JP2015219016A - Image radar device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem of a conventional image radar device that since an image reproduction algorithm for turning observation data into an image is premised on an assumption that a chirp rate uses a certain chirp signal as a pulse signal, the observation data in which the chirp rate changes in the middle cannot be turned directly into an image.SOLUTION: Reflection waves from a plurality of pulse signals from the object of observation that are transmitted at different time intervals are handled as observation data, and the observation data is converted to the waveforms that are equivalent to the reflection waves of a plurality of pulse signals from the object of observation that are transmitted at single time intervals. By applying an image reproduction algorithm to the waveforms generated by this conversion, it is possible to smoothly process image reproduction by using the observation data near boundaries at which a PRF and a chirp rate change.

Description

  The present invention relates to an image radar apparatus that images an observation object using radio waves or the like.

  The image radar device transmits a pulse signal, receives a part of the pulse signal reflected or scattered by the observation target as an echo signal, and performs a series of processes for obtaining a range profile from the received signal between the observation target and the radar. Repeat while changing the relative positional relationship. By synthesizing the time history of the range profile obtained in this way in consideration of the relative positional relationship, the reflection intensity distribution of the radio wave on the observation target is obtained as an image.

  The time history of the range profile is generally given as a two-dimensional distribution. One axis is an elapsed time [s] (hereinafter referred to as fast time) after transmission of a pulse signal, or a range [m] multiplied by 1/2 of the radio wave speed. A range profile is generated. The other axis is the time of observation (hereinafter referred to as slow time) performed while changing the relative positional relationship. A typical aperture radar that obtains a radar image is a synthetic aperture radar. Further, the frequency at which the pulse signal is repeatedly transmitted is called a pulse repetition frequency (PRF: Pulse Repetition Frequency).

In the satellite-borne synthetic aperture radar, the satellite altitude changes depending on the observation position on the earth. Therefore, when performing observation over a long period of time, it is necessary to change the PRF setting during the observation. If the PRF changes during the observation, the pulse signal reception interval is not constant. As a result, the image reproduction processing of the observation data on the premise that the reception interval of the pulse signal is constant does not function, and the image quality is deteriorated due to the resolution deterioration or the side lobe increase. Therefore, before performing the azimuth signal processing that is a component of the image reproduction processing, the conventional image radar apparatus resamples the observation data in the azimuth direction, and performs processing in which the observation data is equally spaced (for example, Patent Document 1).

Japanese Patent No. 5411506

In a conventional image radar apparatus or a synthetic aperture radar that is a representative image radar apparatus, an image reproduction algorithm for imaging observation data is based on the premise that a chirp signal is used as a pulse signal. Yes. Here, the chirp plate represents a change in the transmission frequency per unit time of the chirp signal. Therefore, the observation data in which the chirp plate changes during the process cannot be directly imaged. Therefore, the conventional image radar apparatus uses the position where the chirp plate changes as a boundary and reproduces the observation data separately before and after the boundary. In this case, there is a problem in that imaging near the boundary where the chirp plate changes cannot be performed.

  In the image radar apparatus or the synthetic aperture radar, in order to perform efficient operation while maintaining the conditions of constant peak transmission power, constant average transmission power, and constant transmission duty given by pulse width × PRF, PRF It is necessary to change the pulse width of the pulse signal when. When using a chirp signal as the pulse signal, it is necessary to change the chirp plate in order to change the pulse width. However, the conventional image radar device or synthetic aperture radar assumes a constant chirp plate, so the pulse width There was a problem that could not cope with the change of.

  The present invention has been made to solve the above-described problems. Even when the PRF or the pulse width is changed during the observation, the image radar apparatus can smoothly perform imaging near the boundary where the change occurs. The purpose is to give.

  An image radar apparatus according to the present invention includes a data storage unit that stores data indicating reflected waves from an observation target of a plurality of first signals transmitted by different PRFs, and a reflected wave stored in the data storage unit. And a data conversion unit that converts data to be shown into data equivalent to data indicating reflected waves from the observation target of the second plurality of signals transmitted by a single PRF.

  The image radar apparatus according to the present invention can provide an image radar apparatus that can smoothly image the vicinity of the boundary where the change occurs even when the PRF changes during observation.

1 is a block configuration diagram showing an image radar apparatus 1 according to Embodiment 1 of the present invention. 3 is a flowchart showing the operation of the image radar apparatus 1 according to the first embodiment of the present invention. Observation data during processing of the image radar apparatus 1 according to the first embodiment of the present invention. The block block diagram which shows the image radar apparatus 1 of Embodiment 2 of this invention. The flowchart which shows operation | movement of the image radar apparatus 1 of Embodiment 2 of this invention.

Embodiment 1 FIG.

  In this embodiment, a configuration of an image radar device that converts a reflected wave from an observation target of a pulse signal transmitted at different time intervals into a waveform equivalent to a reflected wave of a pulse signal transmitted at a single time interval. Disclose.

  1 is a block diagram showing an image radar apparatus 1 according to Embodiment 1 of the present invention. In each figure, the same numerals indicate the same or corresponding parts. In FIG. 1, an image radar apparatus 1 includes an observation data storage unit 2 that is a data storage unit that stores observation data observed by a synthetic aperture radar, and a data conversion unit 3 that converts the observation data. The observation data storage unit 2 stores reflected waves from the observation target of pulse signals transmitted at different time intervals as observation data. The data conversion unit 3 converts the observation data stored in the observation data storage unit 2 into a waveform equivalent to the reflected wave from the observation target of the pulse signal transmitted at a single time interval.

  The data conversion unit 3 converts the observation data stored in the observation data storage unit 2 into a pulse-compressed signal and outputs it. The matched filtering unit 4 outputs the time length of the pulse-compressed signal output from the matched filtering unit 4. A waveform of a reflected wave assuming a pulse signal transmitted at a single time interval is generated from an equalization processing unit 5 to be adjusted and a pulse-compressed signal whose time length is adjusted as an output of the equalization processing unit 5 A demodulation processing unit 6 and a post-conversion data storage unit 9 are provided.

  The pulse replica storage unit 7 stores a replica of the pulse signal transmitted from the image radar apparatus 1. Here, the pulse replica stores the pulse waveform of the pulse signal transmitted every time the pulse signal is transmitted or every time the PRF is changed. The demodulation reference signal generator 8 generates an ideal pulse signal assumed by data corresponding to observation of a pulse signal transmitted at a constant PRF, that is, a single time interval, obtained as a final output.

  Next, the configuration within the matched filtering unit 4 will be described. The matched filtering unit 4 performs processing for converting the observation data stored in the observation data storage unit 2 into a pulse-compressed signal. In the present embodiment, a configuration is shown in which this processing is performed in the frequency domain. The Fourier transform A unit 10 performs a Fourier transform in the fast time direction of the observation data stored in the observation data storage unit 2. The power normalization unit 20 normalizes the power of the replica of the pulse signal stored in the pulse replica storage unit 7. The Fourier transform B unit 30 performs a Fourier transform in the fast time direction of the pulse replica whose power is normalized by the power normalization unit 20. The matched filter processing unit 50 performs matched filter processing from the output of the Fourier transform A unit 10 and the output of the Fourier transform B unit 30. The inverse Fourier transform unit 60 performs an inverse Fourier transform on the output of the matched filter processing unit 50 in the fast time direction. That is, in the matched filtering unit 4, the input signal is converted into the frequency domain by the Fourier transform A unit 10 and the Fourier transform B unit 30, and after the matched filter processing is performed by the matched filter processing unit 50, the inverse Fourier transform unit 60. Is done. As a result, a pulse-compressed signal is output.

  The configuration of the equalization processing unit 5 will be described. The reception time equalization processing unit 70 in the equalization processing unit 5 performs processing for making the reception time constant for the output of the inverse Fourier transform A unit 60.

  The configuration of the demodulation processing unit 6 will be described. The Fourier transform D unit 80 in the demodulation processing unit 6 performs inverse Fourier transform on the output of the reception time equalization processing unit 70 in the fast time direction. The Fourier transform C unit 40 performs a Fourier transform in the fast time direction of the output of the demodulation reference signal generation unit 8. The chirp demodulation processing unit 90 demodulates the chirp using the output of the Fourier transform D unit 80 and the output of the Fourier transform C unit 40. The inverse Fourier transform B unit 100 performs inverse Fourier transform on the output of the chirp demodulation processing unit 90 in the fast time direction. The azimuth direction resampling unit 110 applies azimuth direction resampling to the output of the inverse Fourier transform unit B unit 100.

  The post-conversion data storage unit 9 stores the output of the azimuth direction resampling unit 110.

In this description and the following description, the term “part” or “part” means a dedicated electronic circuit or element, but a computer equipped with a general-purpose central processing unit (CPU) has a predetermined value. You may make it comprise in the form of the computer program module which performs a process.

Next, the operation will be described with reference to FIG.

  First, in the matched filtering unit 4, the matched filter processing of the observation data stored in the observation data storage unit 2 and the pulse signal stored in the pulse replica storage unit 7 is performed in the frequency domain, and the observation data storage unit 2 A pulse-compressed signal corresponding to the observation data stored in is output. A specific operation will be described below.

  First, the Fourier transform A unit 10 reads the observation data stored in the observation data storage unit 2 (step ST010).

The characteristics of the observation data will be described with reference to the schematic diagram shown in FIG. In FIG. 3A, it is assumed that echo signals from five point scatterers located in the observation area are received. In FIG. 3A, as the PRF is changed during observation, 1 / PRF, which is the time interval in the slow time direction of the transmitted pulse, changes, and the width of the transmitted pulse also changes. ing. Further, the shape of the transmission pulse does not become an ideal rectangle due to the hardware characteristics of the transmitter but has distortion. As the pulse width is changed, the reception time is changed according to the observation width of the observation area and the transmission pulse width.

First, the Fourier transform A unit 10 performs Fourier transform on the observation data in the fast time direction (step ST015). Next, the power normalization unit 20 reads the pulse replica of the pulse signal from the pulse replica storage unit 7 (step ST020), and normalizes the pulse replica read by the power normalization unit 20 to be 1. (Step ST025). This normalization is performed by calculating the power of the stored pulse replica signal and dividing the pulse replica by the same power. Further, the Fourier transform B unit 30 performs Fourier transform on the output of the power normalization unit 20 in the fast time direction (step ST030).

The matched filter processing unit 50 performs matched filter processing using the output of the Fourier transform A unit 10 and the output of the Fourier transform B unit 30 (step ST050). This matched filter processing is given by equation (1).

ここで、ｆはレンジ周波数、

はマッチドフィルタ処理の出力、Ｓ_rec(f)は観測データをファストタイム方向にフーリエ変換した信号を示し、Ｓ_rep(f)は電力で規格化したパルスレプリカをファストタイム方向にフーリエ変換した信号を表す。

Where f is the range frequency,

Is the output of the matched filter processing, S _rec (f) is the signal obtained by Fourier transform of the observation data in the fast time direction, and S _rep (f) is the signal obtained by Fourier transforming the pulse replica normalized by power in the fast time direction. Represent.

As a result of the division by the spectrum of the pulse replica, the amplitude distortion of each echo signal in the observation data is corrected. Here, the distortion to be corrected is the distortion of the pulse replica of the pulse signal, and the secondary phase change due to chirp modulation is removed by this correction. The pulse replica uses a pulse signal having the same pulse width and chirp rate as each echo signal in the observation data in order to cope with the change of the PRF being observed. As a result, even when the PRF is changed, the secondary phase change due to the chirp modulation can be removed by using the same chirp plate pulse signal as each echo signal in the observation data. In addition, since the division is performed by the normalized pulse replica, the power of the observation data at the output of the matched filter processing is equivalent to that before the matched filter processing.

Here, the matched filter processing is shown by the expression (1), but the present invention is not limited to this, and any similar expression may be used. For example, if the distortion in the shape of a pulse to be transmitted does not matter, a general matched filter process using a complex conjugate of a replica as a matched filter may be used.

The inverse Fourier transform A unit 60 performs inverse Fourier transform on the output of the matched filter processing unit 50 in the fast time direction (step ST060).

This process will be described with reference to FIG. The output of the inverse Fourier transform A unit 60 is a pulse-compressed signal. Then, with the change of the PRF as a boundary, the reception time is different as described with reference to FIG.

Therefore, in the process of step ST070, the reception time is set to a predetermined fixed length regardless of the change of PRF. That is, the reception time equalization processing unit 70 in the matched filtering unit 4 changes the reception time as the PRF is changed. The reception time that changes due to the change in the PRF is set to a fixed length (step ST070). As a method of aligning the reception time to a fixed length, for example, by deleting a portion with a long reception time when the PRF is small, the reception time when the PRF is large may be aligned. Further, the reception time may be aligned by adding a reception time that is insufficient so that the time length of the reception time when the PRF is large becomes a fixed length longer than that. Since pulse compression corresponds to pulse integration processing, the amplitude after pulse compression differs with the change of PRF as a boundary.

In the demodulation processing unit 6, the chirp signal is demodulated with respect to the pulse-compressed signal having the same reception time. This process is performed from step ST080 to step ST120. Specifically, the Fourier transform D unit 80 Fourier transforms the output of the reception time equalization processing unit 70 in the fast time direction (step ST080). Next, demodulation reference signal generation section 8 generates a chirp demodulation reference signal (step ST090). The reference signal for this chirp demodulation is given by equation (2). The reference signal is standardized so that the power is 1.

ここで、Ｔ_uniは所定のパルス幅、tはファストタイム、Ｋ_uniは所定のチャープレートである。

Here, T _uni is a predetermined pulse width, t is a fast time, and K _uni is a predetermined char plate.

The Fourier transform C unit 40 Fourier-transforms the demodulation reference signal in the fast direction (step ST100). Here, a demodulation reference signal is generated in the time domain, and a reference signal for demodulation in the range frequency domain is obtained by Fourier transforming the reference signal. However, the present invention is not limited to this. For example, instead of the demodulation reference signal generation unit 8 and the Fourier transform C unit 40, a portion for generating a demodulation reference signal in the range frequency domain may be provided.

The chirp demodulation processing unit 60 multiplies the output of the Fourier transform D unit 80 by the output of the Fourier transform C unit 40 to perform a chirp demodulation process (step ST110). The chirp demodulation process is given by equation (3).

ここで、

はこの復調処理の出力であり、

はフーリエ変換Ｄ部８０の出力、

はフーリエ変換Ｃ部４０の出力である。

here,

Is the output of this demodulation process,

Is the output of the Fourier transform D unit 80,

Is the output of the Fourier transform C section 40.

  The inverse Fourier transform B unit 100 performs inverse Fourier transform on the output of the chirp demodulation processing unit 60 in the fast time direction (step ST120). The signal after the inverse Fourier transform is a signal having a uniform pulse width as shown in FIG. However, the amplitude has a different value at the PRF change.

  In FIG. 3C, the signal has a uniform pulse width, but the time interval and power of pulses in the azimuth direction are not constant. Therefore, in order to make the time interval and power of the pulses in the azimuth direction constant, the azimuth direction resampling unit 110 resamples the output of the inverse Fourier transform B unit 100 in the azimuth direction (step ST130). This resampling is performed so as not to change the power in the azimuth direction. As a result, as shown in FIG. 3 (d), the observation data after conversion corresponding to the observation data obtained by observation with a fixed PRF with the time interval and the amplitude of the pulses in the slow time direction, that is, the azimuth direction aligned. obtain.

  That is, the observation data can be converted into a waveform equivalent to the reflected wave from the observation target of the plurality of pulse signals transmitted at a single time interval by the conversion process in the post-conversion data storage unit 9. Then, the observation data after conversion is output to the converted data storage unit 9.

  As described above, the pulse width and the chirp plate are aligned in consideration of the change of the pulse width accompanying the change of the PRF in the observation of the synthetic radar with the constant peak transmission power and the constant average transmission power. Observation data equivalent to observation data with a constant chirp can be obtained by PRF.

In a conventional image radar device, when an image including a boundary where the chirp plate changes is to be imaged, the pulse reception interval is separately made equal by resampling before and after the change of the chirp plate as a boundary, and pulse compression is performed. Then, after removing the chirp in the range direction, the image is reproduced by performing the subsequent processing. Alternatively, after the pulse compression is first performed to remove the chirp in the range direction, the pulse reception interval is made equal by resampling, and then the image is reproduced by performing the subsequent processing. When performing such imaging, the image reproduction algorithm that can be used is limited to the algorithm that first performs pulse compression to remove the chirp and then performs other processing, and uses a chirp in the range direction such as a chirp scaling algorithm. Therefore, there is a problem that an algorithm for performing high-speed image reproduction cannot be used. In addition, since resampling is performed without considering the power change per pulse accompanying the change in PRF, power modulation occurs in the vicinity of the PRF change in the signal after the azimuth signal processing, and there is no effect on the reproduced image. There was a problem that natural luminance unevenness occurred.

  On the other hand, in the image radar apparatus of the present embodiment, the entire image including the vicinity of the boundary is reproduced by an arbitrary image reproduction algorithm without dividing the observation data near the boundary where the chirp plate is changed and performing image reproduction. Can be imaged.

Moreover, since the chirp plates of the observation data are aligned, a high-speed image reproduction algorithm using the chirp in the range direction can be used, and high-speed imaging can be performed.

  In addition, matched filter processing is performed by division of pulse replicas normalized by power to correct pulse amplitude modulation due to hardware characteristics, so the sidelobe asymmetry generated by pulse amplitude modulation is corrected. Thus, it is possible to eliminate image quality degradation after reproduction processing that occurs on an image.

  As described above, the image radar apparatus according to Embodiment 1 includes an observation data storage unit 2 that is a data storage unit that stores reflected waves from observation targets of a plurality of pulse signals transmitted at different time intervals, and observation data. A data conversion unit that converts observation data that is a reflected wave stored in the storage unit 2 into a waveform equivalent to a reflected wave from the observation target of a plurality of pulse signals transmitted at a single time interval; It is characterized by that.

  By using this configuration, even when the PRF or the pulse width is changed during observation, it is possible to provide an image radar apparatus that can smoothly image the vicinity of the boundary where the change occurs.

  Further, the plurality of signals transmitted at different time intervals are chirp signals. By using the chirp signal, the pulse compression process can be performed smoothly.

  The plurality of pulse signals transmitted at different time intervals have different time widths. Even in such a case, in the image radar device according to the first embodiment, pulse waveforms corresponding to a plurality of pulse signals having different time widths are stored in the pulse replica unit 4, and a plurality of time widths having different time widths are stored. A pulse signal can also be converted into a waveform equivalent to a reflected wave from an observation target of a plurality of pulse signals transmitted at a single time interval.

  In the image radar apparatus according to the first embodiment, a plurality of signals transmitted at different time intervals include chirp signals having different chirp plates, and a plurality of pulse signals transmitted at a single time interval are single. It is a chirp signal having one chirp plate. By using chirp signals having different chirp plates in this way, chirp signals having different time widths in the same transmission band can be used. In addition, the chirp signal pulse waveform with different time width can be converted into a waveform equivalent to the reflected wave from the observation target of multiple pulse signals transmitted at a single time interval, and the boundary for changing the chirp plate Images can be reproduced smoothly in the vicinity.

  The matched filter processing in the matched filtering unit 4 and the demodulation processing in the demodulation processing unit 6 are performed as processing in the frequency domain. By performing processing in the frequency domain, the filter processing can be changed by product operation for each frequency, and the operation can be performed efficiently.

  The image radar apparatus according to the first embodiment uses a reference signal standardized by matched filter processing or chirp demodulation processing so that the power of the original observation data is not changed. Accordingly, there is an advantage that power is not modulated in the vicinity of the PRF change, and unnatural luminance unevenness does not occur in the image after image reproduction.

Embodiment 2. FIG.

  In the first embodiment, the matched filter process and the chirp demodulation process are performed in the frequency domain, while the second embodiment shows a form in which these processes are performed in the time domain.

  FIG. 4 is a block diagram showing the second embodiment. The same parts as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  In FIG. 4, the time domain matched filter processing unit 51 performs matched filter processing in the time domain. The time domain chirp demodulation processor 91 demodulates the chirp signal in the time domain.

Next, the operation of the second embodiment will be described with reference to FIG. The same operations as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  In step ST011, the time domain matched filter processing unit 51 reads observation data from the observation data storage unit 2.

In step ST051, the time domain matched filter processing unit 51 performs time domain matched filter processing from the observation data read from the observation data storage unit 2 and the output of the power normalization unit 20. This matched filter processing is given by equation (4).

は時間領域マッチドフィルタの出力、ｓ_rec（ｔ）は観測データ、ｈ（ｔ）はマッチドフィルタである。マッチドフィルタは、式（５）で与えられる。 Where t is fast time, u is a variable in the fast time area,

Is the output of the time domain matched filter, s _rec (t) is the observation data, and h (t) is the matched filter. The matched filter is given by equation (5).

ここで、ｓ_reｐ（ｔ）はパルスレプリカの信号である。

Here, s _rep (t) is a pulse replica signal.

  In step ST111, the time-domain chirp signal processing unit 91 performs chirp demodulation processing using the output of the reception time equalization processing unit 70 and the output of the demodulation reference signal generation unit 8. This chirp signal demodulation process is given by equation (6).

As described above, in the second embodiment, the matched filter processing in the matched filtering unit 4 and the demodulation processing in the demodulation processing unit 6 are performed as processing in the time domain. In other words, matched filter processing and chirp demodulation processing are performed in the time domain without performing Fourier transform. As a result, the same effect as in the first embodiment can be obtained without requiring a part for Fourier transform.

  In the first and second embodiments, a signal called a pulse signal may be an arbitrary modulation signal other than the pulse signal. Therefore, the “pulse signal” may be simply called a signal. Further, “a plurality of pulse signals transmitted at different time intervals” may be referred to as “a first plurality of signals transmitted at different time intervals”. “A plurality of pulse signals transmitted at a single time interval” may be referred to as “a second plurality of signals transmitted at a single time interval”. The “reflected wave” may be called “data indicating the reflected wave”.

1: image radar device, 2: observation data storage unit, 3: data conversion unit, 4: matched filtering unit, 5: equalization processing unit, 6: demodulation processing unit, 7: pulse replica storage unit, 8: reference for demodulation Signal generation unit, 9: converted data storage unit, 10: Fourier transform A unit, 20: power normalization unit, 30: Fourier transform B unit, 40: Fourier transform C unit, 50: matched filter processing unit, 51: time Region matched filter processing unit, 60: inverse Fourier transform unit, 70: reception time equalization processing unit, 80: Fourier transform D unit, 90: chirp demodulation processing unit, 91: time domain chirp demodulation processing unit, 100: inverse Fourier transform Part B, 110: Azimuth direction resampling unit

Claims (7)

A data storage unit for storing data indicating reflected waves from the observation target of the first plurality of signals transmitted by different PRFs;
A data conversion unit that converts data indicating the reflected wave stored in the data storage unit into data equivalent to data indicating the reflected wave from the observation target of the second plurality of signals transmitted by a single PRF When,
An image radar apparatus comprising: The image radar apparatus according to claim 1, wherein the first plurality of signals are chirp signals. The image radar apparatus according to claim 1, wherein the first plurality of signals have different time widths. The first plurality of signals includes chirp signals having different chirp plates;
The image radar apparatus according to claim 1, wherein the second plurality of signals are chirp signals having a single chirp plate. The data converter is
A replica storage unit for storing the first plurality of signals;
A matched filtering unit that outputs a plurality of pulse-compressed signals obtained by performing a matched filter process between the data indicating the reflected wave stored in the data storage unit and the signal stored in the replica unit;
An equalization processing unit for aligning time widths of the plurality of pulse-compressed signals output from the matched filtering unit;
A demodulation processing unit that performs demodulation processing using data indicating a waveform of a signal transmitted by the single PRF;
The image radar apparatus according to claim 1, further comprising: The image radar device according to claim 5, wherein the matched filter processing in the matched filtering unit and the demodulation processing in the demodulation processing unit are performed as processing in a frequency domain. The image radar apparatus according to claim 5, wherein the matched filter processing in the matched filtering unit and the demodulation processing in the demodulation processing unit are performed as processing in a time domain.

Images