文章探讨了傅里叶变换在时域和频域中的对称性质，周期性连续时间函数的傅里叶变换为离散非周期频率函数，而离散时间函数的变换则对应连续周期性函数。傅里叶级数、离散傅里叶变换（DFT）、离散傅里叶逆变换（IDFT）以及二维傅里叶变换的概念和表示也被阐述，强调了时间域和频率域之间的转换规则。
傅里叶变换的周期性和离散性在时域与变换域中表现出巧妙的对称关系，即：呈周期性的连续时间函数，其傅里叶变换为离散的非周期频率函数（傅里叶级数的离散频谱）；而非周期性的离散时间函数，其傅里叶变换为连续的周期性函数（抽样信号的频谱呈周期性）。
简而言之，某个域的某属性，对应另一域的另一属性，即周期性对应离散，非周期对应连续；且离散间隔与周期长度对应 f = 1 T f=\frac{1}{T} f=T1，即时间离散间隔对应频率重复周期，频率离散间隔对应时间重复周期。
连续时间与连续频率
X
(
f
)
=
∫
−
∞
∞
x
(
t
)
e
−
j
2
π
f
t
d
t
x
(
t
)
=
∫
−
∞
∞
X
(
f
)
e
j
2
π
f
t
d
f
\begin{align} X(f) &= \int_{-\infty}^\infty x(t)e^{-j2\pi ft} \mathrm dt \\ x(t) &= \int_{-\infty}^\infty X(f)e^{j2\pi ft} \mathrm df \end{align}
X(f)x(t)=∫−∞∞x(t)e−j2πftdt=∫−∞∞X(f)ej2πftdf
连续时间与离散频率
X
(
k
f
1
)
=
1
T
1
∫
T
1
x
(
t
)
e
−
j
2
π
k
f
1
t
d
t
x
(
t
)
=
∑
k
=
−
∞
∞
X
(
k
f
1
)
e
j
2
π
k
f
1
t
\begin{align} X(kf_1) &= \frac{1}{T_1}\int_{T_1} x(t)e^{-j2\pi kf_1t}\mathrm dt \\ x(t) &= \sum_{k=-\infty}^\infty X(kf_1) e^{j2\pi kf_1t} \end{align}
X(kf1)x(t)=T11∫T1x(t)e−j2πkf1tdt=k=−∞∑∞X(kf1)ej2πkf1t
离散时间与连续频率
X
(
f
)
=
∑
n
=
−
∞
∞
x
(
n
T
s
)
e
−
j
2
π
f
n
T
s
x
(
n
T
s
)
=
1
f
s
∫
f
s
X
(
f
)
e
j
2
π
f
n
T
s
d
f
\begin{align} X(f) &= \sum_{n=-\infty}^\infty x(nT_s) e^{-j2\pi fnT_s} \\ x(nT_s) &= \frac{1}{f_s}\int_{f_s} X(f) e^{j2\pi fnT_s}\mathrm df \end{align}
X(f)x(nTs)=n=−∞∑∞x(nTs)e−j2πfnTs=fs1∫fsX(f)ej2πfnTsdf
离散时间与离散频率
X
(
k
f
1
)
=
∑
n
=
0
N
−
1
x
(
n
T
s
)
e
−
j
2
π
N
n
k
x
(
n
T
s
)
=
1
N
∑
k
=
0
N
−
1
X
(
k
f
1
)
e
j
2
π
N
n
k
\begin{align} X(kf_1) &= \sum_{n=0}^{N-1} x(nT_s)e^{-j\frac{2\pi}{N}nk} \\ x(nT_s) &= \frac{1}{N}\sum_{k=0}^{N-1} X(kf_1) e^{j\frac{2\pi}{N}nk} \end{align}
X(kf1)x(nTs)=n=0∑N−1x(nTs)e−jN2πnk=N1k=0∑N−1X(kf1)ejN2πnk
离散傅里叶级数
X
p
(
k
)
=
∑
n
=
0
N
−
1
x
p
(
n
)
e
−
j
2
π
N
n
k
x
p
(
n
)
=
1
N
∑
k
=
0
N
−
1
X
p
(
k
)
e
j
2
π
N
n
k
\begin{align} X_p(k) &= \sum_{n=0}^{N-1} x_p(n)e^{-j\frac{2\pi}{N}nk} \\ x_p(n) &= \frac{1}{N}\sum_{k=0}^{N-1} X_p(k) e^{j\frac{2\pi}{N}nk} \end{align}
Xp(k)xp(n)=n=0∑N−1xp(n)e−jN2πnk=N1k=0∑N−1Xp(k)ejN2πnk
令上式中，
W
N
=
W
=
e
−
j
2
π
N
W_N=W=e^{-j\frac{2\pi}{N}}
WN=W=e−jN2π，有如下形式。
离散傅里叶级数对
X
p
(
k
)
=
DFS
[
x
p
(
n
)
]
=
∑
n
=
0
N
−
1
x
p
(
n
)
W
n
k
x
p
(
n
)
=
IDFS
[
X
p
(
k
)
]
=
1
N
∑
k
=
0
N
−
1
X
p
(
k
)
W
−
n
k
\begin{align} X_p(k) &= \text{DFS}[x_p(n)] = \sum_{n=0}^{N-1} x_p(n)W^{nk} \\ x_p(n) &= \text{IDFS}[X_p(k)] = \frac{1}{N}\sum_{k=0}^{N-1} X_p(k)W^{-nk} \end{align}
Xp(k)xp(n)=DFS[xp(n)]=n=0∑N−1xp(n)Wnk=IDFS[Xp(k)]=N1k=0∑N−1Xp(k)W−nk
离散傅里叶变换
X
(
k
)
=
DFT
[
x
(
n
)
]
=
∑
n
=
0
N
−
1
x
(
n
)
W
n
k
,
0
≤
k
≤
N
−
1
x
(
n
)
=
IDFT
[
X
(
k
)
]
=
1
N
∑
k
=
0
N
−
1
X
(
k
)
W
−
n
k
,
0
≤
n
≤
N
−
1
\begin{align} X(k)&=\text{DFT}[x(n)]=\sum_{n=0}^{N-1} x(n)W^{nk} \ ,\quad 0\le k\le N-1 \\ x(n)&=\text{IDFT}[X(k)]=\frac{1}{N}\sum_{k=0}^{N-1} X(k)W^{-nk}\ ,\quad 0\le n\le N-1 \end{align}
X(k)x(n)=DFT[x(n)]=n=0∑N−1x(n)Wnk ,0≤k≤N−1=IDFT[X(k)]=N1k=0∑N−1X(k)W−nk ,0≤n≤N−1
二维傅里叶变换
F
(
u
,
v
)
=
F
2
D
[
f
(
x
,
y
)
]
=
∬
−
∞
∞
f
(
x
,
y
)
e
−
j
2
π
(
u
x
+
v
y
)
d
x
d
y
f
(
x
,
y
)
=
F
2
D
−
1
[
F
(
u
,
v
)
]
=
∬
−
∞
∞
F
(
u
,
v
)
e
j
2
π
(
u
x
+
v
y
)
d
u
d
v
\begin{align} F(u,v) &= \mathscr F_{2\text{D}}[f(x,y)] = \iint_{-\infty}^\infty f(x,y) e^{-j2\pi(ux+vy)} \mathrm dx\mathrm dy \\ f(x,y) &= \mathscr F^{-1}_{2\text{D}}[F(u,v)] = \iint_{-\infty}^\infty F(u,v) e^{j2\pi(ux+vy)} \mathrm du\mathrm dv \end{align}
F(u,v)f(x,y)=F2D[f(x,y)]=∬−∞∞f(x,y)e−j2π(ux+vy)dxdy=F2D−1[F(u,v)]=∬−∞∞F(u,v)ej2π(ux+vy)dudv
二维傅里叶逆变换
F
(
u
,
v
)
=
DFT
2
D
[
f
(
x
,
y
)
]
=
∑
x
=
0
M
−
1
∑
y
=
0
N
−
1
f
(
x
,
y
)
e
−
j
2
π
(
u
x
M
+
v
y
N
)
f
(
x
,
y
)
=
IDFT
2
D
[
F
(
u
,
v
)
]
=
1
M
N
∑
u
=
0
M
−
1
∑
v
=
0
N
−
1
F
(
u
,
v
)
e
j
2
π
(
u
x
M
+
v
y
N
)
\begin{align} F(u,v) &= \text{DFT}_{2\text{D}}[f(x,y)] = \sum_{x=0}^{M-1}\sum_{y=0}^{N-1} f(x,y) e^{-j2\pi(\frac{ux}{M}+\frac{vy}{N})} \\ f(x,y) &= \text{IDFT}_{2\text{D}}[F(u,v)] = \frac{1}{MN}\sum_{u=0}^{M-1}\sum_{v=0}^{N-1} F(u,v) e^{j2\pi(\frac{ux}{M}+\frac{vy}{N})} \end{align}
F(u,v)f(x,y)=DFT2D[f(x,y)]=x=0∑M−1y=0∑N−1f(x,y)e−j2π(Mux+Nvy)=IDFT2D[F(u,v)]=MN1u=0∑M−1v=0∑N−1F(u,v)ej2π(Mux+Nvy)
扫一扫
1144
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
