Computer Vision: Mubarak Shah Computer Vision Lab University of Central Florida Orlando, FL 32816

Computer Vision
Mubarak Shah
Computer Vision Lab
University of Central Florida
Orlando, FL 32816
Computer Vision
The ability of computers to see.
Image Understanding
Machine Vision
Robot Vision
Image Analysis
Video Understanding
A picture is worth a thousand
words.
A word is worth a thousand
pictures.
A HUNT
Image
2-D array of numbers (intensity values, gray

levels)
Gray levels 0 (black) to 255 (white)
Color image is 3 2-D arrays of numbers
Red
Green
Blue
Resolution (number of rows and columns)
128X128
256X256
512X512
640X480
Image Formats
TIF
PGM
PBM
GIF
JPEG
Video
Sequence of frames
30 frames per second
Formats
AVI
MPEG
Quick Time
Video Clip
Sequence of Images
Digitization
TV camera is analog, need
A to D converter
Frame grabber
Digital Cameras do not need digitization
JVC (MPEG through fire wire, USB)
Sony (MPEG through fire wire, USB)
---
Face Recognition
Simple Approach
Recognize faces (mug shots) using gray levels

(appearance)
Each image is mapped to a long vector of gray

levels
Several views of each person are collected in the

model-base during training
During recognition a vector corresponding to an

unknown face is compared with all vectors in the
model-base
The face from model-base, which is closest to the

unknown face is declared as a recognized face.
Problems and Solution
Problems :
Dimensionality of each face vector will be very large

(250,000 for a 512X512 image!)
Raw gray levels are sensitive to noise, and lighting

conditions.
Solution:
Reduce dimensionality of face space by finding

principal components (eigen vectors) to span the face
space
Only a few most significant eigen vectors can be used

to represent a face, thus reducing the dimensionality
Eigen Vectors and Eigen Values
The eigen vector, x, of a matrix A is a special vector, with
the following property
Where is called eigen value
To find eigen values of a matrix A first find the roots of:
Then solve the following linear system for each eigen
value to find corresponding eigen vector
x Ax
0 ) det( I A
0 ) ( x I A
Example
Eigen Values
Eigen Vectors
1
1
1
]
1
7 0 0
4 3 0
0 2 1
A
1 , 3 , 7
3 2 1

1
1
1
]
1
1
1
1
]
1
1
1
1
]
1
0
0
1
,
0
2
1
,
4
4
1
3 2 1
x x x
Eigen Values
0 ) det( I A
0 )
1 0 0
0 1 0
0 0 1
7 0 0
4 3 0
0 2 1
det(
1
1
1
]
1
1
1
1
]
1
0 )
7 0 0
4 3 0
0 2 1
det(
1
1
1
]
1
7 3, , 1
0 ) 7 )( 3 )( 1 (
0 ) 0 ) 7 )( 3 )(( 1 (

Eigen Vectors
1
0 ) ( x I A
1
1
1
]
1
1
1
1
]
1
1
1
1
]
1
+
1
1
1
]
1
0
0
0
)
1 0 0
0 1 0
0 0 1
7 0 0
4 3 0
0 2 1
(
3
2
1
x
x
x
1
1
1
]
1
1
1
1
]
1
1
1
1
]
1
0
0
0
8 0 0
4 4 0
0 2 0
3
2
1
x
x
x
0 , 0 , 1
0 8 0 0
0 4 4 0
0 0 2 0
3 2 1
3
3 2
2

+ +
+ +
+ +
x x x
x
x x
x
1
1
1
]
1
0
0
1
1
x
Face Recognition
Collect all gray levels in a long vector u:
Collect n samples (views) of each of p persons in matrix A
(MN X pn):
Form a correlation matrix L (MN X MN):
Compute eigen vectors, , of L, which form
a bases for whole face space
u I I N I I N I M I M N
T
( ( , ), , ( , ), ( , ), , ( , ), , ( , ), , ( , )) 11 1 2 1 2 1
[ ]
A u u u u u u
n n
p
n
p
1
1 1
1
2 2
1
, , , , , ,
L AA
T
1 3 2 1
, , , n
Face Recognition
Each face, u, can now be represented as a linear combination
of eigen vectors
Eigen vectors for a symmetric matrix are orthonormal:
i
n
i
i
a u
1
1
'
j i
j i
j
T
i
if 0
if 1
.
Face Recognition
Therefore:
i
T
n
T
i
T T
i
T
i
n
i
i i
T
x
a u

). a a a (a
. ) ( .
n i 2 2 1 1
1
+ + + + +

) . a . a . a . (a .
n i 2 2 1 1 i
T
n i
T
i i
T
i
T
i
T
x
u + + + + +
i i
T
x
a u .
i
T
x i
u a .
Face Recognition
L is a large matrix, computing eigen vectors of a large matrix is
time consuming. Therefore compute eigen vectors of a smaller
matrix, C:
Let be eigen vectors of C, then are the eigen vectors of L:
C A A
T
i i i
C
i i i
T
A A
) ( ) (
i i i
T
A A AA
) ( ) (
i i i
A A L
i
i
A
Training
Create A matrix from training images
Compute C matrix from A.
Compute eigenvectors of C.
Compute eigenvectors of L from eigenvectors of

C.
Select few most significant eigenvectors of L for

face recognition.
Compute coefficient vectors corresponding to

each training image.
For each person, coefficients will form a cluster,

compute the mean of cluster.
Recognition
Create a vector u for the image to be

recognized.
Compute coefficient vector for this u.
Decide which person this image belongs to,

based on the distance from the cluster
mean for each person.

Computer Vision: Mubarak Shah Computer Vision Lab University of Central Florida Orlando, FL 32816

Uploaded by

Copyright:

Available Formats

Computer Vision: Mubarak Shah Computer Vision Lab University of Central Florida Orlando, FL 32816

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Computer Vision: Mubarak Shah Computer Vision Lab University of Central Florida Orlando, FL 32816

Uploaded by

Copyright:

Available Formats

Computer Vision

The ability of computers to see.

2-D array of numbers (intensity values, gray

Gray levels 0 (black) to 255 (white)

Color image is 3 2-D arrays of numbers

Resolution (number of rows and columns)

30 frames per second

TV camera is analog, need

Digital Cameras do not need digitization

JVC (MPEG through fire wire, USB)

Sony (MPEG through fire wire, USB)

Recognize faces (mug shots) using gray levels

Each image is mapped to a long vector of gray

Several views of each person are collected in the

During recognition a vector corresponding to an

The face from model-base, which is closest to the

Dimensionality of each face vector will be very large

Raw gray levels are sensitive to noise, and lighting

Reduce dimensionality of face space by finding

Only a few most significant eigen vectors can be used

Create A matrix from training images

Compute C matrix from A.

Compute eigenvectors of L from eigenvectors of

Select few most significant eigenvectors of L for

Compute coefficient vectors corresponding to

For each person, coefficients will form a cluster,

Create a vector u for the image to be

Compute coefficient vector for this u.

Decide which person this image belongs to,

You might also like