Professional Documents
Culture Documents
Computer Vision #5
Differential Geometry
1. Curvature of curve
2. Curvature of surface
3. Application of curvature
Parameterization of curve
1. curve -- s arc length s
a(s) = ( x(s), y(s) )
a(s)
2. tangent of a curve
a’(s) = ( x’(s), y’(s) ) a ' ( s ) s
a ( s s ) a (s )
3. curvature of a curve
a”(s) = ( x”(s), y”(s) )
|a”(s)| -- curvature a ' ( s s ) a' (s)
a" ( s ) s
a' ( s)
a ' ( s s )
Example (circle)
y
1. Arc length, s
s r s r
2. coordinates x
x r cos r cos s r y r sin r sin s r
a s xs , y s r coss , r sin s
r r
3. tangent
a ' s sin s , cos s
r r a' s
a s
4. curvature
S
a" s k s ns
Corner model and its signatures
s=0
a
b
b c d a
a arc length
d
c
b a a b c
s=0
c d
s=0
Gaussian filter and scale space
+ a
a b e
+ c d
b f h k
+ d
a c e g i j
Curvature of surfaces
K1 K2 0
cylinder:
K1 0 K2 0
ellipsoid:
K1 0 K2 0
hyperboloid:
K1 0 K2 0
Gaussian curvature and mean curvature
Gaussian curvature K1 K 2 K
K1 K 2
Mean curvature H
2
K1 0 K 0
K2 0 H 0
K1 0 K 0
K2 0 H 0
K1 0 K 0
K2 0 H 0
K1 0 K 0
K2 0 H 0
K1 0 K 0
K2 0 H ?
Parabolic points
K 0 Parabolic point
K 0 elliptic point
K 0 hyperbolic point
Principal direction
PD
PD
PD
Lines of curvature
Curvature primal sketches along
lines of curvature
Important formula
1. Surface X u , v x u , v , y u , v , z u , v N
Xv
x, y, z x, y
Xu
2. surface normal
Xu Xv
N u , v
Xu Xv
area EG F 2 dudv
2
eg f
Gaussian curvature 2
EG F
1 eG 2 fF gE
Mean curvature 2
2 EG F
Z
X , X , , Y , , Z ,
r sin cos , r sin sin , r cos
Y
i j k
X X r cos cos r cos sin r sin
r sin sin r sin cos 0
r 2 sin sin cos i sin sin j cos k
r sin d
area EG F dd 2
1
r
Summary
1. curvature of curve
2. curvature of surface
– Gaussian curvature
– mean curvature
Surface Description #2
(Extended Gaussian Image)
Topics
1.Gauss map
3.Application of EGI
Gauss map
gauss map
1D
gauss map 2D
S 2 x, y , z R 3 ; x 2 y 2 z 2 1, N x, y , z
The map N: S→S3 is called the Gauss map.
Characteristics of EGI
EGI is the necessary and the sufficient condition for
the congruence of two convex polyhedra.
S
object K lim Gaussian sphere
O 0 O
O large S S small
(K: small)
small
O
O small S
(K: large)
large S large
O
Gaussian curvature and EGI maps
X v dv O X u du X v dv
X u du
X u X v dudv
N v dv
v dv du
N u du
u
N N v dv
N S N u du N v dv
N N u du N u N v dudv
Since N u and N v exist on the tangential plane at X ,
we can represent them by a linear combination of X u and X v
Nu a11 Xu a21 Xv
N v a12 X u a22 X v
N u N v ( a11a22 a12 a21 ) X u X v
S N u N v a a21
det 11
O Xu Xv a12 a22
N u X u a11 X u X u a21 X v X v
e a11 E a21 F
N v X u a12 X u X v a22 X v X u
f a12 E a22 F
N u X v a11 X u X v a21 X v X v
f a11 F a21G
N v X v a12 X u X v a22 X v X v
g a12 F a22G
e f a11 a21 E F
G
f
g a12 a22 F
S a a21 eg f 2
det 11 k
O a12 a22 EG F 2
O
EGI mass
S
1
EGI map
k
Implementation of EGI
Tessellation of the unit sphere
•all cells should have the same area
have the same shape
occur in a regular pattern
•geodesic dome based on a regular polyhedron
side view
top view
Cylinder Ellipsoid
Determination of attitude using EGI
10 20 0
viewing
direction
5 8 0
EGI table
5 8 0
The complex EGI(CEGI)
An k
dk
Pn k An k e jd k
0
The complex EGI(CEGI)
n1 Gauss mapping n1
A1e jd1 n6
A1
n6
n4 n3 n4
n3
d1
n5 n5
n2
Origin n2
Hand-eye calibration
Photometric Stereo Set-up
Bin-Picking System
Summary
1. Gauss map
2. Extended Gaussian Image
3. Characteristics of EGI
congruence of two convex polyhedra
EGI mass is the inverse of Gaussian curvature
mass center of EGI is at the origin of the sphere
4. Implementation of EGI
Tessellation of the unit sphere
Recognition using EGI
5. Complex EGI
6. Bin-picking system based on EGI
7. Read Horn pp.365-39 pp.423-451