Hough

Question 1

Line fitting challenges

Answer 1

Extra feature points (clutter)

Only some parts of a line are detected

Noise in measured feature points

Question 2

Voting is

Answer 2

A technique where we let each feature vote for all the models that are compatible with it

Question 3

Hough transform

Answer 3

Global shape parametrized by line or circle,

Edge element votes for particular parameter values

Global shape detected by many votes for same parameter value

Question 4

Hough transform outline

Answer 4

Separate images into bins
For each feature point in image, vote in every bin that could have generated this point
Find the bins with the most votes

Question 5

A line in an image corresponds to a

Answer 5

Point in hough space

Question 6

A line is a set of pints such that

Answer 6

X cos theta + y sin theta = distance

Theta is angle of perpendicular line to the x axis

Question 7

Hough transform algorithm

Answer 7

H(theta,d) =0

For each edge point (x,y)
For min theta to max theta
D = x cos theta + y sin theta
H(theta, d) += 1

Find the max if H(theta, d)

Question 8

Dealing with noise in hough transform

Answer 8

Take points only with significant gradient magnitude

Choose grid bin size carefully

Question 9

Hough transform for circles

Answer 9

Alpha = a,b,r

For each edge point [x, y, mag, dir]woth mag > t

For each possible radius value r 
Theta =gradient orientation at x,y
A = x-r cos theta
B= y-r sin theta 
H(a,b,r) += 1

Question 10

Hough pros

Answer 10

Can cope with occulusion

Can detect multiple instances of a model

Robustness to noise

Question 11

Hough cons

Answer 11

Complexity of search time increases exponentially with number of model parameters

Non target shapes can produce spurious peaks in parameter space

Hard to pick good grid size