Granted Patent B2

US 9,147,226 · App. 14/012,006 · Granted Sep 29, 2015

Abstract

Claims (73)

1. A method comprising:

extracting a set of image features from a first image and a corresponding set of image features from a second image;

computing a first transformation matrix between the first image and the second image based on a matching of the set of image features in the first image with the corresponding set of image features in the second image;

computing a second transformation matrix between the first image and the second image based on the matching of the set of image features in the first image with the corresponding set of image features in the second image;

determining a first set of feature points based on the first transformation matrix, and a second set of feature points based on the second transformation matrix;

estimating a transformation matrix between the first set of feature points and the second set of feature points; and

aligning the second image with the first image based on the transformation matrix;

wherein the method further comprises:

determining a set of first transformation matrices and a set of second transformation matrices,

wherein the set of first transformation matrices comprises the first transformation matrix between adjacent images of the first image, adjacent images of the second image and adjacent images of at least one subsequent image, and wherein the set of first transformation matrices include a set of homography transformation matrices, and

wherein the set of second transformation matrices comprises the second transformation matrix between the adjacent images of the first image, adjacent images of the second image and adjacent images of the at least one subsequent image, and wherein the set of second transformation matrices include a set of similarity transformation matrices.

2. The method as claimed in claim 1 further comprising:

determining a set of matching image features between the set of image features from the first image and the corresponding set of image features from the second image, the set of matching image features are used for computing the first transformation matrix and the second transformation matrix.

3. The method as claimed in claim 1 , wherein feature points of the first set of feature points and feature points of the second set of feature points comprises at least one of inliers and corner points.

4. The method as claimed in claim 1 , wherein estimating the transformation matrix is performed based on a feature matching between the first set of feature points and the second set of feature points until a mismatch error is determined to be less than a predetermined threshold value of mismatch error.

5. The method as claimed in claim 1 , further comprising generating a panorama image based on the first image and the aligned second image.

6. The method as claimed in claim 1 , further comprising:

generating a first pre-processed image based on the first transformation matrix; and

generating a second pre-processed image based on the second transformation matrix.

7. The method as claimed in claim 6 , wherein determining the first set of feature points and the second set of feature points comprises determining feature points in the first pre-processed image and the second pre-processed image.

8. The method as claimed in claim 7 , further comprising:

performing feature matching between the first set of feature points and the second set of feature points;

computing the transformation matrix based on the feature matching;

computing a mismatch error based on the transformation matrix; and

modifying the computed transformation matrix based on a determination of the mismatch error being greater than a predetermined threshold value of mismatch error.

9. An apparatus comprising:

at least one processor; and

at least one memory comprising computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least perform:

extract a set of image features from a first image and a corresponding set of image features from a second image;

compute a first transformation matrix between the first image and the second image based on a matching of the set of image features in the first image with the corresponding set of image features in the second image;

compute a second transformation matrix between the first image and the second image based on the matching of the set of image features in the first image with the corresponding set of image features in the second image;

determine a first set of feature points based on the first transformation matrix, and a second set of feature points based on the second transformation matrix;

estimate a transformation matrix between the first set of feature points and the second set of feature points; and

align the second image with the first image based on the transformation matrix;

wherein the apparatus is further caused, at least in part, to:

determine a set of first transformation matrices and a set of second transformation matrices,

wherein the set of first transformation matrices comprises the first transformation matrix between adjacent images of the first image, adjacent images of the second image and adjacent images of at least one subsequent image, and wherein the set of first transformation matrices include a set of homography transformation matrices, and

wherein the set of second transformation matrices comprises the second transformation matrix between adjacent images of the first image, adjacent images of the second image and adjacent images of the at least one subsequent image, and wherein the set of second transformation matrices include a set of similarity transformation matrices.

10. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part, to:

determine a set of matching image features between the set of image features from the first image and the corresponding set of image features from the second image, the set of matching image features are used for computing the first transformation matrix and the second transformation matrix.

11. The apparatus as claimed in claim 9 , wherein feature points of the first set of feature points and feature points of the second set of feature points comprises at least one of inliers and corner points.

12. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part, to:

perform estimation of the transformation matrix based on a feature matching between the first set of feature points and the second set of feature points until a mismatch error is determined to be less than a predetermined threshold value of mismatch error.

13. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part, to:

generate a panorama image based on the first image and the aligned second image.

14. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part, to:

generate a first pre-processed image based on the first transformation matrix; and generate a second pre-processed image based on the second transformation matrix.

15. The apparatus as claimed in claim 14 , wherein the apparatus is further caused, at least in part, to:

determine the first set of feature points and the second set of feature points based on determining feature points in the first pre-processed image and the second pre-processed image.

16. The apparatus as claimed in claim 15 , wherein the apparatus is further caused, at least in part, to:

perform feature matching between the first set of feature points and the second set of feature points;

compute the transformation matrix based on the feature matching;

compute a mismatch error based on the transformation matrix; and

modify the computed transformation matrix based on a determination of the mismatch error being greater than a predetermined threshold value of mismatch error.

17. A computer program product comprising at least one non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium comprising a set of instructions, which, when executed by one or more processors, cause an apparatus to at least perform:

extract a set of image features from a first image and a corresponding set of image features from a second image;

compute a first transformation matrix between the first image and the second image based on a matching of the set of image features in the first image with the corresponding set of image features in the second image;

compute a second transformation matrix between the first image and the second image based on a matching of the set of image features in the first image with the corresponding set of image features in the second image;

determine a first set of feature points based on the first transformation matrix, and a second set of feature points based on the second transformation matrix;

estimate a transformation matrix between the first set of feature points and the second set of feature points; and

align the second image with the first image based on the transformation matrix;

wherein the apparatus is further caused, at least in part, to:

determine a set of first transformation matrices and a set of second transformation matrices,

wherein the set of first transformation matrices comprises the first transformation matrix between adjacent images of the first image, adjacent images of the second image and adjacent images of at least one subsequent image, and wherein the set of first transformation matrices include a set of homography transformation matrices, and

wherein the set of second transformation matrices comprises the second transformation matrix between adjacent images of the first image, adjacent images of the second image and adjacent images of the at least one subsequent image, and wherein the set of second transformation matrices include a set of similarity transformation matrices.

18. The computer program product as claimed in claim 17 , wherein the apparatus is further caused, at least in part, to:

determine a set of matching image features between the set of image features from the first image and the corresponding set of image features from the second image, the set of matching image features being configured to be utilized for computing the first transformation matrix and the second transformation matrix.

19. The method as claimed in claim 1 , further comprising:

generating a first pre-processed image based on the set of first transformation matrices; and

generating a second pre-processed image based on the set of second transformation matrices.

20. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part, to:

generate a first pre-processed image based on the set of first transformation matrices; and

generate a second pre-processed image based on the set of second transformation matrices.

Assignments (8)