Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp...
-
Upload
clemence-may -
Category
Documents
-
view
221 -
download
0
Transcript of Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp...
![Page 1: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/1.jpg)
Edges
![Page 2: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/2.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Gradients and edges• Points of sharp change in
an image are interesting:– change in reflectance
– change in object
– change in illumination
– noise
• Sometimes called edge points
• Points of sharp change in an image are interesting:– change in reflectance
– change in object
– change in illumination
– noise
• Sometimes called edge points
• General strategy– determine image gradient
– now mark points where gradient magnitude is particularly large wrt neighbours (ideally, curves of such points).
• General strategy– determine image gradient
– now mark points where gradient magnitude is particularly large wrt neighbours (ideally, curves of such points).
![Page 3: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/3.jpg)
CS8690 Computer Vision University of Missouri at Columbia
There are three major issues: 1) The gradient magnitude at different scales is different; which should we choose? 2) The gradient magnitude is large along thick trail; how do we identify the significant points? 3) How do we link the relevant points up into curves?
![Page 4: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/4.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Smoothing and Differentiation• Issue: noise
– smooth before differentiation– two convolutions to smooth, then differentiate?– actually, no - we can use a derivative of Gaussian
filter• because differentiation is convolution, and convolution is
associative
• Issue: noise– smooth before differentiation– two convolutions to smooth, then differentiate?– actually, no - we can use a derivative of Gaussian
filter• because differentiation is convolution, and convolution is
associative
![Page 5: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/5.jpg)
CS8690 Computer Vision University of Missouri at Columbia
The scale of the smoothing filter affects derivative estimates, and alsothe semantics of the edges recovered.
1 pixel 3 pixels 7 pixels
![Page 6: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/6.jpg)
CS8690 Computer Vision University of Missouri at Columbia
We wish to mark points along the curve where the magnitude is biggest.We can do this by looking for a maximum along a slice normal to the curve(non-maximum suppression). These points should form a curve. There arethen two algorithmic issues: at which point is the maximum, and where is thenext one?
![Page 7: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/7.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Non-maximumsuppression
At q, we have a maximum if the value is larger than those at both p and at r. Interpolate to get these values.
![Page 8: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/8.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Predictingthe nextedge point
Assume the marked point is an edge point. Then we construct the tangent to the edge curve (which is normal to the gradient at that point) and use this to predict the next points (here either r or s).
![Page 9: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/9.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Remaining issues• Check that maximum value of gradient value is
sufficiently large– drop-outs? use hysteresis
• use a high threshold to start edge curves and a low threshold to continue them.
• Check that maximum value of gradient value is sufficiently large– drop-outs? use hysteresis
• use a high threshold to start edge curves and a low threshold to continue them.
![Page 10: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/10.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Notice• Something nasty is happening at corners
• Scale affects contrast
• Edges aren’t bounding contours
• Something nasty is happening at corners
• Scale affects contrast
• Edges aren’t bounding contours
![Page 11: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/11.jpg)
CS8690 Computer Vision University of Missouri at Columbia
![Page 12: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/12.jpg)
CS8690 Computer Vision University of Missouri at Columbia
fine scalehigh threshold
![Page 13: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/13.jpg)
CS8690 Computer Vision University of Missouri at Columbia
coarse scale,high threshold
![Page 14: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/14.jpg)
CS8690 Computer Vision University of Missouri at Columbia
coarsescalelowthreshold
![Page 15: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/15.jpg)
CS8690 Computer Vision University of Missouri at Columbia
The Laplacian of Gaussian• Another way to detect an
extremal first derivative is to look for a zero second derivative
• Appropriate 2D analogy is rotation invariant– the Laplacian
• Another way to detect an extremal first derivative is to look for a zero second derivative
• Appropriate 2D analogy is rotation invariant– the Laplacian
• Bad idea to apply a Laplacian without smoothing
– smooth with Gaussian, apply Laplacian
– this is the same as filtering with a Laplacian of Gaussian filter
• Now mark the zero points where there is a sufficiently large derivative, and enough contrast
• Bad idea to apply a Laplacian without smoothing
– smooth with Gaussian, apply Laplacian
– this is the same as filtering with a Laplacian of Gaussian filter
• Now mark the zero points where there is a sufficiently large derivative, and enough contrast
![Page 16: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/16.jpg)
CS8690 Computer Vision University of Missouri at Columbia
sigma=2
sigma=4
contrast=1 contrast=4LOG zero crossings
![Page 17: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/17.jpg)
CS8690 Computer Vision University of Missouri at Columbia
We still have unfortunate behaviourat corners
![Page 18: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/18.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Filters are templates• Applying a filter at some
point can be seen as taking a dot-product between the image and some vector
• Filtering the image is a set of dot products
• Applying a filter at some point can be seen as taking a dot-product between the image and some vector
• Filtering the image is a set of dot products
• Insight – filters look like the effects
they are intended to find
– filters find effects they look like
• Insight – filters look like the effects
they are intended to find
– filters find effects they look like
![Page 19: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/19.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Positive responses
Zero mean image, -1:1 scale Zero mean image, -max:max scale
![Page 20: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/20.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Positive responses
Zero mean image, -1:1 scale Zero mean image, -max:max scale
![Page 21: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/21.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Figure from “Computer Vision for Interactive Computer Graphics,” W.Freeman et al, IEEE Computer Graphics and Applications, 1998 copyright 1998, IEEE
![Page 22: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/22.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Normalized correlation• Think of filters of a dot
product– now measure the angle
– i.e normalised correlation output is filter output, divided by root sum of squares of values over which filter lies
• Think of filters of a dot product– now measure the angle
– i.e normalised correlation output is filter output, divided by root sum of squares of values over which filter lies
• Tricks:– ensure that filter has a zero
response to a constant region (helps reduce response to irrelevant background)
– subtract image average when computing the normalizing constant (i.e. subtract the image mean in the neighbourhood)
– absolute value deals with contrast reversal
• Tricks:– ensure that filter has a zero
response to a constant region (helps reduce response to irrelevant background)
– subtract image average when computing the normalizing constant (i.e. subtract the image mean in the neighbourhood)
– absolute value deals with contrast reversal
![Page 23: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/23.jpg)
CS8690 Computer Vision University of Missouri at Columbia
• Sum of squared differences
• Normalize cross-correlation
• Sum of absolute differences
Region based Similarity Metric
![Page 24: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/24.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Orientation representations• The gradient magnitude
is affected by illumination changes– but it’s direction isn’t
• We can describe image patches by the swing of the gradient orientation
• The gradient magnitude is affected by illumination changes– but it’s direction isn’t
• We can describe image patches by the swing of the gradient orientation
• Important types:– constant window
• small gradient mags
– edge window• few large gradient mags in
one direction
– flow window• many large gradient mags
in one direction
– corner window• large gradient mags that
swing
• Important types:– constant window
• small gradient mags
– edge window• few large gradient mags in
one direction
– flow window• many large gradient mags
in one direction
– corner window• large gradient mags that
swing
![Page 25: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/25.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Representing Windows• Types
– constant• small eigenvalues
– Edge • one medium, one small
– Flow• one large, one small
– corner• two large eigenvalues
• Types– constant
• small eigenvalues
– Edge • one medium, one small
– Flow• one large, one small
– corner• two large eigenvalues
H I I Twindow
![Page 26: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/26.jpg)
CS8690 Computer Vision University of Missouri at Columbia
• Compute eigenvalues of G• If smallest eigenvalue of G is bigger than - mark pixel as candidate feature point
• Alternatively feature quality function (Harris Corner Detector)
Point Feature Extraction
![Page 27: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/27.jpg)
CS8690 Computer Vision University of Missouri at Columbia
Harris Corner Detector - Example
![Page 28: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/28.jpg)
CS8690 Computer Vision University of Missouri at Columbia
![Page 29: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/29.jpg)
CS8690 Computer Vision University of Missouri at Columbia
![Page 30: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/30.jpg)
CS8690 Computer Vision University of Missouri at Columbia
![Page 31: Edges. CS8690 Computer Vision University of Missouri at Columbia Gradients and edges Points of sharp change in an image are interesting: –change in reflectance.](https://reader035.fdocuments.net/reader035/viewer/2022062322/5697bfa41a28abf838c9739c/html5/thumbnails/31.jpg)
CS8690 Computer Vision University of Missouri at Columbia