US6002807A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Rotationally invariant correlator
Assignee: Northrop Grumman Systems Corp
•
Filing: 1996-01-11
•
Granted: 1999-12-14
•
Expires: 2016-01-11
Abstract
Abstract A rotationally invariant correlator which correlates an arbitrarily rotated image with a large database of similar stored images. The correlator initially scans the image to produce a two-dimensional array of pixels representative thereof. A plurality of concentric rings of pixel data are then produced from the two-dimensional array of pixels. A Fast Fourier Transform is taken of each ring of pixel data which produces an amplitude spectra of the frequency components of that ring of pixel data, which is then normalized. The normalized amplitude spectra for the image being correlated is then compared with normalized amplitude spectra of similar rings of pixel data for similar images in memory, and any matches are identified. The width of each ring of pixel data is selected to be smaller than the size of the smallest features which are to be resolved, while being sufficiently wide to smooth out extraneous noise. Several of the processing steps for each ring of pixel data are performed in parallel in parallel processing arrangements.
Inventors
- 👤 Joseph R. Guerci
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 11:18
CPC Classifications
G; G06; G06F; G06F17/00; G06F17/10; G06F17/15; G06F17/153; G06V; G06V10/00; G06V10/70; G06V10/74; G06V10/75; G06V10/751; G06V10/7515; G06V20/00; G06V20/60; G06V20/66
Claims Analysis
Total Claims: 32
Claim 1
Independent Claim
176 words
A rotationally invariant method of image correlation comprising: a. scanning an image being correlated to produce a set of pixels representative thereof, wherein each of the pixels has a data value; b. storing the pixel data values in a computer memory; c. using a processor means to identify at least one ring of pixels and the data values for each ring of pixels, wherein the data values for each ring of pixels form a respective ring of pixel data; d. taking the Fast Fourier Transform of each ring of pixel data; e. storing in memory in a database Fast Fourier Transforms of rings of pixel data for similar images; f. comparing and correlating the Fast Fourier Transform of each ring of pixel data for the image being correlated with Fast Fourier Transforms of rings of pixel data stored in the database; and g. identifying any match of the Fast Fourier Transform of each ring of pixel data for the image being correlated with the Fast Fourier Transforms of rings of pixel data stored in the database.
Claim 2
Dependent Claim
176 words
A rotationally invariant method of image correlation comprising: a. scanning an image being correlated to produce a set of pixels representative thereof, wherein each of the pixels has a data value; b. storing the pixel data values in a computer memory; c. using a processor means to identify at least one ring of pixels and the data values for each ring of pixels, wherein the data values for each ring of pixels form a respective ring of pixel data; d. taking the Fast Fourier Transform of each ring of pixel data; e. storing in memory in a database Fast Fourier Transforms of rings of pixel data for similar images; f. comparing and correlating the Fast Fourier Transform of each ring of pixel data for the image being correlated with Fast Fourier Transforms of rings of pixel data stored in the database; and g. identifying any match of the Fast Fourier Transform of each ring of pixel data for the image being correlated with the Fast Fourier Transforms of rings of pixel data stored in the database.
Claim 3
Dependent Claim
45 words
A rotationally invariant method of image correlation as claimed in claim 1, wherein said scanning step produces a two-dimensional array of pixels representative of the image being correlated, and the at least one ring of pixel data is produced from the two-dimensional array of pixels.
Claim 4
Dependent Claim
33 words
A rotationally invariant method of image correlation as claimed in claim 1, wherein said step of using comprises producing a plurality of concentric rings of pixel data from the two-dimensional array of pixels.
Claim 5
Dependent Claim
30 words
A rotationally invariant method of image correlation as claimed in claim 1, wherein said steps of taking, comparing and identifying are performed in parallel for each ring of pixel data.
Claim 6
Dependent Claim
32 words
A rotationally invariant method of image correlation as claimed in claim 1, including the step of analyzing the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 7
Dependent Claim
70 words
A rotationally invariant method of image correlation as claimed in claim 1, wherein said step of taking the Fast Fourier Transform produces amplitude spectra of the frequency components of each ring of pixel data, and further including the step of normalizing the amplitude spectra of the frequency components of each ring of pixel data, and the normalized amplitude spectra is compared with the normalized amplitude spectra of images in memory.
Claim 8
Dependent Claim
64 words
A rotationally invariant method of image correlation as claimed in claim 1, wherein the width of each ring of pixel data is selected in dependence upon the size of the smallest features which are to be resolved, with the width of each ring of pixel data being smaller than the smallest features to be resolved, while being sufficiently wide to smooth out extraneous noise.
Claim 9
Dependent Claim
32 words
A rotationally invariant method of image correlation as claimed in claim 2, including the step of analyzing the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 10
Dependent Claim
33 words
A rotationally invariant method of image correlation as claimed in claim 2, wherein said step of producing comprises producing a plurality of concentric rings of pixel data from the two-dimensional array of pixels.
Claim 11
Dependent Claim
30 words
A rotationally invariant method of image correlation as claimed in claim 9, wherein said steps of taking, comparing and identifying are performed in parallel for each ring of pixel data.
Claim 12
Dependent Claim
32 words
A rotationally invariant method of image correlation as claimed in claim 10, including the step of analyzing the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 13
Dependent Claim
71 words
A rotationally invariant method of image correlation as claimed in claim 11, wherein said step of taking the Fast Fourier Transform produces an amplitude spectra of the frequency components of each ring of pixel data, and further including the step of normalizing the amplitude spectra of the frequency components of each ring of pixel data, and the normalized amplitude spectra is compared with the normalized amplitude spectra of images in memory.
Claim 14
Dependent Claim
64 words
A rotationally invariant method of image correlation as claimed in claim 12, wherein the width of each ring of pixel data is selected in dependence upon the size of the smallest features which are to be resolved, with the width of each ring of pixel data being smaller than the smallest features to be resolved, while being sufficiently wide to smooth out extraneous noise.
Claim 15
Dependent Claim
32 words
A rotationally invariant method of image correlation as claimed in claim 13, including the step of analyzing the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 16
Dependent Claim
193 words
A rotationally invariant image correlator comprising: a. means for scanning an image being correlated to produce a set of pixels representative of the image being correlated, wherein each of the pixels has a data value; b. a computer memory for storing the pixel data values; c. a processor means to identify at least one ring of pixels and the data values for each ring of pixels, wherein the data values for each ring of pixels form a respective ring of pixel data; d. Fast Fourier Transform means for taking the Fast Fourier Transform of each ring of pixel data; and e. a memory database having stored therein Fast Fourier Transforms of rings of pixels data from similar images; wherein the processor means includes means for comparing and correlating the Fast Fourier Transform of each ring of pixel data for the image being correlated with the Fast Fourier Transforms of rings of pixel data stored in said memory databases and for identifying any matches of the Fase Fourier Transform, of each ring of pixel data for the image being correlated, with the Fast Fourier Transforms of rings of pixel data stored in the database.
Claim 17
Dependent Claim
193 words
A rotationally invariant image correlator comprising: a. means for scanning an image being correlated to produce a set of pixels representative of the image being correlated, wherein each of the pixels has a data value; b. a computer memory for storing the pixel data values; c. a processor means to identify at least one ring of pixels and the data values for each ring of pixels, wherein the data values for each ring of pixels form a respective ring of pixel data; d. Fast Fourier Transform means for taking the Fast Fourier Transform of each ring of pixel data; and e. a memory database having stored therein Fast Fourier Transforms of rings of pixels data from similar images; wherein the processor means includes means for comparing and correlating the Fast Fourier Transform of each ring of pixel data for the image being correlated with the Fast Fourier Transforms of rings of pixel data stored in said memory databases and for identifying any matches of the Fase Fourier Transform, of each ring of pixel data for the image being correlated, with the Fast Fourier Transforms of rings of pixel data stored in the database.
Claim 18
Dependent Claim
47 words
A rotationally invariant image correlator as claimed in claim 15, wherein: a. said means for scanning produces a two-dimensional array of pixels for the image being correlated; and b. said processor means produces the at least one ring of pixel data from the two-dimensional array of pixels.
Claim 19
Dependent Claim
29 words
A rotationally invariant image correlator as claimed in claim 15, wherein said processor means produces a plurality of concentric rings of pixel data from the two-dimensional array of pixels.
Claim 20
Dependent Claim
31 words
A rotationally invariant image correlator as claimed in claim 15, wherein said processor means performs data processing operations in parallel for each ring of pixel data for the image being correlated.
Claim 21
Dependent Claim
30 words
A rotationally invariant image correlator as claimed in claim 15, wherein said processor means analyzes the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 22
Dependent Claim
89 words
A rotationally invariant image correlator as claimed in claim 15, wherein: a. said Fast Fourier Transform means produces amplitude spectra of the frequency components of each ring of pixel data for the image being correlated; b. the processing means further including means for normalizing the amplitude spectra of the frequency components of each ring of pixel data for the image being correlated; and c. the memory database has stored therein normalized amplitude spectra for similar images, with which the normalized amplitude spectra of the image being correlated is compared.
Claim 23
Dependent Claim
50 words
A rotationally invariant image correlator as claimed in claim 15, wherein the processor means produces a selected width ring of pixel data, wherein the selected width of the ring of pixel data is smaller than the smallest features to be resolved, while being sufficiently wide to smooth out extraneous noise.
Claim 24
Dependent Claim
30 words
A rotationally invariant image correlator as claimed in claim 16, wherein said processor means analyzes the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 25
Dependent Claim
29 words
A rotationally invariant image correlator as claimed in claim 16, wherein said processor means produces a plurality of concentric rings of pixel data from the two-dimensional array of pixels.
Claim 26
Dependent Claim
31 words
A rotationally invariant image correlator as claimed in claim 23, wherein said processor means performs data processing operations in parallel for each ring of pixel data for the image being correlated.
Claim 27
Dependent Claim
30 words
A rotationally invariant image correlator as claimed in claim 24, wherein said processor means analyzes the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 28
Dependent Claim
89 words
A rotationally invariant image correlator as claimed in claim 25, wherein: a. said Fast Fourier Transform means produces amplitude spectra of the frequency components of each ring of pixel data for the image being correlated; b. the processing means further including means for normalizing the amplitude spectra of the frequency components of each ring of pixel data for the image being correlated; and c. the memory database has stored therein normalized amplitude spectra for similar images, with which the normalized amplitude spectra of the image being correlated is compared.
Claim 29
Dependent Claim
50 words
A rotationally invariant image correlator as claimed in claim 26, wherein the processor means produces a selected width ring of pixel data, wherein the selected width of the ring of pixel data is smaller than the smallest features to be resolved, while being sufficiently wide to smooth out extraneous noise.
Claim 30
Dependent Claim
30 words
A rotationally invariant image correlator as claimed in claim 27, wherein said processor means analyzes the two-dimensional array of pixels to determine the center of the two-dimensional array of pixels.
Claim 31
Dependent Claim
22 words
A rotationally invariant method of image correlation as claimed in claim 1, wherein said scanning step scans an unknown and irregular image.
Claim 32
Dependent Claim
21 words
A rotationally invariant image correlator as claimed in claim 15, wherein said means for scanning scans an unknown and irregular image.
References Cited (Backward Citations)
US3074634A
Cited
US3944978A
Cited
US4267573A
Cited
US4802230A
Cited
US5276629A
Cited
US5426708A
Cited
US5359670A
Cited
US5465303A
Cited
US5812252A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2002-12-13 | 2003-06-14 | 2003-12-14 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2006-12-13 | 2007-06-14 | 2007-12-14 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2010-12-13 | 2011-06-14 | 2011-12-14 | $7700.00 | Window Closed / Lapsed |