Title: Invariants to translation and scaling
1Invariants to translation and scaling
Normalized central moments
2Invariants to rotation
M.K. Hu, 1962 - 7 invariants of 3rd order
3(No Transcript)
4Hard to find, easy to prove
5Drawbacks of the Hus invariants
Dependence
Incompleteness
Insufficient number ? low discriminability
6Consequence of the incompleteness of the Hus set
The images not distinguishable by the Hus set
7Normalized position to rotation
8Normalized position to rotation
9(No Transcript)
10(No Transcript)
11Invariants to rotation
M.K. Hu, 1962
12(No Transcript)
13General construction of rotation invariants
Complex moment
Complex moment in polar coordinates
14Basic relations between moments
15(No Transcript)
16Rotation property of complex moments
The magnitude is preserved, the phase is shifted
by (p-q)a. Invariants are constructed by phase
cancellation
17Rotation invariants from complex moments
Examples
How to select a complete and independent subset
(basis) of the rotation invariants?
18Construction of the basis
This is the basis of invariants up to the order r
19Inverse problem
Is it possible to resolve this system ?
20Inverse problem - solution
21The basis of the 3rd order
This is basis B3 (contains six real elements)
22 Comparing B3 to the Hus set
23Drawbacks of the Hus invariants
Dependence
Incompleteness
24 Comparing B3 to the Hus set - Experiment
The images distinguishable by B3 but not by Hus
set
25 Difficulties with symmetric objects
Many moments and many invariants are zero
26Examples of N-fold RS
N 1 N 2 N 3
N 4 N 8
27 Difficulties with symmetric objects
Many moments and many invariants are zero
28 Difficulties with symmetric objects
The greater N, the less nontrivial invariants
Particularly
29 Difficulties with symmetric objects
It is very important to use only non-trivial
invariants
The choice of appropriate invariants (basis of
invariants) depends on N
30 The basis for N-fold symmetric objects
Generalization of the previous theorem
31(No Transcript)
32 Recognition of symmetric objects Experiment 1
5 objects with N 3
33 Recognition of symmetric objects Experiment 1
Bad choice p0 2, q0 1
34 Recognition of symmetric objects Experiment 1
Optimal choice p0 3, q0 0
35 Recognition of symmetric objects Experiment 2
2 objects with N 1 2 objects with N 2 2
objects with N 3 1 object with N 4 2
objects with N 8
36 Recognition of symmetric objects Experiment 2
Bad choice p0 2, q0 1
37 Recognition of symmetric objects Experiment 2
Better choice p0 4, q0 0
38 Recognition of symmetric objects Experiment 2
Theoretically optimal choice p0 12, q0 0
Logarithmic scale
39 Recognition of symmetric objects Experiment 2
The best choice mixed orders
40 Recognition of circular landmarks
Measurement of scoliosis progress during
pregnancy
41The goal to detect the landmark centers The
method template matching by invariants
42Normalized position to rotation
43Rotation invariants via normalization