1 / 39

# 3D Vision - PowerPoint PPT Presentation

3D Vision. Interest Points. correspondence and alignment. Correspondence: matching points, patches, edges, or regions across images. ≈. correspondence and alignment. Alignment: solving the transformation that makes two things match better. T.

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

## PowerPoint Slideshow about ' 3D Vision' - dominique-derick

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

### 3D Vision

Interest Points

• Correspondence: matching points, patches, edges, or regions across images

• Alignment: solving the transformation that makes two things match better

T

Example: tracking points corresponds two views

x

x

x

frame 22

frame 0

frame 49

Your problem 1 for HW 2!

Human eye movements corresponds two views

Human eye movements corresponds two views

Interest points corresponds two views

• Suppose you have to click on some point, go away and come back after I deform the image, and click on the same points again.

• Which points would you choose?

original

deformed

Overview of Keypoint Matching corresponds two views

1. Find a set of distinctive key- points

2. Define a region around each keypoint

A1

B3

3. Extract and normalize the region content

A2

A3

B2

B1

4. Compute a local descriptor from the normalized region

5. Match local descriptors

Goals for Keypoints corresponds two views

Detect points that are repeatable and distinctive

Choosing interest points corresponds two views

Where would you tell your friend to meet you?

Moravec corner detector (1980) corresponds two views

• We should easily recognize the point by looking through a small window

• Shifting a window in anydirection should give a large change in intensity

Moravec corner detector corresponds two views

flat

Moravec corner detector corresponds two views

flat

Moravec corner detector corresponds two views

flat

edge

Moravec corner detector corresponds two views

corner

isolated point

flat

edge

Window function corresponds two views

Shifted intensity

Intensity

Moravec corner detector

Change of intensity for the shift [u,v]:

Four shifts: (u,v) = (1,0), (1,1), (0,1), (-1, 1)

Look for local maxima in min{E}

Problems of Moravec detector corresponds two views

• Noisy response due to a binary window function

• Only a set of shifts at every 45 degree is considered

• Responds too strong for edges because only minimum of E is taken into account

• Harris corner detector (1988) solves these problems.

Harris corner detector corresponds two views

Noisy response due to a binary window function

• Use a Gaussian function

Harris corner detector corresponds two views

Only a set of shifts at every 45 degree is considered

• Consider all small shifts by Taylor’s expansion

Harris corner detector corresponds two views

Equivalently, for small shifts [u,v] we have a bilinear approximation:

, where M is a 22 matrix computed from image derivatives:

Harris corner detector corresponds two views

Responds too strong for edges because only minimum of E is taken into account

• A new corner measurement

Harris corner detector corresponds two views

Intensity change in shifting window: eigenvalue analysis

1, 2 – eigenvalues of M

direction of the fastest change

Ellipse E(u,v) = const

direction of the slowest change

(max)-1/2

(min)-1/2

Harris corner detector corresponds two views

2

edge 2 >> 1

Corner

1 and 2 are large,1 ~ 2;E increases in all directions

Classification of image points using eigenvalues of M:

1 and 2 are small;E is almost constant in all directions

edge 1 >> 2

flat

1

Harris corner detector corresponds two views

Measure of corner response:

(k – empirical constant, k = 0.04-0.06)

Another view corresponds two views

Another view corresponds two views

Another view corresponds two views

Harris corner detector (input) corresponds two views

Corner response R corresponds two views

Threshold on R corresponds two views

Local maximum of R corresponds two views

Harris corner detector corresponds two views

Harris Detector: Summary corresponds two views

• Average intensity change in direction [u,v] can be expressed as a bilinear form:

• Describe a point in terms of eigenvalues of M:measure of corner response

• A good (corner) point should have a large intensity change in all directions, i.e. R should be large positive

threshold corresponds two views

R

R

x(image coordinate)

x(image coordinate)

Harris Detector: Some Properties

• Partial invariance to affine intensity change

• Only derivatives are used => invariance to intensity shift I I+b

• Intensity scale: I  aI

Harris Detector: Some Properties corresponds two views

• Rotation invariance

Ellipse rotates but its shape (i.e. eigenvalues) remains the same

Corner response R is invariant to image rotation

Harris Detector is rotation invariant corresponds two views

Repeatability rate:

# correspondences# possible correspondences

Harris Detector: Some Properties corresponds two views

• But: non-invariant to image scale!

All points will be classified as edges

Corner !

Harris Detector: Some Properties corresponds two views

• Quality of Harris detector for different scale changes

Repeatability rate:

# correspondences# possible correspondences