Methods and apparatus for shifting perspective in a composite image

• US 7,103,236 B2
• Filed: 08/28/2001
• Issued: 09/05/2006
• Est. Priority Date: 08/28/2001
• Status: Active Grant

• Current Assignee: Adobe Inc.
• Original Assignee: Adobe Systems Incorporated (Adobe Inc.)
• Inventors: Peterson, John
• Primary Examiner(s): Mehta, Bhavesh M.
• Assistant Examiner(s): Rosario, Dennis
• Application Number:
  US09/941,427
• Publication Number:
  US 20030048959A1
• Time in Patent Office: 1,834 Days
• Field of Search: 382/284, 382/293, 382/294, 382/154, 348/39, 348 5- 80
• US Class Current: 382/294
• CPC Class Codes: G06T 3/4038 Image mosaicing, e.g. composing plane images from plane sub-images

Abstract

Methods and apparatus for shifting perspective in a composite image derived from a plurality of images including a first image as a center of projection, and a modified version of a second image that is corrected for perspective distortion relative to the first image. A user input specifies a change in perspective to make the second image the center of projection. In response, a transformation is determined for mapping reference points in the modified version of the second image to reference points in the original, uncorrected second image. The transformation is applied to each of the plurality of images in the composite image, and the transformed images are merged to form a second composite image that has the second image as its center of projection. The methods and apparatus can be implemented as an interactive tool capable of changing perspective based on a single user input.

Claims

1. A method for shifting a perspective of a composite image derived from a plurality of component images including a first image and a second image, the composite image including derived component images derived from the component images, the derived component images including the first image as a center of projection and a modified version of the second image, the modified version of the second image having been corrected for perspective distortion relative to the first image, the method comprising:

receiving an instruction to shift the perspective of the composite image to make the second image the center of projection;

determining a transformation for mapping a set of reference points in the modified version of the second image to a corresponding set of reference points in the second image; and

applying the transform to each of the plurality of derived component images in the composite image to generate the second image and a plurality of modified component images corrected for perspective distortion relative to the second image, each of the modified component images having the second image as their center of projection,

wherein the first image includes a plurality of pixels and has a perimeter that defines a set of vertices; and

applying the transform to the first image based on the transformation includes;

transforming the vertices of the first image; and

transforming the pixels of the first image based on the transformation of the vertices.

2. The method of claim 1, further comprising:

merging a modified version of the first image and the second image to form a second composite image that has the second image as its center of projection.

3. The method of claim 1, wherein the plurality of images includes a third image and the composite image includes a first modified version of the third image, the first modified version of the third image being corrected for perspective distortion relative to the first image, the method further comprising:

transforming the first modified version of the third image based on the transformation to derive a second modified version of the third image, the second modified version of the third image being corrected for perspective distortion relative to the second image; and

merging the modified version of the first image, the second image, and the second modified version of the third image to form a second composite image.

4. The method of claim 1, wherein:

the reference points in the modified version of the second image include four non-collinear and non-coincident points in the modified version of the second image; and

the reference points in the second image include four non-collinear and non-coincident points in the second image.

5. The method of claim 1, wherein:

the second image and the modified version of the second image each include a perimeter; and

the reference points in the second image and the modified version of the second image are vertices on the perimeters of the second image and the modified version of the second image.

6. The method of claim 1, wherein:

the transformation is represented as a transformation matrix.

7. The method of claim 6, wherein:

the transformation matrix is derived from the vertices of the modified version of the second image.

8. The method of claim 7, wherein:

the transformation matrix is further derived from the vertices of the second image.

9. The method of claim 8, wherein the transformation matrix, M, is given by:

10. The method of claim 1, wherein:

the reference points in the modified version of the second image are corner points on a trapezoid formed by a perimeter of the modified version of the second image; and

the reference points in the second image are corner points on a rectangle formed by a perimeter of the second image.

11. The method of claim 1, wherein:

applying the transform to the modified version of the second image alters the shape of a perimeter of the modified version of the second image by moving at least one reference point relative to at least one other reference point.

12. The method of claim 1, wherein:

the modified version of the second image has a perimeter forming a trapezoid; and

transforming the modified version of the second image alters the perimeter of the modified version of the second image to form a rectangle.

13. The method of claim 1, wherein:

the instruction to shift perspective is received as a single user input; and

the determining a transformation and applying the transform are automatically performed in response to the user input.

14. A computer-implemented image processing method, comprising:

providing a composite image derived from a plurality of component images including a first image and a second image, the composite image including derived component images derived from the component images, the derived component images including the first image as a center of projection and a modified version of the second image, the modified version of the second image having been corrected for perspective distortion relative to the first image;

receiving a single user input including an instruction to change a perspective of the composite image to make the second image the center of projection; and

in response to the input, automatically;

determining a transformation for mapping a set of reference points in the modified version of the second image to a set of reference points in the second image;

applying the transform to each of the plurality of derived component images in the composite image to generate the second image and a plurality of modified component images, each of the modified component images having the second image as their center of projection, each of the modified component images being corrected for perspective distortion relative to the second image; and

merging the second image and the plurality of the transformed component images corrected for perspective distortion relative to the second image to form a second composite image that has the second image as its center of projection,

wherein the first image includes a plurality of pixels and has a perimeter that defines a set of vertices; and

applying the transform to the first image based on the transformation includes;

transforming the vertices of the first image; and

transforming the pixels of the first image based on the transformation of the vertices.

15. A computer program product, tangibly stored on a computer-readable medium, for shifting a perspective of a composite image derived from a plurality of component images including a first image and a second image, the composite image including derived component images derived from the component images, the derived component images including the first image as a center of projection and a modified version of the second image, the modified version of the second image having been corrected for perspective distortion relative to the first image, the product comprising instructions operable to cause a programmable processor to:

receive an instruction to shift the perspective of the composite image to make the second image the center of projection;

determine a transformation for mapping a set of reference points in the modified version of the second image to a corresponding set of reference points in the second image; and

apply the transform to each of the plurality of derived component images in the composite image to generate the second image and a plurality of modified component images corrected for perspective distortion relative to the second image, each of the modified component images having the second image as their center of projection,

wherein the first image includes a plurality of pixels and has a perimeter that defines a set of vertices; and

the instructions to apply the transform to the first image based on the transformation include instructions to;

transform the vertices of the first image; and

transform the pixels of the first image based on the transformation of the vertices.

16. The computer program product of claim 15, further comprising instructions operable to cause a programmable processor to:

merge a modified version of the first image and the second image to form a second composite image that has the second image as its center of projection.

17. The computer program product of claim 15, wherein the plurality of images includes a third image and the composite image includes a first modified version of the third image, the first modified version of the third image being corrected for perspective distortion relative to the first image, the product further comprising instructions operable to cause a programmable processor to:

transform the first modified version of the third image based on the transformation to derive a second modified version of the third image, the second modified version of the third image being corrected for perspective distortion relative to the second image; and

merge the modified version of the first image, the second image, and the second modified version of the third image to form a second composite image.

18. The computer program product of claim 15, wherein:

the reference points in the modified version of the second image include four non-collinear and non-coincident points in the modified version of the second image; and

the reference points in the second image include four non-collinear and non-coincident points in the second image.

19. The computer program product of claim 15, wherein:

the second image and the modified version of the second image each include a perimeter; and

the reference points in the second image and the modified version of the second image are vertices on the perimeters of the second image and the modified version of the second image.

20. The computer program product of claim 15, wherein:

the transformation is represented as a transformation matrix.

21. The computer program product of claim 20, wherein:

the transformation matrix is derived from the vertices of the modified version of the second image.

22. The computer program product of claim 21, wherein:

the transformation matrix is further derived from the vertices of the second image.

23. The computer program product of claim 22, wherein the transformation matrix, M, is given by:

24. The computer program product of claim 15, wherein:

the reference points in the modified version of the second image are corner points on a trapezoid formed by a perimeter of the modified version of the second image; and

the reference points in the second image are corner points on a rectangle formed by a perimeter of the second image.

25. The computer program product of claim 15, wherein:

the instructions operable to cause a programmable processor to apply the transform to the modified version of the second image are operable to cause the programmable processor to alter the shape of a perimeter of the modified version of the second image by moving at least one reference point relative to at least one other reference point.

26. The computer program product of claim 15, wherein:

the modified version of the second image has a perimeter forming a trapezoid; and

the instructions operable to cause a programmable processor to transform the modified version of the second image are operable to cause the programmable processor to alter the perimeter of the modified version of the second image to form a rectangle.

27. The computer program product of claim 15 wherein:

the instruction to shift perspective is received as a single user input; and

the product includes instructions to cause the programmable processor to determine the transformation and transform the images automatically in response to the user input.

28. A computer program product, tangibly stored on a computer-readable medium, for processing an image, comprising instructions operable to cause a programmable processor to:

receive a composite image derived from a plurality of component images including a first image and a second image, the composite image including derived component images derived from the component images, the derived component images including the first image as a center of projection and a modified version of the second image, the modified version of the second image having been corrected for perspective distortion relative to the first image;

receive a single user input including an instruction to change a perspective of the composite image to make the second image the center of projection; and

in response to the input, automatically;

determine a transformation for mapping a set of reference points in the modified version of the second image to a set of reference points in the second image;

apply the transform to each of the plurality of derived component images in the composite image to generate the second image and a plurality of modified component images, each of the modified component images having the second image as theft center of projection, each of the modified component images being corrected for perspective distortion relative to the second image; and

merge the second image and the plurality of modified component images corrected for perspective distortion relative to the second image to form a second composite image that has the second image as its center of projection,

wherein the first image includes a plurality of pixels and has a perimeter that defines a set of vertices; and

the instructions to apply the transform to the first image based on the transformation include instructions to;

transform the vertices of the first image; and

transform the pixels of the first image based on the transformation of the vertices.