US4780718A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Sar image encoding for data compression
Assignee: Raytheon Co
•
Filing: 1985-06-17
•
Granted: 1988-10-25
•
Expires: 2005-06-17
Abstract
Abstract A technique is disclosed for encoding SAR image data to achieve data compression. In the image encoding stage, the SAR image is transformed into a list of high reflectivity radar discretes and a small array of frequency filters. In the target list, the location data and intensity levels above the local average background clutter are tabulated for a predetermined number of the highest intensity radar discretes. The array of frequency filters is divided into three zones; the inner, middle, and higher frequency zones relative to the d.c. filter. Only the inner and middle zones of filters are retained and the outer filters are discarded, thus acheiving the desired data reduction. The inner zone filters are quantized to a higher level of precision than the middle zone of filters. The saturation levels of the filters are determined adaptively. In the decoding stage, the original SAR image is reconstructed from the radar discrete list and the small array of frequency filters.
Inventors
- 👤 Ralph E. Hudson
- 👤 Yoji G. Niho
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-06 00:16
CPC Classifications
G; G01; G01S; G01S7/00; G01S7/003; G01S13/00; G01S13/88; G01S13/89; G01S13/90; G01S13/9021; G01S13/9027
Claims Analysis
Total Claims: 24
Claim 1
Independent Claim
197 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (i) extracting from the SAR image a set of the brightest radar discretes to form a set of radar discretes, comprising a sequence of the following steps: (a) processing said SAR image to obtain a low frequency image containing only the low frequency components of the SAR image, comprising the following sequence of steps: transforming said SAR image into the frequency domain, processing said transformed image data to retain the low frequency components thereof and zeroing the other frequency components, and performing an inverse transform on the processed image data back to the spatial domain, thereby obtaining a low frequency image version of the SAR image; (b) removing the low frequency image from the SAR image to form a first difference image, and (c) processing said first difference image to identify said set of radar discretes; and (ii) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 2
Independent Claim
197 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (i) extracting from the SAR image a set of the brightest radar discretes to form a set of radar discretes, comprising a sequence of the following steps: (a) processing said SAR image to obtain a low frequency image containing only the low frequency components of the SAR image, comprising the following sequence of steps: transforming said SAR image into the frequency domain, processing said transformed image data to retain the low frequency components thereof and zeroing the other frequency components, and performing an inverse transform on the processed image data back to the spatial domain, thereby obtaining a low frequency image version of the SAR image; (b) removing the low frequency image from the SAR image to form a first difference image, and (c) processing said first difference image to identify said set of radar discretes; and (ii) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 3
Dependent Claim
86 words
The method of claim 1 wherein said step of processing said SAR image and said set of radar discretes comprises the steps of: (i) extracting said set of radar discretes from said SAR image to obtain a second difference image; (ii) transforming said second difference signal into the frequency domain by using a fast Fourier transformation resulting in an array of complex-valued filters to obtain an array of frequency filters; (iii) discarding the higher frequency filters; and (iv) quantizing the lower and middle frequency level filters.
Claim 4
Dependent Claim
127 words
A data compression encoder for encoding input synthetic aperture radar (SAR) image data representing an image comprising an array of pixels, and wherein each array value indicates the intensity level of a corresponding pixel, comprising: first processing means for processing said image data to identify relatively high intensity pixels and their respective intensity values, said first processing means for defining a high intensity pixel data set representing the locations of said pixels and their respective intensity values; second processing means for processing said input image data and said set to provide an array of frequency filters representing the spectral content of the relatively low intensity image background; whereby said input image data is represented by said high intensity pixel data set and said array of frequency filters.
Claim 5
Dependent Claim
127 words
A data compression encoder for encoding input synthetic aperture radar (SAR) image data representing an image comprising an array of pixels, and wherein each array value indicates the intensity level of a corresponding pixel, comprising: first processing means for processing said image data to identify relatively high intensity pixels and their respective intensity values, said first processing means for defining a high intensity pixel data set representing the locations of said pixels and their respective intensity values; second processing means for processing said input image data and said set to provide an array of frequency filters representing the spectral content of the relatively low intensity image background; whereby said input image data is represented by said high intensity pixel data set and said array of frequency filters.
Claim 6
Dependent Claim
64 words
The invention of claim 3 wherein said input image data is logarithmically encoded, and said first processing means comprises: low pass filter means for providing a low frequency component of said image data; means for subtracting said low frequency component from said input image data to provide a first difference image; and means for identifying the high intensity pixels in said first difference image.
Claim 7
Dependent Claim
42 words
The invention of claim 3 wherein said means for identifying the high intensity pixels comprises sorting means for sorting the pixels by intensity value and selecting a predetermined number of said pixels in descending order in accordance with their respective intensity levels.
Claim 8
Dependent Claim
40 words
The invention of claim 3 wherein said means for identifying the high intensity pixels in said difference image comprises means for comparing the respective pixel intensity to a predetermining threshold and identifying those pixels whose intensity level exceeds said threshold.
Claim 9
Dependent Claim
67 words
The encoder of claim 3 wherein said second processing means comprises: means for extracting said high intensity pixel data set from said input image data to provide a second difference image; Fourier transform means for transforming said second difference image into a transform array of frequency filters; and means for extracting the higher frequency filters from said transform array and for providing said array of frequency filters.
Claim 10
Dependent Claim
31 words
The encoder of claim 7 wherein said means for extracting the higher frequency components from said transform array further further comprises quantizer means for quantizing the lower and middle frequency filters.
Claim 11
Dependent Claim
61 words
The encoder of claim 8 wherein said transform array comprises an inner group of low frequency filters centered about the zero frequency filter and a middle group of middle frequency filters centered about said zero frequency filter, and said quantizer is adapted to quantize the inner group of filters to a higher degree of precision than the middle group of filters.
Claim 12
Independent Claim
156 words
A method for processing synthetic aperture radar (SAR) image data, comprising the step of encoding the SAR image data by: (i) transforming the SAR image into the frequency domain to obtain a first frequency transform data set; (ii) low pass filtering the transformed data set; (iii) transforming the filtering data set back into the spatial domain to obtain a first low frequency version of the SAR image; (iv) extracting the low frequency version from the SAR image to obtain a bright object version of the SAR image; (v) removing the bright object version from said SAR image to obtain a background image; (vi) transforming said background image into the frequency domain; and (vii) low pass filtering the transformed background image data and quantizing the filtered image to obtain a set of low frequency image data, whereby said SAR image data is encoded into said bright object image version and said set of low frequency image data.
Claim 13
Independent Claim
156 words
A method for processing synthetic aperture radar (SAR) image data, comprising the step of encoding the SAR image data by: (i) transforming the SAR image into the frequency domain to obtain a first frequency transform data set; (ii) low pass filtering the transformed data set; (iii) transforming the filtering data set back into the spatial domain to obtain a first low frequency version of the SAR image; (iv) extracting the low frequency version from the SAR image to obtain a bright object version of the SAR image; (v) removing the bright object version from said SAR image to obtain a background image; (vi) transforming said background image into the frequency domain; and (vii) low pass filtering the transformed background image data and quantizing the filtered image to obtain a set of low frequency image data, whereby said SAR image data is encoded into said bright object image version and said set of low frequency image data.
Claim 14
Dependent Claim
44 words
The method of claim 10 wherein said SAR image data comprises a pixel matrix of digitized, logarithmically encoded data describing the image intensity at each pixel and wherein said step (iv) of extracting the filtered image from the SAR image comprises a pixel-by-pixel subtraction.
Claim 15
Dependent Claim
79 words
The method of claim 10 further comprising the step of decoding said encoding SAR image data into a reconstructed SAR image, comprising: (viii) zero filling said set of low frequency image data to include higher frequency components of zero value; (ix) transforming said zero-filled set back to the spatial domain to provide a second low frequency version of the SAR image; and (x) adding said bright object image to said low frequency version to obtain the reconstructed SAR image.
Claim 16
Independent Claim
135 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (a) extracting from the SAR image a set of the brightest radar discretes to form a set of radar discretes, said step comprising: (i) processing said SAR image to obtain a low frequency image containing only the low frequency components of the SAR image; (ii) removing the low frequency image from the SAR image to form a first difference image; and (iii) processing said first difference image to identify said set of radar discretes; and (b) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 17
Independent Claim
135 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (a) extracting from the SAR image a set of the brightest radar discretes to form a set of radar discretes, said step comprising: (i) processing said SAR image to obtain a low frequency image containing only the low frequency components of the SAR image; (ii) removing the low frequency image from the SAR image to form a first difference image; and (iii) processing said first difference image to identify said set of radar discretes; and (b) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 18
Dependent Claim
72 words
The method of claim 13 wherein said step of processing said SAR image to obtain a low frequency image comprises: transforming said SAR image into the frequency domain; processing said transformed image data to retain the low frequency components thereof and zeroing the other frequency components; and performing an inverse transform on the processed image data back to the spatial domain, thereby obtaining a low frequency image version of the SAR image.
Claim 19
Dependent Claim
22 words
The method of claim 14 wherein said step of transforming said SAR image into the frequency domain comprises a fast Fourier transformation.
Claim 20
Independent Claim
124 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (a) digitizing and logarithmically encoding said SAR image data into an array of pixels; (b) extracting from said digitized and encoded SAR image data a set of the brightest radar discretes to form a set of radar discretes, comprising: (i) filtering said SAR image data to provide a low frequency image, and (ii) subtracting said low frequency image from said SAR image pixel-by-pixel; and (c) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 21
Independent Claim
124 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (a) digitizing and logarithmically encoding said SAR image data into an array of pixels; (b) extracting from said digitized and encoded SAR image data a set of the brightest radar discretes to form a set of radar discretes, comprising: (i) filtering said SAR image data to provide a low frequency image, and (ii) subtracting said low frequency image from said SAR image pixel-by-pixel; and (c) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 22
Independent Claim
132 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (a) extracting from the SAR image a set of the brightest radar discretes to form a set of radar discretes; and (b) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image, comprising: (i) extracting said set of radar discretes from said SAR image to obtain a second difference image; (ii) transforming said second difference signal into the frequency domain to obtain an array of frequency filters; (iii) discarding the higher frequency filters; and (iv) quantizing the lower and middle frequency level filters; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 23
Independent Claim
132 words
A method for encoding synthetic aperture radar (SAR) image data, comprising a sequence of the following steps: (a) extracting from the SAR image a set of the brightest radar discretes to form a set of radar discretes; and (b) processing said SAR image and said set of radar discretes to obtain background image data representing only the background clutter component of the SAR image, comprising: (i) extracting said set of radar discretes from said SAR image to obtain a second difference image; (ii) transforming said second difference signal into the frequency domain to obtain an array of frequency filters; (iii) discarding the higher frequency filters; and (iv) quantizing the lower and middle frequency level filters; whereby said SAR image is characterized by said background image data and said set of radar discretes.
Claim 24
Dependent Claim
30 words
The method of claim 17 wherein said step of transforming said second difference signal into the frequency domain comprises a fast fourier transformation resulting in an array of complex-valued filters.
References Cited (Backward Citations)
US3215934A
Cited
US3271765A
Cited
US3885224A
Cited
US3964064A
Cited
US4045795A
Cited
US4225864A
Cited
US4594593A
Cited
JPS58213268A
Cited
JPS58213269A
Cited
JPS5994092A
Cited
EP0123573A1
Cited
US4617567A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1991-10-25 | 1992-04-25 | 1992-10-25 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1995-10-25 | 1996-04-25 | 1996-10-25 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1999-10-25 | 2000-04-25 | 2000-10-25 | $7700.00 | Window Closed / Lapsed |