Edges

Question 1

What are edges in images

Answer 1

Areas with strong intensity contrasts

Question 2

General edge detection strategy

Answer 2

Determine image gradient

Mark points where gradient magnitude is particularly large with respect to neighbors

Question 3

2D difference operators

Answer 3

Prewitt,
Sober,
Roberts

Question 4

Prewitt difference operator

Answer 4

• 101
• 101
• 101

111
000
-1-1-1

Question 5

Sobel difference operator

Answer 5

• 101
• 202
• 101

121
000
-1-2-1

Question 6

Roberts

Answer 6

01
-10

10
0-1

Question 7

Compass gradient masks

Answer 7

Use eight masks with usual compass directions
Select largest response
Orientation is the direction associated with the largest response

Gradient magnitude- max response
Gradient direction- direction of max response

Question 8

Properties of derivative masks

Answer 8

Masks have opposite signs

Sum of masks is zero rather than 1 (like smoothing masks)

1st deriv- high absolute values at contrast

2nd deriv- zero-cropping points at contrast

Question 9

Edge detection steps

Answer 9

1.filtering
(Tradeoff between edge detection and noise reduction)

2. Enhancement
   Emphasize pixels with sig change in local intensity value
3. Detection
   Thresholding
4. Localization
   Location of edge can be estimated

Question 10

Why do we use 2 deriv operators?

Laplacian

Answer 10

Thresholding gradient images produces too many edge points

Question 11

Weakness of laplacian operator

Answer 11

Very sensitive to noise
Can produce double edges
Unable to detect edge direction
(Use zero crossing property for edge location)

Question 12

Laplacian of Gaussian

Answer 12

Aka Marr-Hildreth

Basically you smooth image with Gaussian than apply a Laplacian

Edges will be at zero crossings

Question 13

Problems with zero edge finding

Answer 13

Spaghetti effect (closed loops)

More sophisticated to pull off

(Gradient based edge detection is still used more frequently)