Spatial and Frequency-domain Filters

 Enhancement using arithmetic  Sharpening Spatial filters,

/logic operations, - Foundation,
- Image subtraction, - Step and ramp discontinuities.
- image averaging.
 Frequency Domain,
 Basics of spatial filtering, - Fourier transform.
- Linear spatial filtering,
 Smoothing Frequency-Domain
 Piecewise-linear transform Filters,
functions, - Ideal lowpass filter,
- Smoothing linear filter, - Butterworth lowpass filter,
- box filters. - Gaussian lowpass filter.
 Order-Statistics (Nonlinear)
filters,
- Median filters,
@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)
1. Enhancement Using Arithmetic/Logic Operations
(1/6)
 Arithmetic/logic operations involves:
– images are performed on a pixel-by-pixel basis between two or more images.
For Example;
- subtraction of two images results in a new image.
- whose pixel at coordinates (x, y) is the difference between the pixels in that
same location on the two images being subtracted.

 Actual mechanisms of implementation arithmetic/logic operations

can be done, sequentially one pixel at a time, or in parallel.

- To implement the AND, OR and NOT logic operators because these three
operators are functionally complete.
- Example, performing the NOT operation on a black, 8-bit pixel produces a
white pixel.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

1. Enhancement Using Arithmetic/Logic Operations
(2/6)
 Arithmetic/logic operations involves:
– In the AND or OR image masks, light represents a binary 1 and dark
represents a binary 0.
– Masking sometimes is referring to as region of interest (ROI) processing.

Figure: Arithmetic/Logic Operations such as NOT, AND and OR for images A and B

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

1. Enhancement Using Arithmetic/Logic Operations
(3/6)
(Assignment)

1. Apply AND, OR and NOT operators at different

scene datasets.

2. Make manual ROI processing at face datasets.

3. Develop ROI technique automatically over pedestrian

datasets.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

1. Enhancement Using Arithmetic/Logic Operations
(4/6)
 Arithmetic/logic operations involves:
– In the arithmetic operation, subtraction and addition are the most useful for
image enhancement.
– We consider divide of two images simply
– as multiplication of an image by the reciprocal of the other.

 Types of Arithmetic/logic operations:

1) Image Subtraction:
 The difference between two image f (x,y) and h (x,y), expressed as

g (x,y) = f (x,y) - h (x,y)

- is obtained by` computing the difference between all pixels from f and h.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

1. Enhancement Using Arithmetic/Logic Operations
(5/6)
 Types of Arithmetic/logic operations:
1) Image Subtraction:
– One of the most commercially successful and beneficial uses of image
subtraction is in the area of medical imaging called mask mode radiography.

– In this case h(x,y), the mask, is an X-ray image of a region of a patient’s body
captured by an intensified TV camera located opposite an X-ray source.
– The procedure consists of injecting a contrast medium into the patient’s
bloodstream, taking a series of images having same anatomical region,
– After subtracting this mask from the series of incoming images after injection
of the contrast medium.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

1. Enhancement Using Arithmetic/Logic Operations
(6/6)
 Types of Arithmetic/logic operations:
2) Image Averaging:
 Consider a noisy image g (x, y) formed by the addition of noise n (x, y) to an
original image f (x, y); that is

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

2. Basics of Spatial Filtering (1/3)
 Neighborhood operations work with the :
- values of the image pixels in the neighborhood and the corresponding values of a
subimage that has the same dimensions as the neighborhood.

 This subimage is called as;

- filter,
- mask, kernel,
- template or windowing,

 While, windowing is the most prevalent terminology;

- the values in a filter subimage are refereed to as coefficients, rather than pixels.

 Filtering is a technique for modifying or enhancing an image.

- For example, you can filter an image to emphasize certain features or remove
other features.
- filtering include smoothing, sharpening, and edge enhancement.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

2. Basics of Spatial Filtering (2/3)
 This process consists simply of moving the filter mask from point
to point in an image. At each point (x, y), the response of the filter:
- at that point is calculated using a predefined relationship.

1) Linear Spatial Filtering:

 response is given by a sum of products of the filter coefficients and
the corresponding image pixels in the area spanned by the filter
mask.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

2. Basics of Spatial Filtering (3/3)
1) Linear Spatial Filtering:
- w (-1,-1) f(x-1,y-1) + w (-1,0) f (x-1,y) + ………………
………………………...............+ w(1,0)f(x+1,y) +
w(1,1)f(x+1,y+1)

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (1/10)
 Smoothing filters are used for;
- blurring and for noise reduction.

 Blurring is used in preprocessing step, such as;

- removal of small details from an image to object extraction and
- bridge of small gaps in lines or curves.

 Noise reduction can be accomplished by blurring with a;

- linear filter and
- non-linear filtering.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (2/10)

Types of Smoothing Spatial Filters:

1) Linear (Average) Filters:-
Output of a smoothing, linear spatial filter is simply the average of the
pixels contained in the neighborhood of the filter mask.
- This filters sometimes are classed average filters, they also are refereed to a
lowpass filters.

However, edges also are characterized by sharp transitions on gray levels,

so averaging filters have the undesirable side effect that they blur edges.
- A major use of Averaging Filters is in the reduction of “irrelevant” details in
an image.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (3/10)
Types of Smoothing Spatial Filters:
1) Linear (Average) Filters (For Example)

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (4/10)
(Class Participation)
Types of Smoothing Spatial Filters:
 Define 3x3 averaging smoothing filter. What is the structure of
matrix if we have an;
=> average value = 32,
=> average value = 128,
=> average value = 256,
=> average value = 64,
=> average value = 512,
=> average value = 1024,

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (5/10)

Types of Smoothing Spatial Filters:

2) Box (Weighted average) Filters:-
A spatial averaging filter in which all
coefficients are equal is sometime called a box
filter.

The 2nd mask so-called weighted average,

terminology used to indicate that pixels are
multiplied by different coefficients.

In the mask, (b-f) the pixel at the center of

the mask is multiplied by a higher value than
any other, thus giving this pixel more
importance in the calculation of the average.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (6/10)
Types of Smoothing Spatial Filters:
2) Box (Weighted average) Filters:-
Basic strategy:
- weighting the center point the highest.
- Then, reducing the value of the coefficients as a function of
increasing distance from the origin is simply an attempt to reduce
blurring in the smoothing process.
The general implementation for filtering an M  N image
with a weighted averaging filter of size m  n is given
a b

  w( s, t ) f ( x  s, y  t )
g ( x, y )  s  a t  b
a b

  w( s, t )
s  a t  b

where m  2 a  1, n  2b  1.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (8/10)
Types of Smoothing Spatial Filters:
3) Order-Statistics Filters (Non-Linear):-
Order-statistics filters are non-linear spatial filters whose response
is based on ordering (ranking).
- the pixel contained in the image area encompassed by the filters.
- then, replace the value of the center pixel with the value determined by the
ranking result.

For Example :
- this category used medium filter, which, replaces the value of the
pixel by the medium of the gray levels in the neighborhood of that
pixel.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (9/10)
Types of Smoothing Spatial Filters:
3) Order-Statistics Filters (Non-Linear):-
Medium filters are:
- Particularly effective in the presence of impulse noise, also called salt-and-
pepper noise because of its appearance as white and black dots superimposed on an
image.
- When several values in a neighborhood are the same, all equal values are
grouped.

For Example; Suppose that a 3x3 neighborhood has values

(10,20,20,20,15,20,20,25,100).
These values are sorted as ( 10,15, 20,20,20,20,20,25,100), which results in the
median of 20.
Thus, the principal function of medium filter is to force points with distinct gray
levels to be more like their neighbors.
In this case “eliminated” means forced to the medium intensity of the neighbors.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

3. Smoothing Spatial Filters (10/10)
Types of Smoothing Spatial Filters:
3) Order-Statistics Filters (Non-Linear):-
Example: Use of Median Filtering for Noise Reduction

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4. Image Enhancement in the Frequency Domain
 Fourier’s contribution in this
particular field states that:
– Any function that periodically
repeats itself can be expressed as
the sum of sines and/or cosines of
difference frequencies, each
multiplied by a different coefficient.

– First time practical processing and

meaningful interpretation as :

 Medical monitors and scanners

to modern electronics
communication.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4.1 One-Dimensional Fourier Transform and its
Inverse

The Fourier Transform of a continous function f (t )


F (  )  { f (t )}   f (t )e j 2t dt


The Inverse Fourier Transform of F (  )


f (t )   {F (  )}   F (  )e j 2t d 
1


@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4.1 One-Dimensional Fourier Transform and its
Inverse

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4.2 Filtering in the Frequency Domain (1/5)

 F (u , v) contains all values of f (x,y), modified by the values of

the exponential terms.

 The slowest varying frequency component (u = v = 0)

corresponds to the average gray level of an image.

 As we move away from the origin of the transform,

- the low frequencies correspond to the slowly varying
components of an image.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4.2 Filtering in the Frequency Domain (2/5)

 Steps of filtering in the frequency domain:-

- Filtering in the frequency domain is straight forward. It consists of the
following steps.

1) Multiple the input image by (-1)x+y to center the transform.

2) Compute F (u, v), the DFT of the image.
3) Multiple F (u, v) by the filter function H (u, v).
4) Compute the inverse DFT.

H (u, v) is called filter because it suppressed certain frequencies I

the transform while leaving other unchanged.

G (u , v) = H (u, v) * F(u, v)

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4.2 Filtering in the Frequency Domain (3/5)

 The multiplication of H and F involves two–dimensional functions.

 The first element of H multiplies the first element of F, the second
element of H multiplies the second element of F.

 Each component of H multiplies both the real and imaginary parts

of the corresponding component in F. Such a filter is called zero-
phase-shift filters.
Fourier Filter Inverse
transform function Fourier
H (u,v) transform
Pre-
F (u, v) H (u,v) F (u, v) Post-
Processing
Processed

F (x, y) g (x, y)
Input image Enhanced image
@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)
4.2 Filtering in the Frequency Domain (4/5)
 Low frequencies in the Fourier transform are responsible for the
general gray-level appearance of the image over smooth areas,

 A filter that attenuates high frequencies while “passing” low

frequencies is called a lowpass filter.
- is a filter that passes signals with a frequency lower than a selected cutoff
frequency.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

4.2 Filtering in the Frequency Domain (5/5)
 While, high frequencies are responsible for detail, such as edges
and noise.

 A filter that has the opposite characteristics is appropriately called a

highpass filter.
- is an electronic filter that passes signals with a frequency higher than a certain
cutoff frequency.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters
 These 3 filters cover the range from very sharp (ideal) to very
smooth (Gaussian) filter function.
(1) Ideal, (2) Butterworth and (3) Gaussian filters.

1) Ideal Lowpass filter:-

 A filter that “cut off” all high-frequency components of the Fourier
transform.
- Distance greater than a specified distance Do from the origin of the centered
transform.
 Such a filter is called a two-dimensional ideal filter (ILPF).
Ideal Lowpass Filters (ILPF)
1 if D(u, v)  D0
H (u, v)  
0 if D(u, v)  D0
@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)
5. Smoothing Frequency-Domain Filters

1) Ideal Lowpass filter:-

 In this case, the distance from any
point ( u, v) to the center (origin) of
the Fourier transform is given by:
D (u , v) = [ (u – M/2)2 + (v – N/2)2]1/2

 As the filter radius increase, less and

less power is removed, resulting in
less severe blurring.
- The circles superimposed on the spectrum
having radius of 5, 15, 30, 80, 230 pixels (the circle
of radius 5 is not easily visible).

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters
2) Butterworth Lowpass filter:-
 Butterworth filters are one of the most commonly used digital
filters in motion analysis.
 Cutoff frequency = Maximally flat magnitude filter
- Butterworth filter of order 1 has no ringing.
- Ringing generally is imperceptible in filters of order 2,
- but can become a significant factor in filters of higher order.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters

Butterworth Lowpass Filters (BLPF) of order n and

with cutoff frequency D0
1
H (u , v ) 
1   D(u, v) / D0 
2n

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

Comparison: ILPF vs BLPF

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters
 Gaussian filtering is used to blur images and remove noise and
detail. [ more effective at smoothing images]
 In one dimension, the Gaussian function is;

- where is the standard deviation of the distribution. The

distribution is assumed to have a mean of 0.
 Shown graphically, we see the familiar bell shaped Gaussian
distribution.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters
3) Gaussian Lowpass filter:-
By letting   D0
 D 2 ( u ,v )/2 D02
H (u, v)  e

- where Do is the cutoff frequency. When D (u,v) = Do, the filter is

down to 0.607 of its maximum value.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

5. Smoothing Frequency-Domain Filters
3) Gaussian Lowpass filter:-

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

 Examples: GLPF
3) Gaussian Lowpass filter:-
 Example is from the field of machine
perception, with application to
character recognition.
 (a) shows that the characters in this particular
document have distorted shapes due to lack
of resolution, and many of the characters are
broken.
- a machine recognition system has real
difficulties reading broken characters.
(a) (b)
 (b) shows how well characters can be
“repaired” by this simple process using
Gaussian lowpass filter with Do=80. The
image is size 444x508 pixels.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

 Examples: GLPF
3) Gaussian Lowpass filter:-
- the second is from the printing and publishing industry; and

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)

 Examples: GLPF
3) Gaussian Lowpass filter:-
- the third is related to processing satellite images.

@Copyrights: Digital Image Processing Organized by Dr. Ahmad Jalal (http://portals.au.edu.pk/imc/)