Lecture 04

Spatial Domain Filtering

 Contents

In this lecture we will look at spatial domain filtering:

 What is image filtering?
 Correlation
 Convolution
 Smoothing linear filter
 Median filter
 Sharpening

Image Filtering 2
 What is filtering

 Filtering is the process of replacing a pixel with a value

based on some operations or functions.
 The operations/functions used on the original image
are called filters.
• or masks, kernels, templates, windows…
 In digital image processing, filters are usually used to
• suppress the high frequencies in an image
- i.e., smoothing the image

• suppress the low frequencies in an image

- i.e., enhancing or detecting edges in the image

Image Filtering 3
 What is image filtering

f(x,y) g(x,y)

filtering

filtering

Image Filtering 4
 Image filtering methods

 Spatial domain

 Frequency domain

Image Filtering 5
 Spatial domain methods

Image Filtering 6
 Point processing methods

 Convert a given pixel value to a new pixel value

based on some predefined function

Image Filtering 7
 Point processing methods-Example

Negative Contrast stretching

Thresholding

Histogram
Equalization

Image Filtering 8
 Limitations of Point Operations

 They don’t know where they are in an image

 They don’t know anything about their neighbors
 Most image features (edges, textures, etc) involve a
spatial neighborhood of pixels
 If we want to enhance or manipulate these features,
we need to go beyond point operations

Image Filtering 9
 What Point Operations Can’t Do

 Blurring/Smoothing

Image Filtering 10
 What Point Operations Can’t Do

 Sharpening

Image Filtering 11
 Spatial Filters

 The general processes can be denoted by the

expression:
g ( x, y)  T [ f ( x, y)]
• f(x,y) is the input image
• g(x,y) is the processed image
 T is an operator on f, defined over some
neighborhood of (x,y)
• use a subimage area centered at (x,y)
• shapes of the neighborhood
- Circle, square, rectangular

Image Filtering 12
 Spatial Filters

 Example: 3×3 neighborhood about a point (x,y) in an

image
x

(x,y) (x-1,y-1) (x,y-1) (x+1,y-1)

(x-1,y) (x,y) (x+1,y)

Image f(x,y)
(x-1,y+1) (x,y+1) (x+1,y+1)
y

Image Filtering 13
 Spatial Filters
x

w(-1,-1) w(0,-1) w(1,-1)

Mask

w(-1,0) w(0,0) w(1,0)

Image f(x,y)
w(-1,1) w(0,1) w(1,1)
y
f(x-1,y-1) f(x,y-1) f(x+1,y-1) Mask coefficients

f(x-1,y) f(x,y) f(x+1,y)

Pixels under mask

f(x-1,y+1) f(x,y+1) f(x+1,y+1)

Image Filtering 14
 Operation

 Typically linear combinations

of pixel values.
• e.g., weight pixel values and
add them together.

 Different results can be

obtained using different
weights.
• e.g., smoothing, sharpening,
edge detection

Image Filtering 15
 Example

10 5 3 0 0 0
4 6 1 0 0.5 0 8
1 1 8 0 1 0.5
Local image mask Modified image data
neighborhood

Image Filtering 16
 Correlation

 A filtered image is generated as the center of the mask

visits every pixel in the input image.
n x n mask
h(i,j)

g(i,j)

f(i,j)

filtered
image

Image Filtering 17
 Handling pixels close to boundaries

pad with zeroes wrap around

0 0 0 ……………………….0
0 0 0 ……………………….0

or

Image Filtering 18
 Geometric interpretation of correlation

 Suppose x and y are two n-dimensional vectors:

x  ( x1 , x2 ,..., xn ) y  ( y1 , y2 ,..., yn )

 The dot product of x with y is defined as

x
x. y  x1 y1  x2 y2  ...  xn yn
using vector
θ
notation:
x . y | x || y |cos( )
y

 Correlation generalizes the notion of dot product

Image Filtering 19
 Geometric interpretation of correlation

 cos(θ) measures the similarity between x and y

xy
x . y | x || y |cos( ) or cos( ) 
| x || y |
 Normalized correlation (i.e., divide by lengths)
n n
2 2

  h(k , l ) f (i  k , j  l )
n n
k  l 
N (i, j )  n n
2 2
n n
2 2 2 2
[  h 2 (k , l )]1/ 2 [   f 2 (i  k , j  l )]1/ 2
n n n n
k  l  k  l 
2 2 2 2

Image Filtering 20
 Normalized correlation

 Measure the similarity between images or parts of

images.

mask

Image Filtering 21
 Application example

Image Filtering 22
 Convolution

 Same as correlation except that the mask is flipped,

both horizontally and vertically.
1 2 3 7 8 9 9 8 7
H V
4 5 6 4 5 6 6 5 4
7 8 9 1 2 3 3 2 1
n n n n
2 2 2 2
g (i, j )    h(k , l ) f (i  k , j  l )    h(i  k , j  l ) f (k , l )
n n n n
k  l  k  l 
2 2 2 2

 For symmetric masks (i.e., h(i,j)=h(-i,-j)),

convolution is equivalent to correlation! Notation:
h*f=f*h

Image Filtering 23
 Convolution example

Convolution kernel, h Input Image, f

1 -1 -1 2 2 2 3
1 2 -1 2 1 3 3
1 1 1 2 2 1 2
Rotate 180o 1 3 2 2
1 1 1
-1 2 1
-1 -1 1

Image Filtering 24
 Convolution example

1 1 1 2 2 2 3

-1 2 1 2 1 3 3

-1 -1 1 2 2 1 2
1 3 2 2
1 1 1
2
-1 4 2 2 3 5
Output
2 1
-1 -2 3 3 Image, g
2 2 1 2
1 3 2 2
Input Image, f
Image Filtering 25
 Convolution example
2 2 2 3
1 1 1
2 1 3 3
-1 2 1
2 2 1 2
-1 -1 1
1 3 2 2
1 1 1
2 4
-2 2 2 3 5 4
Output
2 -1
-2 1 3 3 Image, g
2 2 1 2
1 3 2 2
Input Image, f
Image Filtering 26
 Convolution example
1 1 1 2 2 2 3
 Same
-1 2 1 2 1 3 3
-1 -1 1 2 2 1 2
1 3 2 2
1 1 1
2 -2
2 4
2 3 5 4 4
Output
2 -1
1 -3
3 3 Image, g
2 2 1 2
1 3 2 2
Input Image, f
Image Filtering 27
 Convolution example
1 1 1 2 2 2 3
 Same
-1 2 1 2 1 3 3
-1 -1 1 2 2 1 2
1 3 2 2
1 1 1
2 2 2 3
-2 6 1 5 4 4 -2
Output
2 1 3 -3
-3 3 1 Image, g
2 2 1 2
1 3 2 2
Input Image, f
Image Filtering 28
 Convolution example
2 2 2 3
1 1 1
2 1 3 3
-1 2 1
2 2 1 2
-1 -1 1
1 3 2 2

1 2 2 2 3 5 4 4 -2
Output
-1 4
2 1 3 3 9 Image, g
-1 -2
2 2 1 2
1 3 2 2
Input Image, f
Image Filtering 29
 Convolution example

5 4 4 -2
9 6 14 5
11 7 6 5
9 12 8 5
Final output Image, g

Image Filtering 30
 Linear spatial filtering

R   (1, 1) f ( x  1, y  1)   (1, 0) f ( x  1, y)  ...

  (0, 0) f ( x, y )  ...   (1, 0) f ( x  1, y)   (1,1) f ( x  1, y  1)

 Smoothing Linear Filters : known as averaging

filter or lowpass filters.

Box filter Weighted

average filter
Image Filtering 31
 Smoothing linear filters
Original 3x3

5x5 9x9

15x15 35x35

Image Filtering 32
 Smoothing linear filters

original blurred thresholded

Image Filtering 33
 Spatial Filters – smoothing, linear

 Mean filters 1 1 1
• example: 1
 1 1 1
9
1 1 1

Gaussian noise
Original 5×5
3×3 mean
mean filter
filter
Image Filters 34
 Spatial Filters – smoothing, linear

 Mean filters 1 1 1
• example: 1
 1 1 1
9
1 1 1

Salt and pepper 3×3 mean filter

5×5
Image Filters 35
 Spatial Filters – smoothing, linear

 Weighted average filters

• example: 1 2 1
1
 2 4 2
• general expression: 16
1 2 1

a b

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

  w(s, t )
s   at   b

Image Filters 36
 Spatial Filters – smoothing, nonlinear

 Order-statistic filters

• nonlinear spatial filters

• order/rank the pixels contained in the image area

encompassed by the filter

Image Filters 37
 Spatial Filters – smoothing, nonlinear

 Median filters
• replace a pixel value with the median of its neighboring
pixel values
• example:

23 25 26 30 40
Neighborhood values: 22 24 26 27 35
15, 19, 20, 23, 18 20 50 25 34
24, 25, 26, 27, 50 19 15 19 23 33
Median value: 24 11 16 10 20 30

Image Filters 38
 Spatial Filters – smoothing, nonlinear

 Median filters
• have excellent noise-reduction capabilities

V.S.

Gaussian noise removed Gaussian noise removed

by 3×3 mean filter By 3×3 median filter

Image Filters 39
 Spatial Filters – smoothing, nonlinear

 Median filters
• are particularly effective in salt & pepper

V.S.

Salt & pepper removed Salt & pepper removed

by 3×3 mean filter By 3×3 median filter
Image Filters 40
 Spatial Filters – smoothing, nonlinear

 Max filters
• maximum of neighboring pixel values
• useful for finding the brightest points in an image

 Min filters
• minimum of neighboring pixel values
• useful for finding the darkest points in an image

Image Filters 41
 Spatial Filters – sharpening

 Principal objective
• highlight fine detail in an image
• enhance detail that has been blurred
 Sharpening can be accomplished by spatial
differentiation
 For one dimensional function f(x)
• first order derivative
f
 f ( x  1)  f ( x)
x
• second order derivative
2 f
 f ( x  1)  f ( x  1)  2 f ( x)
x 2

Image Filters 42
 Spatial Filters – sharpening

• A sample

(a) a scan line (b) image strip

(c) first derivative (d) second derivative

Image Filters 43
 Spatial Filters – sharpening

 The Laplacian: second derivative of a two dimensional

function f(x,y)
 2
f  2
f
 f  2  2
2

x y
= [f(x+1,y)+f(x-1,y)+f(x,y+1)+f(x,y-1)] -4f(x,y)
 Use a convolution mask to approximate

0 1 0 1 1 1 -1 2 -1
1 -4 1 1 -8 1 2 -4 2
0 1 0 1 1 1 -1 2 -1

Image Filters 44
Spatial Filters – sharpening

 The Laplacian example

Image Filters 45
 Spatial Filters – sharpening

 The Laplacian
• example:

Image Filters 46
 Summary

 We have looked at:

• Spatial domain operations
• Convolution and correlation
• Smoothing linear filter
• Median filter
• Laplacian

 Next time we will continue our talk about frequency domain

filtering

Image Filtering 47