JP4793240B2 - High resolution radar equipment - Google Patents

Description

  The present invention relates to a high-resolution radar mounted on an aircraft, a satellite, or the like, and more particularly to a synthetic aperture radar (SAR) apparatus for observing and imaging the ground surface and the sea surface.

  When imaging observation data in a synthetic aperture radar device that obtains high-resolution images of the ground surface and the sea surface, the average of the amplitude for each range bin is calculated, and a threshold is set based on the average amplitude in the range bin where this is the maximum. It has already been disclosed that processing can be performed only with a range bin having a small estimation error by performing processing using only a range bin having an amplitude of (for example, Patent Document 1). reference.).

JP 2003-215240 A (FIG. 1)

  In conventional synthetic aperture radar, when the error is estimated from the image, it is corrected by the autofocus method on the assumption that the error in the image is constant. The generated Doppler frequency change is used to decompose the signal to obtain a two-dimensional high-resolution image. However, since the platform position information is used when the signal is decomposed, a motion sensor is installed to measure the platform motion. However, since there is an observation error due to the motion sensor, the image is blurred due to this error, the resolution is deteriorated, and there is a problem that the blur of the image cannot be compensated.

This problem will be specifically described with reference to FIG.
FIG. 3 is a conceptual diagram of observation for explaining the problems to be solved by the present invention, and considers a case where a wide area is observed in the radio wave transmission direction (hereinafter, range direction) at a short distance.
In FIG. 3A, the difference between the trajectory (solid arrow) measured by the motion sensor and the actual trajectory (dotted arrow) becomes an error, resulting in blurring of the image.

Here, the synthetic aperture radar projects the platform trajectory onto a reference plane composed of a reference linear trajectory (P1P2) and an imaging range (N1N2, F1F2, etc.) shown in FIG. 3A, and reproduces the image. Process.
At this time, the trajectory difference between the trajectory shown in FIG. 3B (curve P1P2) and the trajectory shown in FIG. 3C (curve P1P2) is the same as the reference plane P1P2N2N1 shown in FIG. Since the reference plane (P1P2F2F1) shown in (c) is different, the error is different between the short distance side (P1P2N2N1) and the long distance side (P1P2F2F1) of the image shown in FIG. There was a problem that the image was not compensated for by the autofocus method because it was not satisfactory.

  The present invention has been made to solve the above-mentioned problems, and to obtain a high-resolution radar apparatus that can obtain a SAR image with good image quality without blurring of an image, which has not been obtained by a conventional method. It is an object.

  The high-resolution radar apparatus according to the present invention includes an image dividing unit that divides a SAR image into a plurality of ranges, a phase error estimating unit that estimates a phase error for each image divided by the image dividing unit, and the phase error estimating unit. The phase error slope estimator for estimating the slope of the phase error, which is a change with respect to the distance, from the phase error of each divided image estimated in step (b), and the phase error slope estimated by the phase error slope estimator for each range bin. And a phase error compensator that estimates and compensates for different phase errors.

  According to the present invention, after the SAR image is divided in the range direction, the phase error for each range is calculated and the image is corrected, and then the autofocus method is used. There is an effect that the obtained high resolution radar apparatus can be obtained.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a high-resolution radar apparatus according to Embodiment 1 of the present invention, in which 1 is a SAR sensor, 2 is a signal transmission / reception unit, 3 is an image reproduction processing unit, 4 is an image division unit, and 5 is a range bin selection. , 6 is an image shift unit, 7 is a window function multiplication unit, 8 is a phase gradient estimation unit, 9 is a phase error slope estimation unit, 10 is a phase error compensation unit, 11 is an iterative determination unit, and 12 is an image.

In FIG. 1, the SAR sensor 1 includes an antenna, a motion sensor, and the like, and is a part that is mounted on a mobile platform such as an aircraft or a satellite.
The SAR sensor 1 is configured to radiate a high-frequency pulse signal generated by the signal transmission / reception unit 2 to a space and receive a reflected signal of the high-frequency pulse signal, and a motion sensor that measures the movement of the mobile platform. Yes.

The signal transmission / reception unit 2 is a part that constitutes transmission / reception means for generating a high-frequency pulse signal and sending it to the SAR sensor 1 and amplifying the signal received by the SAR sensor 1.
The image reproduction processing unit 3 is a unit that reproduces a SAR image that is a two-dimensional high-resolution image from the signal received from the transmission / reception unit 2 and the motion information of the platform measured by the motion sensor.
Here, the two-dimensional image is composed of an azimuth axis and a range axis.

  The image dividing unit 4 is a part that divides the SAR image reproduced by the image reproduction processing unit 3 into images in the range direction in order to estimate the phase error in each image.

FIG. 4 is a conceptual diagram when the SAR image is divided in the range direction, and shows an example where the number of divisions is four.
In FIG. 4, one image before division of 0 to 511 bins in the vertical direction is azimuth in the horizontal direction, an image after division of 0 to 127 bins in the range direction, and an image after division of 128 to 255 bins in the range direction This represents that the image is divided into four images as an image after dividing 256 to 383 bins in the range direction and an image after dividing 384 to 511 bins in the range direction.
Note that the values in the image do not change before and after the division, and when the divided images are combined, the original image is obtained.

The range bin selection unit 5, the image shift unit 6, the window function multiplication unit 7, and the phase gradient estimation unit 8 perform processing for each divided image.
The range bin selection unit 5 is a part that selects a range bin used for processing from the image shift unit 6 to the phase gradient estimation unit 8.

The image shift unit 6 is an image shift unit that searches for an isolated point target in the range bin and shifts the entire image so that the target is positioned at the left end of the image.
The window function multiplication unit 7 is a window function multiplication unit that multiplies the window function in the azimuth direction.
The phase gradient estimator 8 is a phase gradient estimator and is a part that estimates a phase error occurring in a reproduced image.

  Note that the range bin selection unit 5, the image shift unit 6, the window function multiplication unit 7, and the phase gradient estimation unit 8 are based on the autofocus method.

The phase error inclination estimation unit 9 is a part that estimates the phase error inclination from the phase error for each range division estimated by the phase gradient estimation unit 8 and obtains the phase error for each range bin.
Here, the phase error for each range division is a phase error obtained by the phase gradient estimation unit 8 from the range bin selection unit 5 using the image after the range division.

  For example, the divided images are image 1, image 2, image 3, and image 4, the phase error calculated for image 1 is phase error 1, the phase error calculated for image 2 is phase error 2, and image 3 When the phase error obtained for the phase error 3 is the phase error 3 and the phase error obtained for the image 4 is the phase error 4, the phase error after range division corresponds to the phase errors 1 to 4.

Here, formulating the phase change for each range is approximately a linear change with respect to distance, so pay attention to this, and estimate the slope of the phase error with respect to the range by the least squares method from the phase error for each range division. To do.
Further, a phase error for each range bin is obtained using the estimated inclination.
As a specific example, first, the obtained phase error Φ _i after the range division is modeled as an equation (1) with an L-order polynomial.

Here, Φ _i models the phase error of the i-th division after range division, m is an azimuth bin number, L is an arbitrary positive number of 2 or more, a _{i, k} are coefficients of a k-th order polynomial. , M is the total azimuth bin number.

Note that L is an order assumed when the phase error is modeled by an L-order polynomial, and a positive number of 2 or more is used.
At this time, since a _{i, k} changes linearly with respect to the range _{, obtaining} this primary change amount from a _{i, k} is equivalent to estimating the slope of the phase error with respect to the range.
This is expressed as equation (2).

Here, Φ ′ is a phase error for each range bin, n is a range bin number, a _k ′ is a coefficient of a _k-th order phase error for each range bin, a _{k, 0} ′ is a value of a _k ′ in a range bin with n = 0, Δa _k ′ represents the slope of the k-th order phase error with respect to the range. That is, estimating the slope of the phase error is estimating Δa _k ′, and obtaining the phase error for each range bin is obtaining a _{k, 0} ′ from the estimated Δa _k ′ and obtaining Φ ′. It corresponds to that.
Here, the phase error after the range division is modeled by an L-order polynomial, but other modeling means such as modeling by an L-order sine wave can be used.

  In estimating the slope, the least square method may be used, or an estimation means other than the least square method may be used.

The phase error compensation unit 10 is a part that compensates the phase error for each range obtained by the phase error slope estimation unit 9.
Specifically, if the SAR image before compensation is Fourier-transformed in the azimuth direction, S _bef (m, n), and the SAR image after compensation is S _aft (m, n), the compensation operation is expressed by Equation (3). ).

Here, j represents an imaginary unit.
This compensation means is greatly different from the conventional technique in that an exponent part of Expression (3) is added.

The iterative determination unit 11 determines whether the image dividing unit 4, the range bin selection unit 5, the image shift unit 6, the window function multiplication unit 7, the phase gradient estimation unit 8, the phase error slope estimation unit 9, and the phase error compensation unit 10 are repeated. It is a part to do.
When the repetition is performed, the processing after the image dividing unit 4 is executed again on the SAR image in which the phase error is compensated.

  Note that the resolution of the image 12 is reduced by the processing of the image division unit 4, the range bin selection unit 5, the image shift unit 6, the window function multiplication unit 7, the phase gradient estimation unit 8, the phase error slope estimation unit 9, and the phase error compensation unit 10. It is an improved SAR image.

Next, the operation will be described.
First, the SAR sensor 1 transmits the high-frequency pulse signal generated by the signal transmission / reception unit 2 toward the ground surface or the sea surface. The SAR sensor 1 receives a reflected echo from the ground surface or the sea surface, and the signal transmitting / receiving unit 2 amplifies the signal.
At this time, the SAR sensor 1 also measures the movement of the platform.

  From the amplified signal and platform motion information, the image reproduction processing unit 3 performs image reproduction signal processing to generate a SAR image. The image dividing unit 4 divides the reproduced SAR image in the range direction. For each of the divided images, the range bin selection unit 5 selects a range bin to be used for the subsequent processing.

For the selected range bin, the image shift unit 6 searches for an isolated point target in the range bin and shifts the entire image so that the target is located on the left side of the image.
The window function multiplier 7 multiplies the shifted image in the azimuth direction with respect to the shifted image.
For the image multiplied by the window function, the phase gradient estimation unit 8 is a part for estimating the phase error occurring in the image.

The processes from the range bin selection unit 5 to the past have been executed individually for each of the images divided in the range direction, and thereafter, these are executed together.
For the phase error estimated for each range division, the phase error gradient estimation unit estimates the gradient of the phase error with respect to the range by the least square method or the like. Further, the phase error for each range bin is obtained using the estimated slope of the phase error.

Using the obtained phase error for each range bin, the phase error compensator compensates for the phase error for each range bin and improves the resolution of the image.
For the image whose resolution has been improved by the above processing, the iterative determination unit 11 performs repetitive determination of the processing, and when performing repetition, the processing after the range bin dividing unit 4 is executed again.

  For example, in a radar with a resolution of 30 cm, even when the resolution is degraded to 60 to 90 cm due to blurring of the image, for example, when the target is a vehicle or the like, the resolution is 30 cm and it is composed of 4 or 5 points. Therefore, the feature as a vehicle can be extracted. However, when the resolution deteriorates to 90 cm, only one or two points are used, and thus it becomes difficult to extract the feature as a vehicle.

  On the other hand, since the resolution of the image is improved to a predetermined 30 cm by using the first embodiment, the characteristics of the vehicle and the like can be extracted, and the radar can obtain only the performance deteriorated from the design value due to the blur of the image. On the other hand, desired performance can be obtained.

  As described above, the SAR image is divided into a plurality in the range direction, the phase error is estimated for each divided image, the inclination of the phase error is estimated from the estimated phase error, and the phase error for each range bin is obtained. By estimating and compensating for a phase error that differs for each range bin, and estimating and compensating for a phase error that differs for each range, it is possible to improve the resolution.

Embodiment 2. FIG.
FIG. 2 is a block diagram showing a high-resolution radar apparatus according to Embodiment 2 of the present invention. Reference numeral 13 denotes an image division number repetition determination unit, and 1 to 12 are the same as those in Embodiment 1.

  The image division number repetition determination unit 13 stores the images whose resolution has been improved by the repetition determination unit 11, and further repeats the processing after the image division unit with a different number of divisions, and compares the resolutions of the obtained images for each number of image divisions. This is the part that outputs the image with the most improved resolution.

In the first embodiment, if the number of divisions is not optimal, the resolution may not be improved to the design value.
For example, in a case where the resolution is degraded to 60 to 90 cm with respect to a radar having a resolution of 30 cm, the division number 3 is improved to 50 cm, and the division number 4 is improved to 30 cm. However, in the second embodiment, the resolution can always be improved to 30 cm.

  As described above, the SAR image is divided into a plurality in the range direction, the phase error is estimated for each divided image, the inclination of the phase error is estimated from the estimated phase error, and the phase error for each range bin is obtained. By repeatedly executing it, it is possible to estimate and compensate for a different phase error for each range bin, and it is possible to obtain an effect that the resolution can be improved over that of the first embodiment.

It is a figure which shows the structure of the high resolution radar apparatus by Embodiment 1 of this invention. It is a figure which shows the structure of the high resolution radar apparatus by Embodiment 2 of this invention. It is a conceptual diagram of observation for demonstrating the problem which this invention solves. It is a conceptual diagram in the case of dividing | segmenting a SAR image in the range direction.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 SAR sensor, 2 Signal transmission / reception part, 3 Image reproduction process part, 4 Image division part, 5 Range bin selection part, 6 Image shift part, 7 Window function multiplication part, 8 Phase gradient estimation part, 9 Phase error inclination estimation part, 10 A phase error compensation unit, 11 a repetition determination unit, 12 images, and a 13 image division number repetition determination unit.

Claims (3)

An image dividing unit for dividing the SAR image into a plurality of ranges in the range direction;
The range bin used for processing is selected from the SAR image divided in the range direction by the image dividing unit, an isolated point target is searched for in the range bin, and the entire image is shifted so that the target is located at the left end of the image. A phase error estimator that multiplies the window function in the azimuth direction and estimates the phase error occurring in the reproduced image for each divided image ;
A phase error slope estimator that estimates the slope of the phase error, which is a change with respect to the distance, from the phase error of each divided image estimated by the phase error estimator;
A phase error compensator that estimates and compensates for different phase errors for each range bin from the phase error slope estimated by the phase error slope estimator;
A high-resolution radar apparatus comprising: The SAR image is mounted on a mobile platform, radiates a high-frequency pulse signal into space, receives a reflected signal from the ground surface or the sea surface, and a SAR sensor comprising a motion sensor that measures the movement of the mobile platform,
An image reproduction processing unit for reproducing a SAR image, which is a two-dimensional high-resolution image, from the reflected signal obtained by the receiving means of the SAR sensor and the motion information of the platform measured by the motion sensor;
The high resolution radar apparatus according to claim 1, wherein the high resolution radar apparatus is obtained. The high resolution according to claim 1, further comprising an image division number repetition determination unit that compares the resolutions of the images for each number of image divisions obtained by processing with different division numbers and outputs an image with the highest resolution. Resolution radar device.

Images