Digital Image Transforms
© Oxford University Press 2011
• Image Processing Operations
– Spatial Domain
• Directly operates on the pixel intensities
• Can be visualised easily
– Frequency Domain
• Deals with the frequency components of the
images
© Oxford University Press 2011
– Real Time image Processing Applications
• Biometric System – Fingerprint based attendance
• Airport baggaging screening system
• Radar Image monitoring
– Important Requirement
• Speed operations
• Simple operations
– Frequency domain serves both
– A change in the problem representation –
solves the problem faster
© Oxford University Press 2011
NEED FOR IMAGE
TRANSFORMS
© Oxford University Press 2011
Spatial domain and
Frequency domain
• These are different representations of the same
information
• Calculations in the frequency domain are simpler
SPATIAL FREQUENCY
DOMAIN DOMAIN
© Oxford University Press 2011
Example
Transforms Inverse Transforms
Fourier – DFT, 2D DFT Inverse DFT, Inverse 2D DFT
DCT –Discrete Cosine Transforms IDCT
DWT – Discrete Wavelet Transforms IDWT
© Oxford University Press 2011
Spatial Frequencies in Image
Processing
• ‘spatial frequency’ is used to describe the rate
of change of pixel intensity of an image in space.
It can be visualized using the line profile of the
image in any row or column.
• X-axis – Distance of a pixel from the origin
• Y – axis – Grey value (0-255)
• Intensity Profile – Collection of either sinusoidal
signals (sine, cosine, harmonic signals) or non-
sinusoidal signals (square waves)
© Oxford University Press 2011
• High frequency Component
– Rapidly changing data eg., edges
• Low frequency Component
– Gradual change of data
– Contrast and other large scale features of the image
• Intensity Change I sin(ux)
u = Frequency; x= Distance across the image; (-1)≤ I ≤ (1)
2
• For a gray scale image I 127.5 x sin ux 127.5
m
• 2 - To convert the frequency from cycles to radians per image width
m – image width in pixels
© Oxford University Press 2011
Image Profile
© Oxford University Press 2011
© Oxford University Press 2011
Types of image transforms
• Essentially there are two categories of
transforms available
1. Orthogonal
Fourier, DCT
2. Non-sinusoidal.
Haar, Hadamard, Walsh
© Oxford University Press 2011
Practical Applications of Image
Transforms
• Useful in understanding Sampling
• Energy Preservation – Signal energy is
preserved
• Energy Compaction
– large fraction of the average energy of the
image is packed into a few components, in
transform domain
– Uneven distribution of energy
– Many co-eff. have negligible energy; useful in
compression
© Oxford University Press 2011
• Helps to design faster algorithms
• Helps to assess image quality
– Frequency denotes repetitive noise patterns.
Presence or absence of high/low areas
– Spectrum images of a clean and a noisy
image
• Orthogonal transforms
– Co-eff. are all real nos.
• Hermitian transforms
– Co-eff. are all complex nos.
© Oxford University Press 2011
© Oxford University Press 2011
Transforms
T(u,v,x,y) – Forward transformation
I(x,y,u,v) – Inverse tranformation
Main task – Find a suitable kernel
The forward kernel is said to be separable if
T (u, v, x, y) T1 (u, x)T2 (v, y)
It is said to be symmetric if is functionally equal to such that
T (u, v, x, y) T1 (u, x)T1 (v, y)
© Oxford University Press 2011
Introduction to Fourier
Transform
© Oxford University Press 2011
• The inverse Fourier transform is
expressed mathematically as
© Oxford University Press 2011
© Oxford University Press 2011
DFT
© Oxford University Press 2011
© Oxford University Press 2011
© Oxford University Press 2011
© Oxford University Press 2011
© Oxford University Press 2011
FFT
© Oxford University Press 2011
FFT ALGORITHM
© Oxford University Press 2011
FFT
© Oxford University Press 2011
INVERSE FFT
© Oxford University Press 2011
