Deriving an image representation using frequency components of a frequency representation
A method of deriving a representation of an image by processing signals corresponding to the image, the method comprises deriving a function of the image, where a function of a translated, scaled or rotated version of the image is a translated or scaled version of the derived function of the image, and using a plurality of frequency components of a frequency representation of the function to derive a representation of the image.
1. A method of deriving a representation of an image by processing signals corresponding to the image, the method comprising:
deriving, using a processor, a 1-dimensional function of the image, wherein said function is a circus function or a function derived from a circus function, where a function of a translated, scaled, or rotated version of the image is a translated or scaled version of the derived function of the image,
deriving, using said processor, a plurality of frequency coefficients of a frequency representation of the derived function, and
deriving, using said processor, a representation function of the image using the frequency coefficients, wherein the representation function is obtained using the steps of
calculating, using said processor, the magnitude of the frequency coefficients, and
determining, using said processor, a difference between the magnitude of each frequency coefficient and its subsequent frequency coefficient using the equation c(ω)=|F(ω)|−|F(ω+1)|, and comparing c(ω) with a threshold to derive b ω , where |F(ω)|,|F(ω+1)| are the magnitudes of the frequency coefficients and b ω are the values in a binary string.
2. The method of claim 1 wherein the frequency components are determined using a Fourier transform or a Haar transform.
3. A method of deriving a representation of an image by processing signals corresponding to the image, the method comprising:
deriving, using a processor, a 1-dimensional function of the image, wherein said function is a circus function or a function derived from a circus function, where a function of a translated, scaled, or rotated version of the image is a translated or scaled version of the derived function of the image,
deriving, using said processor, a plurality of frequency coefficients of a frequency representation of the derived function, and
deriving, using said processor, a representation function of the image using the frequency coefficients, wherein the representation function is obtained using the steps of:
calculating, using said processor, a logarithm of the magnitude of the frequency coefficients, and
determining, using said processor, a difference between the logarithm of the magnitude of each frequency coefficient and its subsequent frequency coefficient using the equation c(ω)=log|F(ω)|−log|F(ω+1)|, and comparing c(ω) with a threshold to derive b ω , where |F(ω)|,|F(ω+1)| are the magnitudes of the frequency coefficients and b ω are the values in a binary string.
4. The method of claim 1 , further comprising:
deriving the binary string b ω by applying a threshold to each determined difference to derive a series of binary values, wherein applying the threshold provides a binary value of 0 if a said determined difference is less than zero, and a binary value of 1 if a said difference is greater than or equal to zero.
5. The method of claim 1 , further comprising:
deriving two or more representation functions using two or more respective circus functions, and combining said two or more representation functions.
6. The method of claim 5 wherein combining said two or more representation functions comprises concatenating said two or more representation functions.
7. The method of claim 1 further comprising deriving a representation having n digits or bits, further comprising selecting a subset of m digits or bits.
8. The method of claim 1 wherein the derived function is derived using one or more of the following functions:
∫
ξ
(
t
)
ⅆ
t
,
(
∫
ξ
(
t
)
q
ⅆ
t
)
r
,
where
q
>
0
∫
ξ
(
t
)
′
ⅆ
t
,
∫
(
t
-
X
1
)
2
ξ
(
t
)
ⅆ
t
,
where
X
1
=
∫
t
ξ
(
t
)
ⅆ
t
∫
ξ
(
t
)
ⅆ
t
∫
(
t
-
X
1
)
2
ξ
(
t
)
ⅆ
t
∫
ξ
(
t
)
ⅆ
t
,
max
(
ξ
(
t
)
)
,
and
max
(
ξ
(
t
)
)
-
min
(
ξ
(
t
)
)
.
9. The method of claim 1 further comprising normalising the derived function before determining frequency components.
10. The method of claim 9 further comprising deriving an associated circus function or a normalised circus function.
11. The method of claim 10 using one or more of the following functions:
T
(
g
(
t
)
)
=
∫
R
+
rg
(
r
)
ⅆ
r
,
T
(
g
(
t
)
)
=
∫
R
+
r
2
g
(
r
)
ⅆ
r
,
and
P
(
h
(
t
)
)
=
∑
k
h
(
t
k
+
1
)
-
h
(
t
k
)
,
where T(g(t)) is a trace functional, P(h(t)) is a diametrical functional, r≡t−c,c≡median({t k } k ,{|g(t k )|} k ).
12. A method of identifying an image comprising:
deriving, using a processor, a representation of an image using the method of claim 1 ; and
associating, using said processor, the representation with the image.
13. A method of comparing images comprising:
comparing, using a processor, representations of each image derived using the method of claim 1 , wherein the comparison comprises:
determining a Hamming distance using the equation
H
(
B
1
,
B
2
)
=
1
N
∑
N
B
1
⊗
B
2
,
where is the exclusive OR (XOR) operator, N is the length, and B1 and B2 are two different representations both of length N.
14. The method of claim 13 further comprising selecting images based on the comparison of representations.
15. An apparatus for deriving a representation of an image by processing signals corresponding to the image comprising:
a memory; and
a processor configured to execute the method of claim 1 .
16. The apparatus of claim 15 further comprising a control device for controlling operation of the said processor configured to carry out the method.
17. The apparatus of claim 16 further comprising:
one or more databases for storing images or representations of images;
a display, and
an image selector.
18. A non-transitory computer readable medium having stored thereon instructions for executing the method of claim 1 .