Transformation Fourier
Transformation Fourier
Transformation Fourier
Periodic Function: Periodic function is a function that repeats itself at regular intervals. A
function y= f(x) is said to be a periodic function if there exists a positive real number P such that
f(x + P) = f(x), for all x belongs to real numbers. The least value of the positive real number P is
called the fundamental period of a function. This fundamental period of a function is also called
the period of the function, at which the function repeats itself.
The sine function is a periodic function with a period of 2π. Sin(2π + x) = Sinx. The following are
the graphs of some of the periodic functions.
Orthogonal functions: Two non-zero functions, f(x) and g(x) are said to
be orthogonal on 𝑎 ≤ 𝑥 ≤ 𝑏 if,
𝑏
∫ 𝑓(𝑥)𝑔(𝑥)𝑑𝑥 = 0
𝑎
Example:
𝜋
1. ∫−𝜋 sin(𝑚𝑥) cos(𝑛𝑥) 𝑑𝑥 = 0
𝜋
2. ∫−𝜋 sin(𝑚𝑥) 𝑑𝑥 = 0
𝜋
3. ∫−𝜋 cos(𝑛𝑥) 𝑑𝑥 = 0
Where:
f(x) is the function, we want (such as a square wave)
L is half of the period of the function
𝑎0 , 𝑎𝑛 and 𝑏𝑛 are coefficients that we need to calculate.
1 𝐿
𝑎0 = ∫−𝐿 𝑓(𝑥)𝑑𝑥
𝐿
1 𝐿 𝑛𝜋𝒙
𝑎𝑛 = ∫−𝐿 𝑓(𝑥)𝑐𝑜𝑠( )𝑑𝑥
𝐿 𝐿
1 𝐿 𝑛𝜋𝑥
𝑏𝑛 = ∫−𝐿 𝑓(𝑥)𝑠𝑖𝑛( )𝑑𝑥 𝑛 = 1,2,3, …
𝐿 𝐿
Fourier series is a periodic series which mostly used for denotes any kind of wave forms. For
example, if I want to record your voice for one second saying something, it will be found. It is a
Fourier series which may look something like this.
Wi-Fi modulation uses orthogonal combinations of amplitude and phase to encode every bit in a
stream, and even makes room for redundant data to counter the loss due to fading, reflections,
multi-path, etc. All this is synthesized in a DSP engine using FFT and IFFT.
Without Fourier, we could only encode data like old FM radios, and we’d have to wait 1 minute
to load this page, or leave the laptop all night downloading a single you tube video.
Telephones use a ridiculous number of DSP tricks to keep an acceptable voice quality, including
deconvolution, noise reduction, compression, echo cancellation, equalization, etc. All this done by
codec DSPS or FPGA that calculate Fourier millions of Fourier coefficients per second.
If Fourier wasn’t used, phone calls would sound like pre-WW2 radios or even worse: people would
have to talk louder the farther away they called from, calls would be interrupted by noise every
time people walked behind a tree or around cars because of reflections.
All of Internet uses Spatial-domain Fourier transforms to handle pictures and videos.
All Photoshop tricks and video morphing are based on Fourier transforms. Without Fourier,
Pictures would be still edited in photographic labs, with scissors and ink pen, and videos would
look like those funny psychedelic effects from the ‘70s.