Phase difference auto focusing for synthetic aperture radar imaging

• US 4,999,635 A
• Filed: 03/29/1990
• Issued: 03/12/1991
• Est. Priority Date: 03/29/1990
• Status: Expired due to Term

• Current Assignee: Raytheon Company (Rtx Corporation)
• Original Assignee: Hughes Aircraft Company (Rtx Corporation)
• Inventors: Niho, Yoji G.
• Primary Examiner(s): Barron, Jr., Gilberto
• Application Number:
  US07/502,000
• Time in Patent Office: 348 Days
• Field of Search: 342/25
• US Class Current: 342/25.00D
• CPC Class Codes: G01S 13/9011 with frequency domain processing of the SAR signals in azimuth G01S13/9005 takes precedence
• CPC Class Codes: G01S 13/9019 Auto-focussing of the SAR signals

Abstract

The present invention discloses an autofocusing method and apparatus for Synthetic Aperture Radar which employs operation on two or three subarrays of the SAR compressed data to estimate quadratic and cubic phase errors in the data. Focusing of the SAR array may then be accomplished by removing the estimated phase error from the data to sharpen the image. Phase error is estimated by FFT filtering product arrays formed by multiplying the complex conjugate of one subarray by another of the subarrays to form a correlation function. The correlation functions are integrated. A peak detection of the integrated correlation functions yields a peak location in the filter which is proportional to the phase error which may then be calculated.

Claims

1. An apparatus for automatically focusing range compressed SAR data stored in an array of range bins and azimuth lines comprising:

means for unpacking the range compressed data in each range bin;

means for generating a first and second subarray from the unpacked data;

means for conjugating the first subarray;

means for multiplying the conjugated first subarray and the second subarray to create a first product array;

means for performing a fast Fourier transform of the first product array to produce an FFT filter bank output;

means for calculating the magnitude of the FFT filter to form a correlation function;

means for integrating the correlation function over the range bins;

means for detecting a peak in the correlation function after integrating over range bins;

means for calculating a quadratic phase error responsive to the detected peak; and

means for removing the calculated phase error from the range compressed data.

2. An apparatus as defined in claim 1 further comprising:

means for reducing sidelobe outputs from the FFT filter bank output; and

means for centering the FFT filter bank.

3. An apparatus as described in claim 2 further comprising means for normalizing the correlation function by dividing each value of the correlation function by the sum of the magnitudes of the input data to the FFT.

4. An apparatus as described in claim 3 further comprising:

means for generating a third subarray;

means for calculating the complex conjugate of third subarray;

means for multiplying the complex conjugate of the first subarray and the third subarray to obtain a second product array;

means for multiplying the complex conjugate of the third subarray and the second subarray to obtain a third product array;

means for performing a fast Fourier transform on the second and the third product arrays to produce a second and a third filter bank output;

means for determining the magnitude of the second and the third filter bank output to form a second and a third correlation functions;

means for normalizing the correlation function by the sum of magnitudes of the input data to the FFT;

means for integrating the second and the third correlation functions over the range bins;

means for detecting a peak filter in the second and the third correlation functions after integrating over the range bins;

means for calculating a quadratic error estimate responsive to the peak detected in the second and third correlation function;

means for calculating a cubic error estimate responsive to the peak detected in the second and third correlation function; and

means for correcting the range compressed data for both the quadratic and the cubic errors calculated.

9. An apparatus as described in claim 3 further comprising:

means for generating a third subarray,

means for calculating the complex conjugate of the second subarray,

means for multiplying the complex conjugate of the second subarray and the third subarray to obtain a second product array,

means for performing a fast Fourier transform on the second product array to produce a second filter bank output,

means for determining the magnitude of the second filter bank output to form a second correlation function,

means for integrating the second correlation function over the range bins,

means for detecting a peak in the second correlation function after integrating over the range bins,

means for calculating a cubic error estimate responsive to the peak detected in the first correlation function and the peak detected in the second correlation function, and,

means for correcting the range compressed data for the cubic error calculated.

10. An apparatus as described in claim 9 further comprising means for normalizing the second correlation function by the sum of magnitudes of the input data to the FFT.

5. A method for autofocusing a quadratic phase error from range compressed SAR data contained in an array of range bins and azimuth lines comprising the steps of:

creating a first subarray of the SAR data;

creating a second subarray of the SAR data;

calculating the complex conjugate of the first taking the product of the complex conjugate of the first subarray and the second subarray;

performing a fast Fourier transform on the product to produce an FFT filter bank;

detecting the magnitude of the FFT filters to form a correlation function;

integrating the correlation function over the range bins;

detecting a peak in the correlation function after integrating overall range bins;

calculating a quadratic phase error responsive to the detected peak; and

correcting the range compressed data for the quadratic phase error calculated.

6. A method for autofocusing both a quadratic and a cubic phase errors as defined in claim 5 further comprising the steps of:

creating a third subarray of the SAR data;

calculating the complex conjugate of the third taking the product of the complex conjugate of the first subarray and the third subarray to form a second product array;

taking the product of the complex conjugate of the third subarray and the second subarray to form a third product array;

performing a fast Fourier transform on the second and the third product arrays to form a second and a third FFT filter bank;

detecting the magnitude of the second and the third FFT filters to form a second and a third correlation functions;

integrating the second and the third correlation functions over the range bins;

determining a peak in the second and in the third correlation functions after integrating over the range bins;

calculating both a quadratic and a cubic phase errors responsive to the second peak and third peak; and

correcting the range compressed data for both the quadratic and the cubic phase errors.

7. A method for autofocusing a quadratic phase error as defined in claim 5 further comprising the step of normalizing the correlation function by dividing each FFT filter by the sum of magnitudes of the input data to the FFT after detecting the magnitude.

8. A method for autofocusing both a quadratic and a cubic phase errors as defined in claim 6 further comprising the step of normalizing the second and the third correlation functions by dividing each value of the correlation functions by the sum of magnitudes of the respective input data to the FFT after detecting the