Starting from:
$35

$29

Homework 08 Solution







that satisfies the Optical Flow
1.  Smallest Optical Flow (4 pts): What velocity

Constraint Equation
+






+  = 0 has the smallest magnitude |   |? Hint: This can be






solved geometrically as was outlined in class by considering the OFCE in   ,    space.


v
(0,−
(−    ,0)
u

  
2.    Moving Gaussian Blob (6 (pts):  ,  , A) Gaussian=1 blob is observed over2 time to have brightness
−2  2(  2−2( 1+ 2)  +( 1+ 2) )
    a. What are   ,   , and   ? Hint: You should find that these derivatives have a simple form.


    b. The Optical Flow Constraint Equation is + + = 0. Write this out using the results of Part a. and simplify it as much as possible. For example, you should be able to cancel terms that occur in each of , , and .
(  ,   , ) =
+ 1
[(  −
)2
+(− )2
]
3.  Quadratic Optical Flow (8 pts : Suppose the image brightness is given by
0
2
1

2

    a. What are Ix, Iy, and It? Hint: You should find that these derivatives have a simple form.
    b. Express the Optical Flow Constraint Equation + + = 0 in the simplest terms possible for this image sequence.

    c. The equation from b. must hold for all x, y, and t. Find a constant solution for u and v that makes this true, that is, such that u and v do not depend on x, y, and t.
flow, where at each iteration, the optical flow  (  , ),  (  ,   ) is updated according to

4.  Iterative Optical Flow(8 pts): We saw in class an iterative method for computing optical

  (  ,  )
new



2
+ 4




−1









old(  ) −




= [









  ∈neighbors(  ,  )

























old





[  (  ,   )]




2+4]










(  ) −


























































  ∈neighbors(  ,  )








a.  Show that this is equivalent to


2
+ 4
−    









old(  ) −

  (  ,   )
new





1












  ∈neighbors(  ,  )


































[  (  ,   )]

=
4    
2 + 4    2
+ 16
[−

2
+ 4]








old
(  ) −


































































  ∈neighbors(  ,  )












(  , )































=   ̅










(  
  ̅

+
  ̅
+  )




b.  Show that this is equivalent to update equations






















new





old







4



old




old





















2
+ 2
+

































































































(  , ) =   ̅







(  
  ̅

+
  ̅
+  )





































new





old







4



old




old





















2
+ 2
+

























































































where   ̅
old

old




































,   ̅
are the averages of the 4 neighbors of  (  , ),  (  ,   ). Hint: You only

need to show this for  new because
new follows an identical derivation.




c.  In the case that
= 0, what do the update equations reduce to?









More products