Digital Image Processing

Lecture # 6
Image Enhancement in Spatial Domain- II
ALI JAVED

Lecturer

SOFTWARE ENGINEERING DEPARTMENT

U.E.T TAXILA

Email:: ali.javed@uettaxila.edu.pk

Office Room #:: 7

 Presentation Outline
 Local/ Neighborhood Operations on images

 Local Enhancement through Spatial Filtering

 Smoothing Spatial Filtering

 Linear Filters
 Non-Linear Filters

 Linear Smoothing Spatial Filtering

 Average Filters
 Weighted Average Filters

 Non-Linear Smoothing Spatial Filtering

 Median Filters
 Minimum Filters
 Maximum Filters

 Border Pixel Treatment

 Image Histogram

 Histogram Equalization
 Neighborhood Operations on Images
 Neighbourhood operations simply operate on a larger neighbourhood of
pixels than point operations

 Neighbourhoods are mostly selected a square regions around a central

pixel

 Any size and any shape filter are possible

Origin x

(x, y)
Neighbourhood

3/31/2011
y Image f (x, y)
 Local Operations through Spatial Filtering
 The output intensity value at (x,y) depends not only on the
input intensity value at (x,y) but also on the specified
number of neighboring intensity values around (x,y)

 Spatial masks (also called window, filter, kernel,

template) are used and convolved over the entire image for
local enhancement (spatial filtering)

 The size of the mask determines the number of

neighboring pixels which influence the output
value at (x,y)

 The values (coefficients) of the mask determine the

nature and properties of enhancing technique

3/31/2011
 Basics of Spatial Filtering

 Given the 3×3 mask with coefficients: w1, w2,…, w9

 The mask covers the pixels with gray levels: z1, z2,…, z9

 z gives the output intensity value for the processed

image (to be stored in a new array) at the location of z5 in
the input image

3/31/2011
 Local Operations through Spatial Filtering
x
a b c r s t
d
g
e
h
f
i
* u
x
v
y
w
z
Original Image Filter
Simple 3*3 Pixels
e 3*3 Filter
Neighbourhood
eprocessed = v*e +
r*a + s*b + t*c +
u*d + w*f +
y Image f (x, y) x*g + y*h + z*i

 The above is repeated for every pixel in the original image to

generate the smoothed image

3/31/2011
Local Operations through Spatial Filtering

3/31/2011
Types of Spatial Filtering

3/31/2011
 Spatial Filtering for Smoothing
 For blurring/noise reduction;

 Blurring is usually used in preprocessing steps, e.g., to remove

small details from an image prior to object extraction, or to bridge
small gaps in lines or curves

 Equivalent to Low-pass spatial filtering in frequency domain

because smaller (high frequency) details are removed based on
neighborhood averaging (averaging filters)

 Implementation:
 The simplest form of the spatial filter for averaging is a square
mask (assume m×m mask) with the same coefficients 1/m2 to
preserve the gray levels (averaging).

 Applications: Reduce noise; smooth false contours

 Side effect: Edge blurring

3/31/2011
 Smoothing Spatial Filters
 One of the simplest spatial filtering operations we can
perform is a smoothing operation

 Simply average all of the pixels in a neighbourhood

around a central value

 Especially useful in removing noise from images

1/ 1/ 1/
9 9 9

1/ 1/ 1/
9 9 9

1/ 1/ 1/
9 9 9

Simple Averaging Filter 3/31/2011

 Smoothing Spatial Filters
x
104 100 108

99 106 98

95 90 85
*
1/ 100
104
9
1/ 108
9
1/
9
Original Image
Simple 3*3 1/
99 1/ 1/
3*3 Smoothing Pixels
9 98
9 106 9
Neighbourhood 195
/9 190
/9 185
/9
Filter
e = 1/9*106 +
1/ *104 + 1/ *100 + 1/ *108 +
9 9 9
1/ *99 + 1/ *98 +
9 9
y 1/ *95 + 1/ *90 + 1/ *85
Image f (x, y) 9 9 9
= 98.3333
 The above is repeated for every pixel in the original image to
generate the smoothed image
3/31/2011
Spatial Filtering for Smoothing:: Example

3/31/2011
Spatial Filtering for Smoothing:: Example

3/31/2011
 Weighted Smoothing Spatial Filters
 More effective smoothing filters can be generated by allowing
different pixels in the neighbourhood different weights in the
averaging function

 Pixels closer to the central pixel are more important

 Often referred to as a weighted averaging

1/ 2/ 1/
16 16 16

2/ 4/ 2/
16 16 16

1/ 2/ 1/
16 16 16

Weighted Averaging Filter 3/31/2011

 Order Statistics Filtering
 Nonlinear spatial filters

 Output is based on order of gray levels in the masked area (sub-

image)

 Examples: Median filtering, Max & Min filtering

 Min: Set the pixel value to the minimum in the neighbourhood

 Max: Set the pixel value to the maximum in the

neighbourhood

 Median: The median value of a set of numbers is the midpoint

value in that set (e.g. from the set [1, 7, 15, 18, 24] 15 is the
median). Sometimes the median works better than the average

3/31/2011
 Order Statistics Filtering
Median Filtering
 Assigns the mid value of all the gray levels in the mask to the
center of mask;

 Particularly effective when

 the noise pattern consists of strong, spiky components ( salt-

and-pepper)

 edges are to be preserved

 Force points with distinct gray levels to be more like their

neighbors

3/31/2011
Order Statistics Filtering

3/31/2011
 Order Statistics Filtering
Minimum Filtering
 Assigns the minimum value of all the gray levels in the mask to the center
of mask;

 Particularly effective when

 Only Salt noise is present

Maximum Filtering
 Assigns the maximum value of all the gray levels in the mask to the center
of mask;

 Particularly effective when

 Only Pepper noise is present

 Edges are to be preserved

 Force points with distinct gray levels to be more like their neighbors
3/31/2011
Order Statistics Filtering

3/31/2011
Strange Things Happen At The Edges!
 At the edges of an image we are missing pixels to form a
neighbourhood

Origin x
e e

e e e
y Image f (x, y) 3/31/2011
 Border Pixels Treatment
 Mask operation near the image border

 Problem arises when part of the mask is located outside the

image plane; to handle the problem:

 Truncate the image

 Discard the problem pixels (e.g. 512x512 input 510x510 output, if

mask size is 3x3)

 Image Padding

 expand the input image by padding zeros or 255

(512x512 input 514x514output)

 Image padding is not good; creates artificial lines or edges on the

border
3/31/2011
 Border Pixels Treatment
 Replicate border pixels

 We normally use the gray levels of border pixels to fill up the expanded
region (for 3x3 mask). For larger masks a border region equal to half of the
mask size is mirrored on the expanded region.

3/31/2011
 Histogram
 The histogram of an image represents the frequency of pixels in an image

 The histogram of an image shows the distribution of gray levels in the

image

 Massively useful in image processing especially in image segmentation

 Histogram of an image with gray level (0 to L-1):

 How to obtain histogram?

 For B bit image, initialize 2^B counters with 0

 Loop over all pixels x,y
 When encouraging gray level f(x,y)=I, increment counter # i
Histogram : Example
 Histogram : Example
 A selection of images and their Histograms

 Note that the high contrast image has the most evenly spaced histogram

 Histograms of low contrast images are located in certain portions and not in
the entire gray scale range
 Histogram Equalization
 The idea is to find a transformation s=T(r) to be applied to each pixel of the
input image f(x,y) such that a uniform distribution of gray levels in the entire
range results for the output image g(x,y)
 Histogram Equalization
 Spreading out the frequencies in an image (or equalising the
image) is a simple way to improve dark or washed out images

 The formula for histogram equalisation is given where

sk  T (rk )
 rk: input intensity
 sk: processed intensity
 k: the intensity (e.g 0.0 – 1.0)
k
  pr ( r j )
 nj: the frequency of intensity j
 n: the sum of all frequencies
j 1
k nj

j 1 n
Histogram Equalization: Example
 Histogram Equalization: Example

Initial Image Image After Equalization

Notice that the minimum value (52) is now 0 and the maximum value (154) is now 255.
Histogram Equalization: Example
 Histogram Equalization: Example

No. of pixels

6
2 3 3 2 5

4 2 4 3 4

3
3 2 3 5
2
2 4 2 4
1
Gray level
4x4 image
0 1 2 3 4 5 6 7 8 9
Gray scale = [0,9]
histogram
 Histogram Equalization: Example
Gray
0 1 2 3 4 5 6 7 8 9
Level(j)

No. of 0 0 6 5 4 1 0 0 0 0
pixels

0 0 6 11 15 16 16 16 16 16

k nj 6 11 15 16 16 16 16 16
s 0 0 / / / / / / / /
j 1 n 16 16 16 16 16 16 16 16

3.3 6.1 8.4

sx9 0 0 9 9 9 9 9
3 6 8
 Histogram Equalization: Example

No. of pixels

6
3 6 6 3 5

8 3 8 6 4

3
6 3 6 9
2
3 8 3 8
1
Output image
0 1 2 3 4 5 6 7 8 9
Gray scale = [0,9] Gray level
Equalized histogram
Local Enhancement through Histogram Processing

 The histogram processing methods discussed earlier are global in

the sense that pixels are modified by a transformation function
based on the gray level content of the entire image

 Although this global approach is suitable for overall enhancement.

There are cases in which it is necessary to enhance details over
small areas in the image

 The histogram equalization can be easily adapted to local

enhancement

 The procedure is to define a neighborhood and move the center of

this area from pixel to pixel

 At each location, the histogram of the points in the neighborhood

is computed and then histogram transformation function is
obtained

3/31/2011
Local Enhancement through Histogram Processing

 Map the intensity of the pixel centered in the neighborhood

 Move to the next location and repeat the procedure

3/31/2011
 Use of Histogram Statistics for Image Enhancement

Average Intensity

L 1
m   ri p(ri )
M 1 N 1
1

MN
  f ( x, y )
x 0 y 0
i 0

Variance
M 1 N 1
1
    f ( x, y)  m
2

MN x 0 y 0

3/31/2011
Use of Histogram Statistics for Image Enhancement

Local average intensity

L 1
msxy   ri psxy (ri )
i 0

sxy denotes a neighborhood

Local variance
L 1
 2
s xy   (ri  msxy ) 2 psxy (ri )
i 0

 The following statistical conditions are used here for enhancement

 g(x,y) = { E.f(x,y) if mSxy <= k0MG AND k1DG <= <= k2DG }

 g(x,y) = f(x,y) otherwise

3/31/2011
Use of Histogram Statistics for Image Enhancement

3/31/2011
Any question