Image processing methods and apparatus for separable, general affine transformation of an image

• US 6,215,915 B1
• Filed: 02/20/1998
• Issued: 04/10/2001
• Est. Priority Date: 02/20/1998
• Status: Expired due to Term

• Current Assignee: Cognex Corporation
• Original Assignee: Cognex Corporation
• Inventors: Reyzin, Igor
• Primary Examiner(s): Mehta, Bhavesh
• Assistant Examiner(s): PATEL, KANJIBHAI B
• Application Number:
  US09/027,432
• Time in Patent Office: 1,145 Days
• Field of Search: 382/296, 382/276, 382/295, 382/300, 382/277, 345/436, 345/437, 345/438, 345/432, 345/475
• US Class Current: 382/296
• CPC Class Codes: G06T 3/00 Geometric image transformations in the plane of the image

Abstract

A separable, general affine transformation according to the invention permits an image to be concurrently rotated, scaled, translated, skewed, sheared, or otherwise transformed via a sequence of one-dimensional transformations. A general affine transformation of an image in two dimensions is achieved by generating an “intermediate” image via affine transformation of the source along a first axis. The intermediate image is then subjected to affine transformation along a second axis, e.g., perpendicular to the first. The resultant image may be used in place of that which would be produced by a single two-dimensional transformation of the source image (e.g., in a single pass).

Claims

1. An image processing method of general, separable affine transformation of a source image, the method comprising the steps of

A. generating an intermediate image by affine transformation of the source image along a first axis relative to the source image in accord with a partial transformation matrix L, wherein the partial transformation matrix L is a left triangular matrix;

B. generating a destination image by affine transformation of the intermediate image along a second axis relative to the source image in accord with a partial transformation matrix U, wherein the partial transformation U is a right triangular matrix;

C. where coordinates in the source image are related to coordinates in the destination image in accord with the relations;

$\left[\begin{array}{c}{x}_s\\ y_s\end{array}\right]=M·<br><br>\left[\begin{array}{c}{x}_d\\ y_d\end{array}\right]+\left[\begin{array}{c}{x}_o\\ y_o\end{array}\right]$$M=\left[\begin{array}{cc}{e}_{11}& e_{12}\\ e_{21}& e_{22}\end{array}\right]$

where

(x_d,y_d) represents a coordinate in the destination image;

(x_s,y_s) represents a coordinate in the source image;

(x₀,y₀) is an offset to be effected by the transformation;

e₁₁, e₁₂, e₂₁, and e₂₂ are elements of a general affine transformation matrix M; and

D. where L and U are decompositions of the transformation matrix M, such that M=L·

U and

where the destination image represents a general affine transformation of the source image.

2. A method according to claim 1, wherein step (A) comprises

i. defining a mapping between coordinates in the intermediate image and those in the source image; and

ii. determining a pixel intensity for at least a selected coordinate in the intermediate image as a function of intensities of one or more pixels in a vicinity of a coordinate in the source image to which the selected coordinate maps.

3. A method according to claim 2, wherein step (A)(i) comprises defining a mapping between locations in the intermediate image having two integer coordinates and those in the source image having at least one integer coordinate.

4. A method according to claim 2, wherein step (A)(ii) comprises determining a pixel intensity for the selected coordinate in the intermediate image by any of interpolation and extrapolation of the intensities of one or more pixels in the vicinity of the coordinate in the source image to which the selected coordinate maps.

5. A method according to claim 1, wherein step (B) comprises

i. defining a mapping between coordinates in the destination image and those in the intermediate image;

and

ii. determining a pixel intensity for at least a selected coordinate in the destination image as a function of intensities of one or more pixels in a vicinity of a coordinate in the intermediate image to which the selected coordinate maps.

6. A method according to claim 5, wherein step (B)(i) comprises defining a mapping between locations in the destination image having two integer coordinates and those in the source image having at least one integer coordinate.

7. A method according to claim 5, wherein step (B)(ii) comprises determining a pixel intensity for the selected coordinate in the destination image by any of interpolation and extrapolation of the intensities of one or more pixels in the vicinity of the coordinate in the intermediate image to which the selected coordinate maps.

8. A method according to claim 1, wherein the first axis and the second axis are substantially perpendicular to one another.

9. A method according to claim 1, wherein the partial transformation matrix L is in accord with the relation $L=\left[\begin{array}{cc}{l}_{11}& 0\\ l_{21}& l_{22}\end{array}\right]$

 

wherein l₁₁, l₂₁ and l₂₂ are parameters of L and wherein l₁₁=1.

10. A method according to claim 1, wherein the partial transformation matrix U is in accord with the relation $U=\left[\begin{array}{cc}{u}_{11}& u_{12}\\ 0& u_{22}\end{array}\right]$

 

wherein u₁₁, u₁₂ and u₂₂ are parameters of U and wherein u₂₂=1.

11. An image processing method for general, separable affine transformation of a source image, the method comprising the steps of

A. generating an intermediate image by affine transformation of the source image along a first axis, such generating step including

i. defining a mapping between coordinates in the intermediate image and coordinates in the source image, wherein said step of defining the mapping includes the steps of (1) selecting a plurality of coordinates for the intermediate image and (2) calculating for each of those selected coordinates a corresponding coordinate in the source image,

ii. determining a pixel intensity for at least a selected coordinate in the intermediate image as a function of intensities of one or more pixels in a vicinity of a coordinate in the source image to which the selected coordinate maps,

B. generating a destination image by affine transformation of the intermediate image along a second axis, such generating step including

i. defining a mapping between coordinates in the destination image and those in the intermediate image in accord with the relations

12. A method according to claim 11, wherein

step (A)(i) comprises defining a mapping between locations in the intermediate image having two integer coordinates and those in the source image having at least one integer coordinate, and

step (B)(i) comprises defining a mapping between locations in the destination image having two integer coordinates and those in the source image having at least one integer coordinate.

13. A method according to claim 11, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

14. A method according to claim 11, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

15. A method according to claim 11, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

16. A method according to claim 11, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

17. An image processing system for carrying out a method for general, separable affine transformation of a source image, the method comprising the steps of

A. generating an intermediate image by affine transformation of the source image along a first axis relative to the source image in accord with partial transformation matrix L, wherein the partial transformation matrix L is a left triangular matrix, and

B. generating a destination image by affine transformation of the intermediate image along a second axis relative to the source image in accord with a partial transformation matrix U, wherein the partial transformation matrix U is a right triangular matrix;

C. where coordinates in the source image are related to coordinates in the destination image in accord with the relations;

$\left[\begin{array}{c}{x}_s\\ y_s\end{array}\right]=M·<br><br>\left[\begin{array}{c}{x}_d\\ y_d\end{array}\right]+\left[\begin{array}{c}{x}_o\\ y_o\end{array}\right]$$M=\left[\begin{array}{cc}{e}_{11}& e_{12}\\ e_{21}& e_{22}\end{array}\right]$

where

(x_d, y_d) represents a coordinate in the destination image;

(x_s, y_s) represents a coordinate in the source image;

(x_o, y_o) is an offset to be effected by the transformation;

e₁₁, e₁₂, e₂₁, and e₂₂ are elements of a general affine transformation matrix M, and

D. where L and U are decompositions of a transformation matrix M, such that M=L·

U, and

where the destination image represents a general affine transformation of the source image.

18. A system according to claim 17, wherein step (A) comprises

i. defining a mapping between coordinates in the intermediate image and those in the source image; and

ii. determining a pixel intensity for at least a selected coordinate in the intermediate image as a function of intensities of one or more pixels in a vicinity of a coordinate in the source image to which the selected coordinate maps.

19. A system according to claim 18, wherein step (A)(i) comprises defining a mapping between locations in the intermediate image having two integer coordinates and those in the source image having at least one integer coordinate.

20. A system according to claim 18, wherein step (A)(ii) comprises determining a pixel intensity for the selected coordinate in the intermediate image by any of interpolation and extrapolation of the intensities of one or more pixels in the vicinity of the coordinate in the source image to which the selected coordinate maps.

21. A system according to claim 17, wherein step (B) comprises

i. defining a mapping between coordinates in the destination image and those in the intermediate image;

and

ii. determining a pixel intensity for at least a selected coordinate in the destination image as a function of intensities of one or more pixels in a vicinity of a coordinate in the intermediate image to which the selected coordinate maps.

22. A system according to claim 21, wherein step (B)(i) comprises defining a mapping between locations in the destination image having two integer coordinates and those in the source image having at least one integer coordinate.

23. A system according to claim 21, wherein step (B)(ii) comprises determining a pixel intensity for the selected coordinate in the destination image by any of interpolation and extrapolation of the intensities of one or more pixels in the vicinity of the coordinate in the intermediate image to which the selected coordinate maps.

24. A system according to claim 17, wherein the first axis and the second axis are substantially perpendicular to one another.

25. A system according to claim 24, wherein the first axis and the second axis are substantially aligned with respective axes of the destination image.

26. A system according to claim 17, wherein the partial transformation matrix L is in accord with the relation $L=\left[\begin{array}{cc}{l}_{11}& 0\\ l_{21}& l_{22}\end{array}\right]$

 

wherein l₁₁, l₂₁ and l₂₂ are parameters of L and wherein l₁₁=1.

27. A system according to claim 17, wherein the partial transformation matrix U is in accord with the relation $U=\left[\begin{array}{cc}{u}_{11}& u_{12}\\ 0& u_{22}\end{array}\right]$

 

wherein u₁₁, u₁₂ and u₂₂ are parameters of U and wherein u₂₂=1.

28. An image processing system for carrying out a method for general, separable affine transformation of a source image, the method comprising the steps of

A. generating an intermediate image by affine transformation of the source image along a first axis, such generating step including

i. defining a mapping between coordinates in the intermediate image and those in the source image, wherein said step of defining the mapping includes the steps of (1) selecting a plurality of coordinates for the intermediate image and (2) calculating for each of those selected coordinates a corresponding coordinate in the source image;

and

ii. determining a pixel intensity for at least a selected coordinate in the intermediate image as a function of intensities of one or more pixels in a vicinity of a coordinate in the source image to which the selected coordinate maps,

B. generating a destination image by affine transformation of the intermediate image along a second axis, such generating step including

i. defining a mapping between coordinates in the destination image and those in the intermediate image in accord with the relations

29. A system according to claim 28, wherein

step (A)(i) comprises defining a mapping between locations in the intermediate image having two integer coordinates and those in the source image having at least one integer coordinate, and

step (B)(i) comprises defining a mapping between locations in the destination image having two integer coordinates and those in the source image having at least one integer coordinate.

30. A system according to claim 28, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

31. A system according to claim 28, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

32. A system according to claim 28, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

33. A system according to claim 28, wherein the defining step comprises determining mappings between coordinates in the source image and those of the intermediate image in accord with the relations

34. An image processing method for general, separable affine transformation of a source image, the method comprising the steps of

A. generating an intermediate image by affine transformation of the source image along a first axis, such generating step including

i. defining a mapping between coordinates in the intermediate image and those in the source image in accord with the relations

35. A method according to claim 34, wherein

step (A)(i) comprises defining a mapping between locations in the intermediate image having two integer coordinates and those in the source image having at least one integer coordinate, and

step (B)(i) comprises defining a mapping between locations in the destination image having two integer coordinates and those in the source image having at least one integer coordinate.

36. A method according to claim 34, wherein defining step (B)(i) comprises determining mappings between coordinates in the intermediate image and those of the destination image in accord with the relations

37. An image processing system for carrying out a method for general, separable affine transformation of a source image, the method comprising the steps of

A. generating an intermediate image by affine transformation of the source image along a first axis, such generating step including

i. defining a mapping between coordinates in the intermediate image and those in the source image in accord with the relations

38. A system according to claim 37, wherein

step (A)(i) comprises defining a mapping between locations in the intermediate image having two integer coordinates and those in the source image having at least one integer coordinate, and

step (B)(i) comprises defining a mapping between locations in the destination image having two integer coordinates and those in the source image having at least one integer coordinate.

39. A system according to claim 37, wherein defining step (B)(i) comprises determining mappings between coordinates in the intermediate image and those of the destination image in accord with the relations