IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0247603
(2008-10-08)
|
등록번호 |
US-8155452
(2012-04-10)
|
발명자
/ 주소 |
|
출원인 / 주소 |
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
12 인용 특허 :
35 |
초록
▼
A method for correlating or finding similarity between two data sets. The method can be used for correlating two images with common scene content in order to find correspondence points between the data sets. These correspondence points then can be used to find the transformation parameters which whe
A method for correlating or finding similarity between two data sets. The method can be used for correlating two images with common scene content in order to find correspondence points between the data sets. These correspondence points then can be used to find the transformation parameters which when applied to image 2 brings it into alignment with image 1. The correlation metric has been found to be invariant under image rotation and when applied to corresponding areas of a reference and target image, creates a correlation surface superior to phase and norm cross correlation with respect to the correlation peak to correlation surface ratio. The correlation metric was also found to be superior when correlating data from different sensor types such as from SAR and EO sensors. This correlation method can also be applied to data sets other than image data including signal data.
대표청구항
▼
1. A method for determining similarity between a plurality of signals, comprising: selecting a reference data set representing selected characteristics of a defined subject matter;collecting data to define a target data set, said target data set representing said selected characteristics measured fo
1. A method for determining similarity between a plurality of signals, comprising: selecting a reference data set representing selected characteristics of a defined subject matter;collecting data to define a target data set, said target data set representing said selected characteristics measured for at least an overlap area of said defined subject matter common to said target data set and said reference data set;calculating for each one of a plurality of sub-regions within said overlap area, a set of normalized cross-correlation values using said reference data set and said target data set to evaluate a plurality of possible positions of the target data within the reference data;calculating for said plurality of sub-regions within said overlap area a set of phase correlation values using said reference data set and said target data set to evaluate a plurality of possible position of the target data within the reference data;calculating an element by element product of the normalized cross-correlation set and the phase correlation set to determine a phase-rho correlation set for each of said plurality of sub-regions within the overlap area;determining a correlation surface peak location for each sub-region defined by identifying a highest value in the phase-rho correlation set for each sub-region; andusing point sets corresponding to the correlation surface peak locations from selected sub-regions in the overlap area to determine a transformation that minimizes the distance between the reference data set point locations and the corresponding point locations in the target data set to align the target data set with the reference data set;wherein at least one of said reference data and said target data is selected from the group consisting of image data, rf signal data, and audio data collected by a sensor. 2. The method according to claim 1, further comprising selecting each of said reference data set and said target data set to be one-dimensional sets. 3. The method according to claim 1, further comprising selecting each of said reference data set and said target data set to be two-dimensional data sets. 4. The method according to claim 3, further comprising: selecting a two dimensional portion of said target data set to define a template;selectively moving said template to each of said plurality of possible positions within each said sub-region;calculating said normalized cross-correlation set and said phase correlation set using said two dimensional portion of said target data defined by said template and a two dimensional portion of said reference data defined within said sub-region. 5. The method according to claim 4, wherein said phase-rho correlation set defines a three-dimensional surface where the x, y locations represent the center position of the template with respect to the reference data set, and the z value is the phase-rho correlation value for each of the x, y locations. 6. The method according to claim 5, further comprising: using a peak or maximum z value in each said three dimensional surface calculated for each said template to determine a plurality of correspondence points for associating said reference data and said target data;using said correspondence points in an optimization routine which minimizes a distance between the corresponding points. 7. The method according to claim 6, wherein said distance between each corresponding point set is minimized simultaneously under predetermined transformations selected from the group consisting of translation, rotation, and scale. 8. The method according to claim 6, further comprising, using a result of said optimization routine to determine a transformation which optimally brings the target data set into alignment with the reference data set. 9. The method according to claim 3, further comprising selecting said target data set and said reference data set to each comprise image data. 10. A computer system for determining similarity between a plurality of signals, comprising: a least one data store configured for storing a reference data set representing selected characteristics of a defined subject matter, and for storing data which defines a target data set, said target data set representing said selected characteristics measured for at least an overlap area of said defined subject matter common to said target data set and said reference data set;at least one processing device configured to:calculate for each one of a plurality of sub-regions within said overlap area, a set of normalized cross-correlation values using said reference data set and said target data set to evaluate a plurality of possible positions of the target data within the reference data;calculate for said plurality of sub-regions within said overlap area a set of phase correlation values using said reference data set and said target data set to evaluate a plurality of possible position of the target data within the reference data;calculate an element by element product of the normalized cross-correlation set and the phase correlation set to determine a phase-rho correlation set for each of said plurality of sub-regions within the overlap area;determine a correlation surface peak location for each sub-region defined by identifying a highest value in the phase-rho correlation set for each sub-region; andto use point sets corresponding to the correlation surface peak locations from selected sub-regions in the overlap area to determine a transformation that minimizes the distance between the reference data set point locations and the corresponding point locations in the target data set to align the target data set with the reference data set;wherein at least one of said reference data and said target data is selected from the group consisting of image data, rf signal data, and audio data collected by a sensor. 11. The computer system according to claim 10, wherein each of said reference data set and said target data set are one-dimensional sets. 12. The computer system according to claim 10, wherein each of said reference data set and said target data set are two-dimensional data sets. 13. The computer system according to claim 12, wherein a two dimensional portion of said target data set defines a template, and said at least one processing device is further configured for: selectively moving said template to each of said plurality of possible positions within each said sub-region; andcalculating said normalized cross-correlation set and said phase correlation set using said two dimensional portion of said target data defined by said template and a two dimensional portion of said reference data defined within said sub-region. 14. The computer system according to claim 13, wherein said phase-rho correlation set defines a three-dimensional surface where the x, y locations represent the center position of the template with respect to the reference data set, and the z value is the phase-rho correlation value for each of the x, y locations. 15. The computer system according to claim 14, wherein said at least one processing device is further configured for: using a peak or maximum z value in each of said three dimensional surface calculated for each said template to determine a plurality of correspondence points for associating said reference data and said target data; andusing said correspondence points in an optimization routine which minimizes a distance between the corresponding points. 16. The computer system according to claim 15, wherein said at least one processing device is further configured for simultaneously minimizing a distance between each corresponding point set using predetermined transformations selected from the group consisting of translation, rotation, and scale. 17. The computer system according to claim 15, wherein said at least one processing device is further configured for using a result of said optimization routine to determine a transformation which optimally brings the target data set into alignment with the reference data set. 18. The computer system according to claim 12, wherein said target data set and said reference data set to each comprise image data. 19. A non-transitory machine readable media programmed with a set of instructions for determining similarity between a plurality of signals, comprising: selecting a reference data set representing selected characteristics of a defined subject matter;collecting data to define a target data set, said target data set representing said selected characteristics measured for at least an overlap area of said defined subject matter common to said target data set and said reference data set;calculating for each one of a plurality of sub-regions within said overlap area, a set of normalized cross-correlation values using said reference data set and said target data set to evaluate a plurality of possible positions of the target data within the reference data;calculating for said plurality of sub-regions within said overlap area a set of phase correlation values using said reference data set and said target data set to evaluate a plurality of possible position of the target data within the reference data;calculating an element by element product of the normalized cross-correlation set and the phase correlation set to determine a phase-rho correlation set for each of said plurality of sub-regions within the overlap area;determining a correlation surface peak location for each sub-region defined by identifying a highest value in the phase-rho correlation set for each sub-region; andusing point sets corresponding to the correlation surface peak locations from selected sub-regions in the overlap area to determine a transformation that minimizes the distance between the reference data set point locations and the corresponding point locations in the target data set to align the target data set with the reference data set;wherein at least one of said reference data and said target data is selected from the group consisting of image data, rf signal data, and audio data collected by a sensor. 20. A method for determining similarity between a plurality of signals, comprising: selecting a two-dimensional reference data set representing image data associated with a defined subject matter;collecting data to define a target data set, said target data set representing two dimensional image data for at least an overlap area of said defined subject matter common to said target data set and said reference data set;calculating for each one of a plurality of sub-regions within said overlap area, a set of normalized cross-correlation values using said reference data set and said target data set to evaluate a plurality of possible positions of the target data within the reference data;calculating for said plurality of sub-regions within said overlap area a set of phase correlation values using said reference data set and said target data set to evaluate a plurality of possible position of the target data within the reference data;calculating an element by element product of the normalized cross-correlation set and the phase correlation set to determine a phase-rho correlation set for each of said plurality of sub-regions within the overlap area;determining a correlation surface peak location for each sub-region defined by identifying a highest value in the phase-rho correlation set for each sub-region; andusing point sets corresponding to the correlation surface peak locations from selected sub-regions in the overlap area to determine a transformation that minimizes the distance between the reference data set point locations and the corresponding point locations in the target data set to align the target data set with the reference data set. 21. The method according to claim 20, further comprising: selecting a two dimensional portion of said target data set to define a template;selectively moving said template to each of said plurality of possible positions within each said sub-region;calculating said normalized cross-correlation set and said phase correlation set using said two dimensional portion of said target data defined by said template and a two dimensional portion of said reference data defined within said sub-region.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.