Correlatability analysis for sparse alignment
First Claim
1. A computer-implementable method of estimating a geometrical relationship between a first image and a second image, said second image including a noise component, said method comprising the steps of:
- (a) determining a location and size of each one of a plurality of image patches for each of said images, based on the noise component included in said second image and correlation information derived from said first image, wherein the size of each of the plurality of image patches is determined using a Cramer-Rao lower bound of two-dimensional shift estimation;
(b) identifying a plurality of first image areas in said first image and a corresponding plurality of second image areas in said second image, based on the location and size of each one of said plurality of image patches, wherein each first image area of said first image corresponds to a related second image area of said second image; and
(c) determining a geometrical relationship between said first and second images by comparing, for each one of said first image areas, information located within said first image area with information located within said corresponding related second image area.
1 Assignment
0 Petitions
Accused Products
Abstract
Disclosed herein is a method of estimating a geometrical relationship between a first image (101) and a second image (102), wherein the second image (102) includes a noise component. The method determines a location and size of each one of a plurality of image patches (201), based on the noise component included in the second image (102) and correlation information derived from the first image (101). The method then identifies a plurality of first image areas in the first image and a corresponding plurality of second image areas in the second image, based on the location and size of each one of the plurality of image patches. Each first image area of the first image (101) corresponds to a related second image area of the second image (102). The method then determines a geometrical relationship between the first and second images (101, 102) by comparing, for each one of the first image areas, information located within the first image area with information located within the corresponding related second image area.
-
Citations
17 Claims
-
1. A computer-implementable method of estimating a geometrical relationship between a first image and a second image, said second image including a noise component, said method comprising the steps of:
-
(a) determining a location and size of each one of a plurality of image patches for each of said images, based on the noise component included in said second image and correlation information derived from said first image, wherein the size of each of the plurality of image patches is determined using a Cramer-Rao lower bound of two-dimensional shift estimation; (b) identifying a plurality of first image areas in said first image and a corresponding plurality of second image areas in said second image, based on the location and size of each one of said plurality of image patches, wherein each first image area of said first image corresponds to a related second image area of said second image; and (c) determining a geometrical relationship between said first and second images by comparing, for each one of said first image areas, information located within said first image area with information located within said corresponding related second image area. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)
-
-
16. A non-transitory computer readable storage medium having recorded thereon a computer program executable by a processor to estimate a geometrical relationship between a first image and a second image, said second image including a noise component, said program comprising:
-
(a) code for determining a location and size of each one of a plurality of image patches for each of said images, based on the noise component included in said second image and correlation information derived from said first image, wherein the size of each of the plurality of image patches is determined using a Cramer-Rao lower bound of two-dimensional shift estimation; (b) code for identifying a plurality of first image areas in said first image and a corresponding plurality of second image areas in said second image, based on the location and size of each one of said plurality of image patches, wherein each first image area of said first image corresponds to a related second image area of said second image; and (c) code for determining a geometrical relationship between said first and second images by comparing, for each one of said first image areas, information located within said first image area with information located within said corresponding related second image area.
-
-
17. A computer-implementable method of estimating a geometrical relationship between a first image and a second image, said second image including a noise component, said method comprising the steps of:
-
(a) determining a location and size of each one of a plurality of image patches for each of said images, based on the noise component included in said second image and correlation information derived from said first image, wherein a first image patch, having a location corresponding to a first region of said first image, has a smaller size than a second image patch having a location corresponding to a second region of said first image, when said first region contains more correlation information than said second region; (b) identifying a plurality of first image areas in said first image and a corresponding plurality of second image areas in said second image, based on the location and size of each one of said plurality of image patches, wherein each first image area of said first image corresponds to a related second image area of said second image; and (c) determining a geometrical relationship between said first and second images by comparing, for each one of said first image areas, information located within said first image area with information located within said corresponding related second image area.
-
Specification