1 / 31

# - PowerPoint PPT Presentation

Mosaics con’t. CSE 455, Winter 2010 February 10 , 2010. Announcements. The Midterm: Due this Friday, Feb 12, at the beginning of class Late exams will not be accepted Additional Office Hour today: 2:30 to 3:30 in CSE 212 (the normal place). Review From Last Time. How to do it?.

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.

## PowerPoint Slideshow about '' - akiko

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

Mosaics con’t

CSE 455, Winter 2010

February 10, 2010

• The Midterm:

• Due this Friday, Feb 12, at the beginning of class

• Late exams will not be accepted

• 2:30 to 3:30 in CSE 212 (the normal place)

• Similar to Structure from Motion, but easier

• Basic Procedure

• Take a sequence of images from the same position

• Rotate the camera about its optical center

• Compute transformation between second image and first

• Transform the second image to overlap with the first

• Blend the two together to create a mosaic

• If there are more images, repeat

Input

Goal

• How to account for warping?

• Translations are not enough to align the images

• Homographies!!!

x4

x1

x3

x2

x5

x7

x6

Structure from Motion: Image reprojection

p1,1

p1,3

p1,2

Image 1

Image 3

R3,t3

R1,t1

Image 2

R2,t2

Panoramas: Image reprojection

x4

x1

x3

x2

x5

x7

x6

p1,2

Image 3

Image 2

Image 1

R3

R2

R1

mosaic plane

Image reprojection

• The mosaic has a natural interpretation in 3D

• The images are reprojected onto a common plane

• The mosaic is formed on this plane

where no pixel

maps to

Image warping with homographies

image plane in front

image plane below

• Take a sequence of images from the same position

• Rotate the camera about its optical center

• Compute correspondence between second image and first

• Compute transformation between second image and first

• Transform the second image to overlap with the first

• Blend the two together to create a mosaic

• If there are more images, repeat

ü

• Take a sequence of images from the same position

• Rotate the camera about its optical center

• Compute correspondence between second image and first

• Compute transformation between second image and first

• Transform the second image to overlap with the first

• Blend the two together to create a mosaic

• If there are more images, repeat

ü

• Compute correspondence between second image and first

• Compute transformation between second image and first

• What kind of transformation

• Homography!!!

• Do we know the correspondence?

• Guess some matches

• Compute a transformation using those matches

• Check if the transformation is good

• Randomly choose a set of K potential correspondences

• Typically K is the minimum size that lets you fit a model

• How many for a

• Translation

• rotation?

• Affine?

• Homography?

• Fit a model (e.g., translation, homography) to those correspondences

• Count the number of inliers that “approximately” fit the model

• Need a threshold on the error

• Repeat as many times as you can

• Choose the model that has the largest set of inliers

• Refine the model by doing a least squares fit using ALL of the inliers

RAndom SAmple Consensus

Select one match, count inliers(in this case, only one)

RAndom SAmple Consensus

Select one match, count inliers(4 inliers)

Find “average” translation vectorfor largest set of inliers

• Same basic approach works for any transformation

• Translation, rotation, homographies, etc.

• Very useful tool

• General version

• Randomly choose a set of K correspondences

• Typically K is the minimum size that lets you fit a model

• Fit a model (e.g., homography) to those correspondences

• Count the number of inliers that “approximately” fit the model

• Need a threshold on the error

• Repeat as many times as you can

• Choose the model that has the largest set of inliers

• Refine the model by doing a least squares fit using ALL of the inliers

• Take a sequence of images from the same position

• Rotate the camera about its optical center

• Compute correspondence between second image and first

• Compute transformation between second image and first

• Transform the second image to overlap with the first

• Blend the two together to create a mosaic

• If there are more images, repeat

ü

ü

• Take a sequence of images from the same position

• Rotate the camera about its optical center

• Compute correspondence between second image and first

• Compute transformation between second image and first

• Transform the second image to overlap with the first

• Blend the two together to create a mosaic

• If there are more images, repeat

ü

ü

• Stitch pairs together, blend, then crop

• Error accumulation

• small errors accumulate over time

=

1

0

1

0

Feathering

1

0

1

Effect of window size

left

right

1

0

1

Effect of window size

1

Good window size

• “Optimal” window: smooth but not ghosted

• Doesn’t always work...