次式で与えられる。

ここで、ＰＮ（ｎ），ＰＮ’（ｎ）は，異種ＰＮ系列である。

Here, starting from (x, y, z) = (0, 0, H) (H: satellite altitude), the demodulated signal components of the even number of received pulse waves from the start of transmission by the adjacent different PN sequences are as follows: It is given by the formula.

Here, PN (n) and PN ′ (n) are heterogeneous PN sequences.

ここで、ａは、スケールであり、周波数軸上を掃引するパラメータである。ｂは、シフトであり、時間軸上を掃引するパラメータである。ψ（ｔ）としては、次式で与えられるガボール・ウェーブレット関数を採用している。

ここで，ω_０は，振動の中心周波数であり、搬送角周波数である。The wavelet transform is performed by using the wavelet function ψ (t) and the signal component of Equation (1) as shown in Equation (3).

Here, a is a scale and is a parameter that sweeps on the frequency axis. b is a shift and is a parameter that sweeps on the time axis. As ψ (t), a Gabor wavelet function given by the following equation is adopted.

Here, ω ₀ is the center frequency of vibration and the carrier angular frequency.

  In Non-Patent Document 2, the position and speed of a moving target composed of five point targets are specified based on the above-described wavelet transform theory. Expression (3) is a complex function, and a parameter a is determined from the amplitude information, and using this a, the target position, shape, and velocity are estimated from the phase information.

Takano Furuno, Hiromasa Ikuno “Synthetic Aperture Radar Applying PN Series,” Science Review (B), vol. J88-B, no. 7, pp. 1348-1358, July 2005. Takano Furuno, Hiromasa Ikuno, “Identification of Moving Targets Using Synthetic Aperture Radar Applying PN Sequence,” Science Review (B), vol. J89-B, no. 11, pp. 2143-2150, Nov. 2006.

  If the number of point targets is about 5 as shown in Non-Patent Document 2, the moving target is imaged by sweeping parameters a and b as shown in Non-Patent Document 2. However, it is difficult to image an aggregate composed of a large number of point targets. Also, the correspondence between the parameter a and the three-dimensional positions of a plurality of point targets including the velocity vector is not clear.

  The present invention has been made in order to solve the above-described problems, and an object of the present invention is to combine a plurality of point targets to enable imaging of a collection of point targets with high resolution. An aperture radar apparatus is obtained.

In order to solve the above-mentioned problem, the invention according to claim 1 is a synthetic aperture radar apparatus that demodulates a reflected wave obtained by irradiating the entire observation region with a transmission wave modulated with a predetermined modulation signal and performs wavelet transform. ,
The observation area is divided into a plurality of parts, each corresponding to an individual identifier,
The identifier is obtained by wavelet transforming a pseudo signal having phase information on the distance and velocity between each observation region and a radar, and differentiating the output to obtain a maximum point. Calculated, memorized,
Imaging processing is performed from all local maximum points and a set of point target luminance points of the observation signal subjected to the wavelet transform processing on the reflected wave of the radar wave irradiated to the observation region from the synthetic aperture radar device. It is characterized by that.
Among the identifiers calculated from the local maximum points obtained by differentiation after the wavelet transform process, the observation signal can be increased in resolution from the local maximum points.

大点であることを特徴とする合成開口レーダ装置。
前記ウェーブレット変換は、フーリエ変換と違って、時間軸に局在する波形を適切に検出できるという特徴を有している。According to a second aspect of the present invention, in the first aspect of the present invention, the local maximum point is ω0 as a center frequency of vibration of a wavelet function, a is a scale, 1 / a is x, and a radius

A synthetic aperture radar device characterized by being a large point.
Unlike the Fourier transform, the wavelet transform has a feature that a waveform localized on the time axis can be detected appropriately.

を作成し、前記観測領域を前記識別子に対応させることを特徴とする。
前記極大点と観測領域の点目標輝度点は、１対１に対応しており、前記観測領域を前記識別子によって構成することができる。The invention according to claim 3 is the invention according to claim 1 or 2, wherein the identifier is obtained by using the wavelet transform parameter a and the other wavelet transform parameter b.

And the observation area is made to correspond to the identifier.
The local maximum point and the point target luminance point of the observation area have a one-to-one correspondence, and the observation area can be configured by the identifier.

  According to a fourth aspect of the present invention, in the first to third aspects of the invention, the observation signal is imaged by performing a collation process between the all identifiers and the luminance point of the wavelet transform output as the observation signal. To do.

  According to a fifth aspect of the present invention, in the first to fourth aspects of the present invention, when performing the collation process, a noise component mixed in the wavelet transform output is calculated, and a threshold process for suppressing the noise component is also performed. It is characterized by carrying out.

  As described above, by applying the identifier defined in the present application to a target composed of a large number of point targets, it is possible to provide a synthetic aperture radar apparatus that realizes wide area and high resolution.

The figure which shows an example of one point target wavelet transform output of the difference process for 8 pulses. Design specifications of the synthetic aperture radar using the PN sequence used for the calculation. The figure which shows an example of the positional information on a point target, speed, and an identifier. The figure which shows an example of a receiver noise. The figure which shows a mode that the identifier at the time of adding noise changes.

  Hereinafter, embodiments of the present invention will be described by taking a satellite-borne synthetic aperture radar as an example.

似信号は，非特許文献１を参照すると次式で表示される。

The similar signal is displayed by the following formula when Non-Patent Document 1 is referred to.

ここで、ＰＮ（ｎ）は｛０｝と｛１｝の系列を｛−１｝と｛１｝に変換して得られる巡回符号系列であり、τ０はＰＮ波形のパルス時間幅、Ｔ（＝Ｎτ０）はコード長がＮのＰＮ系列で位相変調したパルス幅、ｍは、任意の整数（０，１・・）、Ｔ_０（１／ＰＲＦ）はパルス繰返し周期である。δ_ｉ，ｊはｉ＝ｊのとき１，ｉ≠ｊのとき０となる．ｉを次数とすると、ＰＮ系列のコード長は２^ｉ−１で与えられる。ｃは光

与えられる。

Here, PN (n) is a cyclic code sequence obtained by converting a sequence of {0} and {1} into {−1} and {1}, τ0 is a pulse time width of the PN waveform, and T (= Nτ0) is a pulse width obtained by phase modulation with a PN sequence having a code length of N, m is an arbitrary integer (0, 1...), And T ₀ (1 / PRF) is a pulse repetition period. δ _{i, j} is 1 when i = j and 0 when i ≠ j. If i is the order, the code length of the PN sequence is given by 2 ⁱ −1. c is light

Given.

度である。また、Ｈは、衛星の高度で、Ｖは衛星の速度であり、ｙ軸の正方向に飛行しているものとする。なお，以上述べたパルス波については、アンテナ放射パターンの効果は考慮に入れていない。

Degree. Further, H is the altitude of the satellite, V is the speed of the satellite, and it is assumed that it is flying in the positive direction of the y-axis. Note that the effect of the antenna radiation pattern is not taken into consideration for the pulse wave described above.

  The pseudo signal of Expression (5) can be demodulated by the same PN sequence as that of the transmission wave by referring to Non-Patent Document 1.

  Unlike the Fourier transform, the wavelet transform for the demodulated signal has a feature that a waveform localized on the time axis can be detected appropriately. Therefore, it is possible to detect a change due to a difference in position and speed of the point target. The wavelet transform is given by the inner product of the wavelet function ψ (t) and the signal component as shown in the following equation.

ここで、式（７）の信号成分は、前記ＰＮ系列による復調後の点目標関数の差分であり、次式で与えられる。

Here, the signal component of Equation (7) is the difference of the point target function after demodulation by the PN sequence, and is given by the following equation.

ここで、（ｘ，ｙ，ｚ）＝（０，０，Ｈ）（Ｈ：衛星の高度）を起点として、送信開始から偶数個の前記擬似信号の隣接異種ＰＮ系列による復調信号成分は、次式で与えられる。

Here, starting from (x, y, z) = (0, 0, H) (H: satellite altitude), the demodulated signal component of the even number of pseudo signals adjacent to the pseudo signal from the start of transmission is It is given by the formula.

ここで，ＰＮ（ｎ），ＰＮ’（ｎ）は，異種ＰＮ系列である． （８）

ラメータである。ｂは，シフトであり、時間軸上を掃引するパラメータである。ψ（ｔ）としては、次式で与えられるガボール・ウェーブレット関数を採用する。

Here, PN (n) and PN ′ (n) are heterogeneous PN sequences. (8)

It is a parameter. b is a shift and is a parameter that sweeps on the time axis. As ψ (t), a Gabor wavelet function given by the following equation is adopted.

ここで、ω_０は、振動の中心周波数であり、搬送角周波数とする。
点目標をｔ＝ｂに固定し、パルス列のうち，ｍ＝０としたときの応答として、式（７）のウェーブレット変換を具体的に表現すると次式で表される。

Here, ω ₀ is the center frequency of vibration and is the carrier angular frequency.
As a response when the point target is fixed at t = b and m = 0 in the pulse train, the wavelet transform of Equation (7) is specifically expressed by the following equation.

る点目標の位置と速度の情報を有している。パラメータ１／ａが大きくなり、ウェーブレット変換出力も大きくなることから、式（１１）は単調増加減少関数である。従って、式（１１）の極大点条件を求めるために、パラメータ１／ａで微分すると次式が得られる。

Information on the position and speed of the point target. Since the parameter 1 / a increases and the wavelet transform output also increases, equation (11) is a monotonically increasing / decreasing function. Therefore, in order to obtain the maximum point condition of Expression (11), the following expression is obtained by differentiating with the parameter 1 / a.

が近接し、前記半径が極小となる場合である。

Are close to each other and the radius is minimized.

In order to confirm the above-described result, the wavelet transform output of the point target of the 8-pulse difference process is shown in FIG. The wavelet transform output is shown as amplitude information on the (b, 1 / a) plane. In this figure, the maximum amplitude is b ₃₅ and 1 / a ₃₁ . This maximum amplitude point becomes the maximum point of the point target.

集合を識別子ＩＤ（ｂ，１／ａ）として、これを用いて前記点目標の位置と速度ベクトルを識別する。すなわち、前記識別子１とその要素を次式のように表す。Here, FIG. 2 shows the design parameters of the synthetic aperture radar to which the PN sequence used in the calculation is applied.

The set is used as an identifier ID (b, 1 / a) to identify the position and velocity vector of the point target. That is, the identifier 1 and its elements are expressed as follows:

Here, the total number of identifiable elements of the identifier 1 is M × N. The maximum value of the shift b shown in Equation (14) and Equation 1 is the pulse width T. Further, in order to improve the resolution, the scale 1 / a shown in the second formula of the equation (14) is such that the radius of the circle of the equation (12) does not overlap with the adjacent circle and is minimized as described above. To decide. This radius is referred to as a point radius. The point radius increases ω ₀ (related to resolution)

Position information including speed, which is additional information defined by the identifier 1 and the equation (6), is expressed as the following equation.
Here, the first term is position information and the second term is speed information. The unit of position coordinates and speed is m and m / s. Here, in the numerical example of the present application, T = 3 × 10 ⁻⁵ , 1 / a _N = 3 × 10 ⁷ , M = 60, and N = 60.

  The observation area distributed in the three-dimensional space is rectangular or square, and the set of point targets is in a stationary state (speed 0) in a steady state, and the observation area is divided using the point target as a circle. . Based on the position information that is additional information of the identifier 1 to the observation area, the identifier 1 is made to correspond to the dividing point. When the observation region is determined, collation processing is performed between the identifier 1 and the wavelet transform output, which is an observation signal, using the speed shown in Expression (15) as a parameter. The collation process can be realized by a difference process between the identifier 1 and the observed value. An example of the result of the collation process is shown in FIG. As is apparent from FIG. 3, the three point targets are separated and moved. In addition, by applying the wavelet transform using the Gabor wavelet function, as shown in the following equation, a resolution of 3.7 m calculated from the code length of the PN sequence from the design specifications of FIG. As shown in 16), it can be seen that the distance between the point targets can be identified up to 0.5 m in the x and y directions, and the resolution is increased.

  Actually, the correspondence between the identifier 1 and the observation signal becomes difficult due to the noise of the observation equipment or the error such as clutter existing in the space. Next, numerical calculation is performed assuming that the receiver noise 2 of the synthetic aperture radar gives an error to the bright spot of the observed signal. The wavelet transform when there is an error is expressed as the following equation by the same calculation method as in equation (11).

Here, n ₀ (t) is receiver noise 2, which is added to the demodulated observation signal component and wavelet transformed as shown in equation (17).
When the equation (17) is differentiated by the parameter 1 / a, the condition that the wavelet transform takes the maximum point when the observation signal includes noise is given by the following equation.

を座標軸として定義すると、

致する。以上から、前記観測信号に前記雑音２を含んだ場合でも、極大点条件は成立し、所定の１／ａで極大値をとることが分かる。更に、前記雑音２を含んでいない式（１２）と比較すると、円の中心も半径も異なり、雑音の量によっては、極大点となる１／ａの値は異なる。前記識別子１は、前記雑音２を含んでいないため、この点目標を別のものと見なすようになり、識別に対する誤差が発生する。前記雑音２を抑制するために、前記照合処理を実施する際に、閾値処理を併せておこなう。シミュレーション結果を以下に示す。

Is defined as a coordinate axis,

I agree. From the above, it can be seen that even when the observed signal includes the noise 2, the maximum point condition is satisfied and the maximum value is obtained at a predetermined 1 / a. Furthermore, compared to the equation (12) that does not include the noise 2, the center and radius of the circle are different, and the value of 1 / a that is the maximum point differs depending on the amount of noise. Since the identifier 1 does not include the noise 2, the point target is regarded as a different one, and an error in identification occurs. In order to suppress the noise 2, a threshold value process is also performed when the matching process is performed. The simulation results are shown below.

Receiver noise 2 is shown in FIG. 4 as white noise generated by random numbers. The mean and variance are 0.031 and 0.289, respectively. In Expression (18), when the received power amount is 1 and the noise 2 is added from −40 dB to −10 dB, the state in which the observation signal wavelet transform output changes is expressed as ID (b ₃₅ , 1 / a ₄₂ ) and the phase 3 calculated from the second equation of the equation (17) are collectively shown in FIG. As is apparent from the results, when a noise of −20 dB or more is added, the value of 1 / a changes due to the change in the condition of the maximum point due to the noise 2. Eventually, the noise addition amount for which the identifier 1 is effective is up to −30 dB in which the element ID (b ₃₅ , 1 / a ₄₂ ) of the identifier 1 is equivalent in this calculation example. Further, as shown in FIG. 5 , the phase amount change between −40 dB and −30 dB is 0.05 rad. Therefore, the threshold PHΔ is given from the information of the phase 3 of the identifier 1 and is set as the following equation.

Here, PH is a phase in an ideal state (no error). As described above, Δ is determined from the phase amount change in a range where the element of the identifier 1 does not change. Specifically, as described above, the maximum value of the phase amount change of the identifier 1 is Δ in a range in which the noise 2 is added to the phase of the ideal state and the identifier element does not change. In this calculation example, the threshold PHΔ is −1.07 rad. When the phase amount of the observed value deviates from the threshold value, the observed value is removed, and the matching process between the next identifier 1 and the observed value is performed.
Although the noise 2 is generated from the receiver, noise generated by weather such as clutter can be similarly evaluated.
The final image processing is obtained by mapping the result of the difference processing, that is, the moving component, on the observation area when the moving speed of the point target assembly is zero.

  As a result of adopting wavelet transform and identifier of Gabor wavelet function, the observation area was widened and the resolution was improved. This enables fine imaging by identifying point targets that move at minute speeds. The present invention can be applied to early detection of disasters such as earthquakes, debris flows, landslides, and tsunamis, and observation of crops and vegetation growth. The radar according to the present invention can be observed in real time by being mounted on a satellite. In particular, by mounting on a quasi-zenith satellite, it is possible to observe constantly as with a geostationary satellite.

1 ... identifier 2 ... noise 3 ... phase

Claims (5)

A synthetic aperture radar apparatus that demodulates a reflected wave obtained by irradiating the entire observation area with a transmission wave modulated with a predetermined modulation signal, and performs wavelet transform using a scale a and a shift b that are wavelet transform parameters ,
The observation area is divided into a plurality of parts, each of which corresponds to an individual identifier.The individual identifier is a wavelet transform of a pseudo signal having the phase information of the distance and speed between the observation area and the radar, The output is differentiated to obtain a local maximum point. The local maximum point is calculated and stored over the entire observation region, and the total local maximum point and the reflected wave of the radar wave irradiated to the observation region from the synthetic aperture radar device are calculated. On the other hand, a synthetic aperture radar apparatus characterized in that an imaging process is performed from an aggregate of point target luminance points of an observation signal subjected to the wavelet transform process. The synthetic aperture radar device according to claim 1,
The maximum point is the center frequency of vibration of the wavelet function ω0, a is a scale, b is a shift, M ₀ and N ₀ are positive integers, 1 / a is x, and the radius is

A synthetic aperture radar device characterized by being a point. The synthetic aperture radar device according to claim 1, wherein
An identifier using the wavelet transform parameter a and the shift b which is the other wavelet transform parameter

And the observation area is associated with the identifier. The synthetic aperture radar device according to claim 1, wherein
A synthetic aperture radar apparatus, wherein the observation signal is imaged by a collation process between the identifier and a luminance point of the wavelet transform output as an observation signal. The synthetic aperture radar device according to claim 4 ,
A synthetic aperture radar device characterized in that, when executing the matching process, a noise component mixed in the observation signal is calculated and a threshold process for suppressing the noise component is performed in parallel.

Images