The relevant features for the examination task are enhanced The irrelevant features for the examination task are removedreduced Here the input and output image are both digital image in color or gray scale ID: 814688
Download The PPT/PDF document "Image Enhancement Image enhancement refe..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Image Enhancement
Image enhancement refers to the class of image processing operations whose goal is to produce an output digital image that is visually more suitable as appearance for its visual examination by a human observer
The relevant features for the examination task are enhanced
The irrelevant features for the examination task are removed/reduced
Here the input and output image are both digital image in color or gray scale.
Slide2Slide3The following operations are generally done on the image during image enhancement
noise removal;
geometric distortion correction;
edge enhancement;
contrast enhancement;
image zooming;
image subtraction;
pseudo-coloring.
Slide4Image enhancement techniques can be divided into two broad categories:
1.Spatial domain methods,
which
operate directly on pixels,
and
2.frequency
domain methods,
which
operate on the Fourier transform of an image
.
Unfortunately
, there is no general theory for determining what is `good' image enhancement when it comes to human perception. If it looks good, it is good!
we will discuss them one by one……
Slide5Slide6Spatial Domain Methods
Slide7Spatial Domain: Content
Slide8Point Operations
Slide9Contrast Stretching
Increasing the contrast of the output image(say J) such that J maps the values in original image(say I) such that some of the data is saturated at low and high intensities of I.
Slide10Contrast Stretching
Clipping:
Thresholding
:
Output : Binary image
v
a
b
u
v
a
v
b
L
L
Point Operations
Slide11Contrast stretching:
When
When
v
a
b
u
v
a
v
b
L
L
Slide12Contrast Stretching
Contrast Stretching, a = 80, b= 160
Slide13Original Image
Contrast stretched
Clipped
Thresholded
Slide14Grey level: Gamma
C
orrection
In this operation, the value of
is corrected to enhance the image.
narrow I range to a wider dynamic range
to a narrow range
Thus in this way the gamma is corrected as required.
Its predominantly used for image
capture,display
and printing
v
u
Gamma
Correction
Original
gamma=0.45
gamma=2.20
v
is the output and u is the input image
Slide16Image negative, image transform
As is evident by name, it is used to take the complement of the image, or its negative
(
s:out
r:input)
s
r
Slide17Log Transform
In log transform, we compress dynamic range of images that have large variation in intensities.
It’s the most common display for Fast Fourier Transforms(FFTs)
s
r
Slide18Bit Extraction
Range Compression
:
Image Subtraction and Change Detection
medical imaging application : display blood-flow paths
automated inspection of printed circuits
security monitoring
Other Point Operations
Slide19Histogram
The histogram (h) of an image with L gray levels is defined as:
,
Where,
is the k-
th
gray level and
is the number of pixels with intensity
.
Histogram Processing
Slide22Histogram equalization
The goal of histogram equalization is to equalize the histogram, it spreads the histogram of input image over a large range of intensities
v is uniformly distributed over (0,1).
There can be other methods to equalize the histogram as well, thus we have some different choices for F(u). Some of them are:
etc
Histogram Equalization Result
Slide25Histogram Equalization: its types
SHE: Standard Histogram Equalization
BHE: Bi Histogram Equalization
RMHE: Recursive Mean-Separate Histogram Equalization
CHE: Clipped Histogram Equalization
BOHE: Block-Overlapped Histogram Equalization
GACHE: Gain-Adjustable Clipped Histogram Equalization
They are used for contrast enhancement of the image
Slide26Contrast enhancement by various types of HE
(a) Original (b) SHE (c) BHE (d) CHE
Slide27Discrete case
u
w
Example
Find the transformation between
u
and
v
0
1
2
3
0
1
2
3
Histogram
Specification
Slide28Histogram Matching
Histogram matching refers to the comparison of histogram of one image to that of other.
For example, in the below image, the histograms of the image of a type and Brain-image are matched.
Slide29Histogram mapping of image 1
:
Histogram of image 1 is
:
Histogram matching to image 2 can be done as
:
Thus in this way histograms of several images are matched.
Spatial Operations
They refer to the operations that are applied to the pixels having some common features and which can be enhanced in group.
There are various spatial operations that are applied on the image, some of them are
Slide31It is used in
noise smoothing, low pass filtering, sub-sampling
Revisiting simple random variable theory
Spatial Averaging and Spatial Low-pass Filtering
Slide32-
Mean-filtering
-
Noise
reduction
Spatial Averaging
Slide33Spatial averaging masks
a(
k,l
)
(a) 2 x 2 window
(b) 3 x 3 window
(c) 5-point weighted
averaging
l
k
0
0
0
0
l
k
Spatial averaging masks
a
(
k, l
)
l
k
Example of Spatial
Averaging
Slide34Spatial Filtering
Original image
Lowpass filtered image
Highpass filtered image
Bandpass filtered image
Slide35Directional Smoothing
Directional smoothing is a technique which is used to overcome the problems associated with simple averaging mask.in this method we:
Compute
spatial average along several directions
Take
the result from the direction giving the smallest
changes before
& after filtering
Slide36It is used to
protect the edges from blurring while
smoothing process is going on..
l
k
Directional Smoothing
Slide37It can be of two types, vertical smoothing and diagonal smoothing.
In vertical smoothing we smoothen the image vertically while in diagonal smoothing diagonal smoothening is carried on.
Vertical smoothing
Diagonal smoothing
Slide38Median filtering
It is a non-linear smoothing operator which has the following advantages
resilient
to statistical outliers
incurs
less blurring
simple
to
implement
median value ξ over a small window of size
Nw
odd window size is commonly used
3x3
, 5x5,
7x7
5-pixel
“+”-shaped window
Slide39Example
Median Filtering
Slide4040
Comparison-Median Filters with Different Sizes
Original
3x3
5x5
9x9
Slide41Experiment-Gaussian Noise
Noisy
(Gaussian)
3x3
5x5
9x9
Slide42Unsharp Masking
Unsharp masking
is an image manipulation technique
for increasing
the apparent sharpness of photographic images.
The
"unsharp" of the name derives from the fact that the
technique uses
a blurred, or "unsharp", positive to create a "mask" of
the original
image.
The
unsharped
mask is then combined with
the negative
, creating a resulting image sharper than the original.
Slide43Steps of Unsharp
Masking
Blur
the image
Subtract
the
blurred version
from the
original (this
is called the mask)
Add the “mask” to
the original image.
Slide44Signal
Low-pass
High-pass
(1)
(2)
(3)
(1)+
(3)
Unsharp Masking and
Crispening
Sharpened signal
Blurred signal
Unsharp mask
Slide45Image of moon with unsharp mask
Original Image
Image with unsharp mask applied
Slide46Contrast
ratio:
Inverse contrast ratio: enhance weak edges
Inverse Contrast Ratio Mapping and Statistical Scaling
Slide4747
Replication
Linear
Interpolation
P-
th
order linear interpolation
Linear interpolation and Magnification
Interlacing by
p
rows and
columns zeros
Convolve
H
H
H
1
2
p
Slide48Original image
Lowpass filtered image
Interpolation
Subsampled
(1/4)
Zero interlace
(2x)
1st Zooming
2nd Zero interlace
(4x)
2nd Zooming
Slide4949
The goal is
to attract the attention of observers
A mapping from Gray scale to Color
Color
coordinate
transformation
Display
Feature
extraction
Input
images
R
G
B
False Color and Pseudo Color
Slide50Color image enhancement
The basic idea behind color image enhancement is to enhance every component image by filtering out R,G,B and then mixing the enhanced versions so as to enhance the colored image.
Slide51Basic idea
example
:
(Histogram equalization )
Coordinate
conversion
Monochrome
image enhancement
Monochrome
image enhancement
Monochrome
image enhancement
Inverse
coordinate
transformation
Display
R
G
B
Input
image
Color Image Enhancement
Slide52END