Spatial Operations

Introduction
Fundamentals
Image Enhancement Techniques
Speech and Image Processing Lab

IMAGE ENHANCEMENT TECHNIQUES
Jisa David & Seena Symon

M.Tech. Signal Processing
Jisa David & Seena Symon M.Tech. Signal Processing Speech and Image Processing Lab IMAGE ENHANCEME
Introduction
Fundamentals
INTRODUCTION
DIGITAL IMAGE & IMAGE PROCESSING

Image:2D function f(x,y) where x and y are spatial coordinates
and f is the intensity.
Digital image :Representation of a 2D image as a finite set of
digital values( pixels).
Digital Image Processing:Processing a digital image by means
of a digital computer.
Introduction
Fundamentals
FUNDAMENTALS
STEPS IN IMAGE PROCESSING

1 Image Aquisition
2 Image Enhancement
3 Image Restoration
4 Morphological processing
5 Segmentation
6 Representation and Description
7 Recognition
Introduction
Spatial Domain
Fundamentals
Frequency Domain
IMAGE ENHANCEMENT TECHNIQUES
IMAGE ENHANCEMENT
Processing of an image so that the result is more suitable than orignal
image for Specific Application.
Done in Two domain

1 Spatial Domian
2 Frequency Domain
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL DOMAIN
Gray Level Transformation

Image Negatives
Log Transformations
Power-Law transformation
Piecewise Linear Transformation
Contrast Stretching
Gray Level Slicing
Bit Plane Slicing
Spatial Filtering
Neighborhood Averaging
Median Filtering
Histogram processing
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL DOMAIN
GRAY LEVEL TRANSFORMATION

Operated on individual pixels intensity values: s = T(r). r:
original intensity, s: new intensity
Data independent pixel-based enhancement method.
Approaches
1 Image negatives
2 Log transform
3 Power law transform
Introduction
Spatial Domain
Fundamentals
Frequency Domain

IMAGE NEGATIVE
Function reverses the order from black to white so that the
intensity of the output image decreases as the intensity of the
input increases.
s= T(r) = L-1-r
Used mainly in medical images and to produce slides of the screen
Figure: (a)Original mammogram ,(b)Negative image
Introduction
Spatial Domain
Fundamentals
Frequency Domain
LOG GRAY LEVEL TRANSFORM

s = T(r) = c log(1+r)
expand dark value to enhance details of dark
area.Example:Fourier 2D Transform.
Figure: (a)Fourier Spectrum ,(b)Transformed image with c=1
Introduction
Spatial Domain
Fundamentals
Frequency Domain
POWER LAW GRAY-LEVEL TRANSFORM

s = T(r) = crγ
Gamma correction: To compensate the built-in power law
compression due to display characteristics. CRT intensity to
voltage response is a power function with γ=1.8-2.5.
Figure: (A)Original Image, (B)Response of monitor to A (C)Gamma

Corrected Image,(D)Response of monitor to C
Introduction
Spatial Domain
Fundamentals
Frequency Domain
Figure: a) log transformation b)power-law transformation
Introduction
Spatial Domain
Fundamentals
Frequency Domain
PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

Allow more control on the complexity of T(r)
Examples
Contrast stretching
Gray-level slicing
Bit-plane slicing
Introduction
Spatial Domain
Fundamentals
Frequency Domain
CONTRAST STRETCHING
To increase the dynamic range of the gray levels in the image
being processed.
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.
Introduction
Spatial Domain
Fundamentals
Frequency Domain
CONTRAST STRETCHING
Figure: (A)Transformation function,(B) A low Contrast Image,(C)Result of

Contrast Stretching,(D)Result of Thresholding
Introduction
Spatial Domain
Fundamentals
Frequency Domain

GRAY LEVEL SLICING
1 To highlight a specific range of gray levels in an image (e.g. to
enhance certain features).
2 The second approach is to brighten the desired range of gray
levels but preserve the background and gray-level tonalities in the
image
Figure: (a)Transformation function of (1),(b)Transformation function of (2)
Introduction
Spatial Domain
Fundamentals
Frequency Domain
Figure: a) original image b)Transformed image
Introduction
Spatial Domain
Fundamentals
Frequency Domain
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.
Application in image compression
Introduction
Spatial Domain
Fundamentals
Frequency Domain
BIT PLANE SLICING
Introduction
Spatial Domain
Fundamentals
Frequency Domain
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.
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 Z2 + ......... + W9 Z9 (1)
(for 3x3 filter)
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
Figure: Mechanism of Spatial Filtering

Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
NEIGHBORHOOD AVERAGING
Reduce the noise
Averaging using M neighborhood Pixels
As M increases, the variability of the pixel values at each location
decreases
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
NEIGHBORHOOD AVERAGING
Figure: (a)image, (b)Corrupted Image,(c)-(f) result of averaging with M

=8,16,64,128
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
MEDIAN FILTERING
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).
Median filters are nonlinear.
Median filtering reduces noise without blurring edges and other

sharp details.
Median filtering is particularly effective when the noise pattern

consists of strong, spike-like components. (Salt-and-pepper
noise.)
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
MEDIAN FILTERING
Figure: (a)Image correpted by Salt & Pepper Noise, (b)Noise reduction

with 3x3 averaging mask, (c)Noise reduction with 3x3 median filtering.
Introduction
Spatial Domain
Fundamentals
Frequency Domain
SPATIAL FILTERING
EDGE ENHANCEMENT
digital image processing filter that improves the apparent
sharpness of an image or video
Application in TV broad cast and DVD
Introduction
Spatial Domain
Fundamentals
Frequency Domain
HISTOGRAM PROCESSING
Data-dependent pixel-based image enhancement method.
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 No: of 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(an estimate of the

probability of occurrence of gray level rk )
Introduction
Spatial Domain
Fundamentals
Frequency Domain
HISTROGRAM PROCESSING
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
FILTERING IN FREQUENCY DOMAIN
Compute Fourier transform of image
Multiply the result by a filter transfer function (or simply filter).
Take the inverse transform to produce the enhanced image.
Figure: Basic Step for Filtering in the Frequency Domain
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
ENHANCEMENT IN FREQUENCY DOMAIN

Types of enhancement that can be done:
Lowpass filtering: Reduce the high-frequency content – blurring
or smoothing
Highpass filtering: increase the magnitude of high-frequency
components relative to low-frequency components – sharpening.
Figure: (a)Frequency response of Low pass fillter and (b)High pass filter
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
BUTTERWORTH LOW PASS FILTER

This filter does not have a sharp discontinuity establishing a clear
cutoff between passed and filtered frequencies
Transfer functiion
1
H(u, v) = 1+[D(u,v)/D0 ]2n
where D(u, v) = [(u − M/2)2 + (v − N/2)2 ]
where n -order of the filter,M-Number of rows,N -Number of

columns,D0 -Cut off frequency
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
Figure: a) original image b)Noisy image c)LPF image

Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
BUTTERWORTH HIGH PASS FILTER

Edges and sharp transitions in grayvalues in an image contribute
significantly to high-frequency content of its Fourier transform.
Image sharpening in the Frequency domain can be done by
attenuating the low-frequency content of its Fourier transform.
Transfer functiion
1
H(u, v) = 1+[D0 /D(u,v)]2n
where D(u, v) = [(u − M/2)2 + (v − N/2)2 ]

columns,D0 -Cut off frequency
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
Figure: a)Original image b)HPF image
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FREQUENCY DOMAIN
BUTTERWORTH BAND PASS FILTER

Band-boost filtering boosts certain midrange freqencies and
partially corrects for blurring, but does not boost the very high
(most noise corrupted) frequencies.
Obtained by cascading low pass and High pass filters.
Transfer functiion
1
H(u, v) = (1+[D(u,v)/D1 ]2n )(1+[D0 /D(u,v)]2n )
where D(u, v) = [(u − M/2)2 + (v − N/2)2 ]

columns,D0 - Lower Cut off frequency ,D1 -Upper Cut off
frequency
Introduction
Spatial Domain
Fundamentals
Frequency Domain
FALSE COLORING
A false-color image is an image that depicts a subject in colors
that differ from human perception of the same subject.
A false-color image is derived from a greyscale image by mapping
each pixel value to a color according to a table or function.
False-color images are frequently used for viewing satellite
images, such as from weather satellites.
Figure: (a)Original Image, (b)False Color Image

Introduction
Spatial Domain
Fundamentals
Frequency Domain
REFERENCE
Rafael C. Gonzalez, richard E. Woods,Digital Image

Processing,Pearson Education.
Anil K. Jain, Fundamentals of Digital Image Processing,Prentice
Hall
Rafael C. Gonzalez, richard E. Woods,Digital Image Processing
using MATLAB,Pearson Education.
www.wikipedia.org

Spatial Operations

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Spatial Operations

Încărcat de

Drepturi de autor:

Formate disponibile

Introduction

Speech and Image Processing Lab

Jisa David & Seena Symon

DIGITAL IMAGE & IMAGE PROCESSING

STEPS IN IMAGE PROCESSING

IMAGE ENHANCEMENT TECHNIQUES

Done in Two domain

Gray Level Transformation

GRAY LEVEL TRANSFORMATION

Data independent pixel-based enhancement method.

3 Power law transform

GRAY LEVEL TRANSFORMATION

Figure: (a)Original mammogram ,(b)Negative image

GRAY LEVEL TRANSFORMATION

LOG GRAY LEVEL TRANSFORM

Figure: (a)Fourier Spectrum ,(b)Transformed image with c=1

GRAY LEVEL TRANSFORMATION

POWER LAW GRAY-LEVEL TRANSFORM

Figure: (A)Original Image, (B)Response of monitor to A (C)Gamma

GRAY LEVEL TRANSFORMATION

Figure: a) log transformation b)power-law transformation

GRAY LEVEL TRANSFORMATION

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

Figure: (A)Transformation function,(B) A low Contrast Image,(C)Result of

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

Figure: (a)Transformation function of (1),(b)Transformation function of (2)

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

Figure: a) original image b)Transformed image

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

BIT PLANE SLICING

PIECE-WISE LINEAR GRAY-LEVEL TRANSFORM

BIT PLANE SLICING

Use of spatial masks for image processing (spatial filters)

(for 3x3 filter)

Figure: Mechanism of Spatial Filtering

Figure: (a)image, (b)Corrupted Image,(c)-(f) result of averaging with M

Median filters are nonlinear.

Median filtering reduces noise without blurring edges and other

Median filtering is particularly effective when the noise pattern

Figure: (a)Image correpted by Salt & Pepper Noise, (b)Noise reduction

Data-dependent pixel-based image enhancement method.

The histogram of a digital image with gray levels from 0 to L-1 is

rk is the kth gray level

nk is the No: of pixels in the image with that gray level

n is the total number of pixels in the image

Normalized histogram: p(rk )=nk /n(an estimate of the

Figure: Basic Step for Filtering in the Frequency Domain

ENHANCEMENT IN FREQUENCY DOMAIN

BUTTERWORTH LOW PASS FILTER

where D(u, v) = [(u − M/2)2 + (v − N/2)2 ]

where n -order of the filter,M-Number of rows,N -Number of

Figure: a) original image b)Noisy image c)LPF image

BUTTERWORTH HIGH PASS FILTER

where D(u, v) = [(u − M/2)2 + (v − N/2)2 ]

where n -order of the filter,M-Number of rows,N -Number of

Figure: a)Original image b)HPF image

BUTTERWORTH BAND PASS FILTER

where D(u, v) = [(u − M/2)2 + (v − N/2)2 ]

where n -order of the filter,M-Number of rows,N -Number of

Figure: (a)Original Image, (b)False Color Image

Rafael C. Gonzalez, richard E. Woods,Digital Image

S-ar putea să vă placă și