Scribd
Upload
18 views

Image Processing: Gaurav Gupta

This document discusses various image enhancement techniques, including: 1) Spatial domain methods that operate directly on pixel values, such as contrast stretching, sharpening filters, and median filtering for noise reduction. 2) Histogram processing techniques like equalization to improve an image's contrast. 3) Frequency domain filtering that applies transformations in the frequency domain rather than the spatial domain, allowing sharpening or blurring of images.

Uploaded by

Snehal Chothe
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
18 views

Image Processing: Gaurav Gupta

This document discusses various image enhancement techniques, including: 1) Spatial domain methods that operate directly on pixel values, such as contrast stretching, sharpening filters, and median filtering for noise reduction. 2) Histogram processing techniques like equalization to improve an image's contrast. 3) Frequency domain filtering that applies transformations in the frequency domain rather than the spatial domain, allowing sharpening or blurring of images.

Uploaded by

Snehal Chothe
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 38

Image Processing

Lecture 4

-Gaurav Gupta
Today

 Image Enhancement Techniques


 Spatial Domain Method
 Histogram Methods
 Frequency Domain Methods
Enhancement
Techniques

Frequency Domain
Spatial
Operates on FT of
Operates on pixels
Image
Spatial Domain Methods

 In these methods a operation (linear or non-


linear) is performed on the pixels in the
neighborhood of coordinate (x,y) in the input
image F, giving enhanced image F’
 Neighborhood can be any shape but
generally it is rectangular ( 3x3, 5x5, 9x9 etc)

g(x,y) = T[f(x,y)]
Grey Scale Manipulation

 Simplest form of window (1x1)


 Assume input gray scale values are in range
[0, L-1] (in 8 bit images L = 256)
 Nth root Transformation
s = c (r)n
contd…

 Linear: Negative, Identity


 Logarithmic: Log, Inverse Log
 Power-Law: nth power, nth root
Image Negative

Image Negative: s=L–1–r


Log Transformation

s = c log(1+r)
c: constant
 Compresses the dynamic range of images with
large variations in pixel values
Power Law Transformation

 s = crγ
 C,  : positive
constants
 Gamma
correction
Contrast Stretching
 To increase the dynamic range of the gray
levels in the image being processed.
contd…

 The locations of (r1,s1) and (r2,s2) control the shape


of the transformation function.

 If r1= s1 and r2= s2 the transformation is a linear function and


produces no changes.
 If r1=r2, s1=0 and s2=L-1, the transformation becomes a
thresholding function that creates a binary image.
 Intermediate values of (r1,s1) and (r2,s2) produce various
degrees of spread in the gray levels of the output image,
thus affecting its contrast.
 Generally, r1≤r2 and s1≤s2 is assumed.
Example
Bit-Plane Slicing
 To highlight the contribution made to the total
image appearance by specific bits.

 i.e. Assuming that each pixel is represented by 8


bits, the image is composed of 8 1-bit planes.
 Plane 0 contains the least significant bit and plane
7 contains the most significant bit.
 Only the higher order bits (top four) contain
visually significant data. The other bit planes
contribute the more subtle details.
Histogram Processing
 The histogram of a digital image with gray
levels from 0 to L-1 is a discrete function
h(rk)=nk, where:
 rk is the kth gray level
 nk is the # pixels in the image with that gray level
 n is the total number of pixels in the image
 k = 0, 1, 2, …, L-1
 Normalized histogram: p(rk)=nk/n
 sum of all components = 1
 Types of processing:

 Histogram equalization
 Histogram matching (specification)
 Local enhancement
Histogram Equalization
k nj k
sk  T (rk )     pr ( r j )
j 0 n j 0

 Histogram equalization (HE) results are similar to


contrast stretching but offer the advantage of full
automation, since HE automatically determines a
transformation function to produce a new image with
a uniform histogram.
Histogram Matching (or
Specification)
 Histogram equalization does not allow
interactive image enhancement and
generates only one result: an approximation
to a uniform histogram.

 Sometimes though, we need to be able to


specify particular histogram shapes capable
of highlighting certain gray-level ranges.
Method
 Specify the desired density function and obtain the
transformation function G(z):
z z
ni
v  G ( z )   p z ( w)  
0 i 0 n

pz: specified desirable PDF for output

– Apply the inverse transformation function z=G-


1
(s) to the levels obtained in step 1.
Image Smoothing or Averaging
 A noisy image:

g ( x, y )  f ( x , y )  n ( x , y )
 Averaging M different noisy images:
M
1
g ( x, y ) 
M
 g ( x, y )
i 1
i

 As M increases, the variability of the pixel values at each


location decreases.

 This means that g(x,y) approaches f(x,y) as the number of


noisy images used in the averaging process increases.
Example
Spatial Filtering

 Use of spatial masks for image processing


(spatial filters)

 Linear and nonlinear filters

 Low-pass filters eliminate or attenuate high


frequency components in the frequency
domain (sharp image details), and result in
image blurring.
a b
g(x, y)    w(s,t) f (x  s, y  t)
sa tb

a=(m-1)/2 and b=(n-1)/2,


m x n (odd numbers)

For x=0,1,…,M-1 and y=0,1,…,N-1

The basic approach is to sum products between the mask


coefficients and the intensities of the pixels under the mask at a
specific location in the image:
R  w1 z1  w2 z 2  ...  w9 z9 (for a 3 x 3 filter)
Neighborhood Averaging
General Spatial Filter
Non-linear Filter

 Median filtering (nonlinear)

 Used primarily for noise reduction (eliminates


isolated spikes)

 The gray level of each pixel is replaced by the


median of the gray levels in the neighborhood of
that pixel (instead of by the average as before).
Sharpening Filters

 The main aim in image sharpening is to


highlight fine detail in the image
 With image sharpening, we want to enhance
the high-frequency components; this implies
a spatial lter shape that has a high positive
component at the centre
Derivatives

 First derivative
f
 f (x  1)  f (x)
x
 Second derivative

 2 f
 f (x  1)  f (x 1)  2 f (x)
x 2
Observations

 1st order derivatives produce thicker edges in


an image
 2nd order derivatives have stronger response
to fine detail
 1st order derivatives have stronger response
to a gray lever step
 2nd order derivatives produce a double
response at step changes in gray level
 Since the sum of all the weights is zero,
the resulting signal will have a zero DC
value
Frequency Domain Methods

 We simply compute the Fourier transform of


the image to be enhanced, multiply the result
by a filter (rather than convolve in the spatial
domain), and take the inverse transform to
produce the enhanced image.
 Low pass filtering involves the elimination of
the high frequency components in the image.
It results in blurring of the image
Frequency Domain Methods
ThE eNd

You might also like