OriginsandInterpretationsoftheConceptof

Convolutions
Presented
By

Craig Rykal
John Rivera
Justin Malaise
Jeff Swanson
Ben Rougier

From

University of Wisconsin

STOUT

Under the direction of Dr. Steve Deckelman

Abstract
The convolution of two functions is an important concept in a number of areas of
pure and applied mathematics such as Fourier Analysis, Differential Equations,
Approximation Theory, and Image Processing. Nevertheless convolutions often
seem unintuitive and difficult to grasp for beginners. This project explores the
origins of the convolution concept as well as some computer graphic and physical
interpretations of convolution which illustrate various ways the technique of
smoothing can be used to solve some real world problems.

Convolutions

F (t ) f ( x)k (t x)dx
R

Convolutions can be thought of as a method of

averaging unruly functions.
Unruly Functions include:
Functions with sharp or jagged edges
Discontinuous functions

Origin of Convolutions
n

Weighted Averages

j 1

Substitute

yj
n

y f (x) into the function to receive

j 1

Now we need

j 1

f (x j )

f ( x j ) to return a function

Therefore we create a new function

j 1

Which returns a function

f (x x j )

One more substitution of

Gives us the function:

j g j (x)

g
j 1

( x) f ( x x j )

This function will give us a new discrete function, but we

need this function in continuous form
To accomplish this, change the function to an integration
This gives us:

g
(
t
)
f
(
x

t
)
dt

Which is a Convolution!

Examples
Take the function

0 for x -h

f ( x) 1 for h x h
0 for h x

Convolve the function with itself to receive the new function

F (t )

for
t 2h
0
t 2h for 2h t 0

2h t for
0
for

0 t 2h
2h t

One more convolution gives use the continuous function

t 3h

2
(
t

3
h
)
/2

G (t ) t 2 3h 2
(t 3h) 2 / 2

3h t h
h t h
h t 3h
3h t

Convolutions are also useful in smoothing functions of more

than one dimension

F ( s, t ) f ( x, y )k ( s x, t y )dxdy
RR

Antialiasing

Aliasing

Blurred

Deconvolution

Radius of the Earth

A
Sun

Eratosthenes
Earth

7.2

Length ( A) 787 KM
c Circumference

230 B.C
Alexandria, Egypt

787 7.2

c
360

767 * 360
c
7.2

c
r
2

r 6103.59

Figuring out the Distance of The Earth to the Sun

Using Parallax
Parallax - any alteration in the relative apparent positions of objects

SUN

produced by a shift in the position of the observer.

R = radius of Earth.
= angle between observation points with relation to the center of the
Earth.
P & Q are observation points on the Earth.
X = distance between the center of the sun and the center of the Earth.

How To find the Distance:

If we know |PQ| =>

/X = Cos ()

=>

P
Q

|PQ| = R= |PQ| / R

angle of center of sun to P & Q

X = R/Cos () = R(Sec ()

The only problem is that there must be a very accurate measurement of in

order to get an accurate distance. Hipparchus (130 BC) and Ptolemy (150 AD)
used the value of the diameter of the earth given by Eratosthenes ( 195 BC) and
estimated the distance to be 10 million miles. We know that the distance is
93 million miles away. Thus, even close but not precise values give huge
errors.

Earth

The Diameter of a Star

WE SEE:

TRUE IMAGE(shape):
(x inside disc)
f ( x)
otherwise.
0

x0

x x0
f ( x) D(
)

Note: x are all unknown quantities.

Where D(x) is the characteristic function of the unit circle.

D( x) 1

for x 1,

D( x) 0

for x 1

Fundamental Problem:
Calculate from f

The Photographic Plate

X
O

X Y

|| Y ||

Y
Let

: Brightness at point source O

': Brightness at point source Y

K t ( X ) : Brightness at X
' K t ( X Y ) : Brightness at X arising from point source Y.
K t ( X ) : Apparent Brightness (At time t)

We Get:
n

A(x) " Brightness at X" j K t ( X Y j )

2

,..., n
2

Y ,Y

j1

Brightness of Yj given point sources

,...,Y n

Point sources of light

||Y || ,||Y || ,...,||Y ||n

1

Given are very small

Lets pass from discrete to the continuous model:

n

K
j 1

f (Y ) :
dA(Y ) :

( X Y j)

to

f (Y ) K ( X Y )dA(Y )

Brightness at point Y.
Kt (X
Integrate with respect to Y.

TRUE IMAGE(brightness) = f(x)

ACTUAL IMAGE = f*Kt

Y ) : Blurring

effect(kernel) of
the atmosphere at
time t.

earth

1 a.u. = distance from the sun to the earth

is found experimentally

sun

star

1 a.u.

earth

d = sec()

Knowing d and will allow us to find the

diameter of the star.

The Fundamental Problem:

Extract f from f*Kt where Kt
is an unknown random function.
Labeyries Idea:
Use Averaging and Fourier Transforms

1. Averaging Kt:
To get a fixed Kt or the Average Blur:
Average the image received at various times,

~
K 1K
n

j 1

t( j)

t (1), t (2),..., t (n)

: AVERAGE BLUR

Average the Convolutions,

n
~
1
1 n

f
*

f * K t( j)
K t( j) f * K

n
j

1
n j 1

2. Using the Convolution Theorem for Fourier Transforms:

f * K t : FOURIER TRANSFORM of our image(convolution)

t

: Under the Convolution Theorem for Fourier Transforms

Take the sequence:

, ,
t (1)

t ( 2)

t ( 3)

,...,

t (n)

t( j)

is RANDOM, due to Kt being RANDOM.

The only zeros(roots) the

t( j)

will have in common with all other

t( j)

Are the zeros of

Superimposing

t( j)

forms

( )

Clearly the zeros of

j 1

t( j)

( ) f ( )

f ( )

Will stand out as the zeros of

( )

j 1

t( j)

x x0 2 i x
f ( ) D(
)e
dA( x)

Substitute y = x x0

y 2 i(y x )
f ( ) D( ) e
dA( y )

Move out the constants

f ( )

2 i x0

y 2 i y
2 D( ) e dA( y)

Substitute w = y/

y = w

f ( )

2 i x0

D(w) e

2 i w

dA( w)

By the definition of Fourier Transforms

2 i w

D(w) e
2

dA( w) D( )

f ( )

We see that the zeros of

2 i x0

D( )

f ( ) Are the same as the zeros of D( )

TO COMPUTE , we need to know the location of the rings of zeros of

D( )

The zeros occur on the circle of radius = r0,

i.e.

D ( ) 0

When = r0

Which should be visually evident from our picture.

D ( )

2 J 1 ( )

Where J1 is the first order Bessel function

3
5
7
9
x
x
x
x
x
J 1 ( x) 2 231!2! 25 2!3! 27 3!4! 29 4!5!

The zeros Occur on the circle of radius = r1

i.e.

D ( ) 0

When = r1

NOTE: r1 can be found analytically.

So If

D ( r 0 ) 0

then

D(r1) 0

and

r r
0

And thus

1
0

Accuracy:
Labeyries method gives results consistent to
Michelsons interferometer results.
The method has been applied to over 30 stars already.

References
Fourier Analysis by T.W. Korner,
Cambridge University Press, 1988
Convolutions and Computer Graphics
by Anne M. Burns, College Mathematics
Journal, 1992
Dr. Steve Deckelman