The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Magnitude and Phase The Fourier Transform: Examples, Properties, Common Pairs Remember: complex numbers can be thought of as (real,imaginary) or (magnitude,phase). Magnitude: Phase: CS 450: Introduction to Digital Signal and Image Processing Bryan Morse BYU Computer Science Real part Imaginary part Magnitude Phase |F | φ (F ) = = <(F )2 + =(F )2 ) tan−1 =(F <(F ) How much of a cosine of that frequency you need How much of a sine of that frequency you need Amplitude of combined cosine and sine Relative proportions of sine and cosine The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Example: Fourier Transform of a Cosine Example: Fourier Transform of a Cosine Z F (u) = ∞ −∞ ∞ f (t) = cos(2πst) f (t) e −i2πut Z = Z = = = cos(2πst) e −i2πut Frequency Domain 1 2 δ(u cos(2πst) dt −∞ ∞ −∞ −∞ 0 except when u = ±s 1 1 δ(u − s) + δ(u + s) 2 2 0 for all u 0.8 0.5 0.6 0.2 0.4 0.6 1 0.8 0.2 -1 -10 The Fourier Transform: Examples, Properties, Common Pairs Odd and Even Functions Sinusoids ∗ for real-valued signals Odd f (−t) = −f (t) Anti-symmetric Sines Transform is imaginary∗ 0.4 -0.5 The Fourier Transform: Examples, Properties, Common Pairs Even f (−t) = f (t) Symmetric Cosines Transform is real∗ − s) + 12 δ(u + s) 1 1 cos(2πst) [cos(−2πut) + i sin(−2πut)] dt Z ∞ cos(2πst) cos(−2πut) dt + i cos(2πst) sin(−2πut) dt −∞ −∞ Z ∞ Z ∞ cos(2πst) cos(2πut) dt − i cos(2πst) sin(2πut) dt Z = −∞ ∞ Spatial Domain dt 1/2 Spatial Domain f (t) -5 Frequency Domain F (u) cos(2πst) 1 2 [δ(u + s) + δ(u − s)] sin(2πst) 1 2i [δ(u + s) − δ(u − s)] 5 10 The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Constant Functions Delta Functions Spatial Domain f (t) Frequency Domain F (u) 1 δ(u) a a δ(u) Spatial Domain f (t) Frequency Domain F (u) δ(t) 1 The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Square Pulse Square Pulse Spatial Domain f (t) 1 if −a/2 ≤ t ≤ a/2 0 otherwise Frequency Domain F (u) sinc(aπu) = sin(aπu) aπu The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Triangle Comb Spatial Domain f (t) 1 − |t| if −a ≤ t ≤ a 0 otherwise Frequency Domain F (u) Spatial Domain f (t) Frequency Domain F (u) sinc2 (aπu) δ(t mod k) δ(u mod 1/k ) The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Gaussian Differentiation Spatial Domain f (t) e−πt Frequency Domain F (u) 2 e−πu 2 Spatial Domain f (t) Frequency Domain F (u) d dt 2πiu The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Some Common Fourier Transform Pairs More Common Fourier Transform Pairs Spatial Domain f (t) Cosine cos(2πst) Sine sin(2πst) Unit 1 Constant a Delta δ(t) Comb δ(t mod k) Frequency Domain F (u) Deltas 12 [δ(u + s) + δ(u − s)] Deltas 12 i [δ(u + s) − δ(u − s)] Delta δ(u) Delta aδ(u) Unit 1 Comb δ(u mod 1/k ) Spatial Domain f (t) 1 if −a/2 ≤ t ≤ a/2 Square 0 otherwise 1 − |t| if −a ≤ t ≤ a Triangle 0 otherwise 2 Gaussian e−πt d Differentiation dt The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Properties: Notation Properties: Linearity Let F denote the Fourier Transform: F = F(f ) Let F −1 denote the Inverse Fourier Transform: f = F −1 (F ) Frequency Domain F (u) Sinc sinc(aπu) Sinc2 sinc2 (aπu) Gaussian Ramp 2 e−πu 2πiu Adding two functions together adds their Fourier Transforms together: F(f + g) = F(f ) + F(g) Multiplying a function by a scalar constant multiplies its Fourier Transform by the same constant: F(af ) = a F(f ) The Fourier Transform: Examples, Properties, Common Pairs The Fourier Transform: Examples, Properties, Common Pairs Properties: Translation Change of Scale: Square Pulse Revisited Translating a function leaves the magnitude unchanged and adds a constant to the phase. If f2 = f1 (t − a) F1 = F(f1 ) F2 = F(f2 ) then |F2 | = |F1 | φ (F2 ) = φ (F1 ) − 2πua Intuition: magnitude tells you “how much”, phase tells you “where”. The Fourier Transform: Examples, Properties, Common Pairs Rayleigh’s Theorem Total “energy” (sum of squares) is the same in either domain: Z ∞ Z ∞ 2 2 |f (t)| dt = |F (u)| du −∞ −∞