2014 Visikom 04 Grayscale Image Analysis (1)
-
Upload
fajar-ilman-saputra -
Category
Documents
-
view
221 -
download
0
Transcript of 2014 Visikom 04 Grayscale Image Analysis (1)
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
1/92
GRAYSCALE IMAGANALYSIS
Kuliah Visi KomputerJurusan Teknik Inform2014
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
2/92
Review Minggu 3: Summary
Images as 2-D signals
Linear and non-linear filters
Fourier Transform
Discrete Fourier Transform
Hybrid Images
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
3/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
4/92
Single Pixel Manipulation
The value of g(x,y) depends directly on the vf(x, y)
This is a greyscale transformation or simplmapping
Common mapping functions:
Identity transformation Thresholding
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
5/92
Common Mapping Functions
lightening increase contra
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
6/92
Common Mapping Functions
enhance contrast in dark regions enhance contrast in
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
7/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
8/92
Histogram Equalization
Ideally we want the im
spread uniformly over Image data squashed into asmall ran e of re values
stretch out the histogram to produce a more uniformdistribution
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
9/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
10/92
A very noisy
image
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
11/92
Using Neighborhood Pixels
So far we modified values of single pix What if we want to take neighborhood
account?
A common application of linear filtering
image smoothing using an averaging
or averaging maskor averaging kerne
Each oint in the smoothed ima e
The averaging filter is also known as the box filter.
Each pixel under the mask is multiplied by 1/9,
summed, and the result is placed in the output image
The mask is successively moved across the image.
That is, we convolvethe image with the mask.
1
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
12/92
Example of 2-D Convolution
Convolution with box filters of size 1,5,15,3,9,35 (reading or
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
13/92
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
],[],[],[
,
lnkmflkgnmh
lk
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
14/92
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 10
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
],[],[],[
,
lnkmflkgnmh
lk
f l
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
15/92
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 10 20
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
],[],[],[
,
lnkmflkgnmh
lk
fil i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
16/92
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 10 20 30
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
],[],[],[
,
lnkmflkgnmh
lk
I fil i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
17/92
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 10 20 30 30
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
],[],[],[
,
lnkmflkgnmh
lk
I filt i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
18/92
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 10 20 30 30
?
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
],[],[],[
,
lnkmflkgnmh
lk
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
19/92
I filt i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
20/92
0 10 20 30 30 30
0 20 40 60 60 60
0 30 60 90 90 90
0 30 50 80 80 90
0 30 50 80 80 90
0 20 30 50 50 60
10 20 30 30 30 30
10 10 10 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 90 0 90 90 90 0 0
0 0 0 90 90 90 90 90 0 0
0 0 0 0 0 0 0 0 0 0
0 0 90 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
[.,.]h[.,.]f
Image filtering . , . =
1 1 1
1 1 1
1 1 1
. , .
19
],[],[],[
,
lnkmflkgnmh
lk
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
21/92
Practice with Linear Filters
Original
?0 0 0
0 1 0
0 0 0
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
22/92
0 0 0
0 1 0
0 0 0
Practice with Linear Filters
Original Filtered
(no chang
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
23/92
0 0 0
0 0 1
0 0 0
Practice with Linear Filters
Original
?
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
24/92
Practice with Linear Filters
Original Shifted
By 1 pi
0 0 0
0 0 1
0 0 0
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
25/92
Practice with Linear Filters
Original
-(Note that filter sums to 1)1
9
1 1 1
1 1 1
1 1 1
0 0 0
0 2 0
0 0 0
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
26/92
Practice with Linear Filters
Original Sharpening filter
- Accentuates differenc
average
-19
1 1 1
1 1 1
1 1 1
0 0 0
0 2 0
0 0 0
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
27/92
Sharpening
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
28/92
Other Averaging Filters
One expects the value of a pixel to be more closely relatevalues of pixels close to it than to those further away
Accordingly it is usual to weight the pixels near the centremask more strongly than those at the edge
1 1 1
1 2 1
1 1 1
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
29/92
Other Averaging Filters
One expects the value of a pixel to be more closely rthe values of pixels close to it than to those further a
Accordingly it is usual to weight the pixels near the cthe mask more strongly than those at the edge
1
10
1 1 1
1 2 11 1 1
Important filter: Gaussian
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
30/92
0.003 0.013 0.02
0.013 0.059 0.09
0.022 0.097 0.15
0.013 0.059 0.09
0.003 0.013 0.02
5 x 5,
Slide credit: Christ
Important filter: Gaussian
Use Matlabsfspecial function to create a Gaussian filter.
is also know
Gaussian Filter
Smoothing with box filter
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
31/92
Smoothing with box filter
Example Box Filter Smoothing
Smoothing with Gaussian filter
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
32/92
Smoothing with Gaussian filter
Example Gaussian Smoothing
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
33/92
Key Properties of Linear Filt
Linearity:filter(f1+ f2) = filter(f1) + filter(f2)
Shift invariance: same behavior regardless of pix
filter(shift(f)) = shift(filter(f))
Any linear, shift-invariant operator can be representedconvolution
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
34/92
Key Properties of Linear Filt
Commutative:a* b= b* a Conceptually no difference between filter and signal
Associative:a* (b* c) = (a* b) * c Often apply several filters one after another: (((a* b1)
This is equivalent to applying one filter: a * (b1* b2* b
Distributes over addition: a* (b+ c) = (a* b) Scalars factor out: ka * b = a * kb = k (a* b)
Identity:unit impulse e= [0, 0, 1, 0, 0],a* e= a
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
35/92
Gaussian Filters
Linear filters Remove high-frequency components
the image (low-pass filter)
Images become more smooth
Convolution of a Gaussian with a Gaus
another Gaussian
So can smooth with small-width kernel, re
and get same result as larger-width kerne
Separability of the Gaussian fil
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
36/92
Separability of the Gaussian fil
Gaussian Filters
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
37/92
Practical Matters
What about near the edge? The filter window falls off the
edge of the image
Need to extrapolate
Methods:
clip filter (black) wrap around
copy edge
reflect across edge
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
38/92
Practical Matters
Methods (MATLAB): clip filter (black): imfilter(f, g, 0)
wrap around: imfilter(f, g, circ
copy edge: imfilter(f, g, repl
reflect across edge: imfilter(f, g, symm
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
39/92
Practical Matters MATLAB: filter2(g, f, shape)
shape = full: output size is sum of sizes of fan shape = same: output size is same as f
shape = valid: output size is difference of sizes
f
gg
gg
f
gg
gg
g
g
full same
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
40/92
Low-Pass Filtering
Removing all highspatial frequencies fsignal to retain only lowspatial frequen
called low-pass filtering.
Old Spectrum New Spectrum Low-Pass
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
41/92
High-Pass Filtering
Removing all lowspatial frequencies frsignal to retain only highspatial freque
called high-pass filtering.
Old Spectrum New Spectrum High-Pass
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
42/92
Low-Pass FilteringAn Exa
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
43/92
Low-Pass FilteringAn Exa
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
44/92
High-Pass FilteringAn Exa
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
45/92
High-Pass FilteringAn Exa
S
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
46/92
Filtering in the Spatial Doma
Low-pass filtering -> convolve the imagbox / Gaussian filter
High-pass filtering -> ?
Filt i i th S ti l D
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
47/92
Filtering in the Spatial Doma Low-pass filtering -> convolve the image with a box / Gaussian
High-pass filtering ->
Since the sum of the weights is 0, the resulting signal will havevalue (i.e., the average value or the coefficient of the zero freqcomponent) .
To display the image you will need to take the absolute values.
-1/9 -1/9 -1/9
-1/9 8/9 -1/9
-1/9 -1/9 -1/9
N Li Filt i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
48/92
Non-Linear Filtering Neighbourhood averaging or Gaussian smoothing will tend to blur edges becaus
frequency components are attenuated
An alternative approach is to use median filtering. Here we set the grey level toof the pixel values in the neighbourhood
Example: pixel values in 3 3 neighbourhood
Sort the values 10 15 20 20 20 20 20 25 100
Pixels with outlying values are forced to become more like their neighbours
10 20 20
20 15 2020 25 100
20
median v
M di Filt i A E
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
49/92
Median FilteringAn Examp
Medremo
Med
smoo
imagblurr
edge
Hi h B t Filt i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
50/92
High-Boost Filtering Here we take the original image and boost the high frequency compon
Can think of HighPass = OriginalLowPass. Thus
HighBoost = b*OriginalLowPass
= (b-1)*Original + OriginalLowPass
= (b-1)*Original + HighPass
bis the boosting factor.
When b=1, HighBoost = HighPass
High-boost filteringuseful for emphasizing high frequencies while low frequency components of the original image to aid in the interpretaimage
Hi h B t Filt i ( t )
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
51/92
High Boost Filtering (cont.)
How can we perform high-boost filterinspatial domain? -1/9 -1/9 -1/9-1/9 w/9 -1/9
-1/9 -1/9 -1/9
where w = 9*b-
Original High Boosted
intermediate result
High Boost
contrast stretc
H hi Filt i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
52/92
Homomorphic Filtering
The brightness of an image point(, function of the illumination at that pointreflectance of the object at that point, i.
, = , (, )illumination0 , < reflectanc
0 ,
H hi Filt i ( t
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
53/92
Homomorphic Filtering (cont
Let , = log , = log , + log ,
In the frequency domain, we have , = , + ,
Fourier transform
of log( , )
Fourier transform
of log( , )
If we apply a high
then we can supp
frequency illumina
components and
reflectance comp
Homomorphic Filtering (cont.
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
54/92
FFTHigh-boost
filteringFFT-1(, ) log exp
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
55/92
Smoothing and Sub samplin
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
56/92
Smoothing and Sub-samplin
In many Computer Vision applications, sub-samplin
needed, e.g., to build an image pyramid, or
simply to reduce the resolution for efficient storage, trprocessing,
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
57/92
Throw away every other row and
column to create a 1/2size image
Sub sampling Issues
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
58/92
Sub-sampling Issues
Sub-sampling may be dangerous. WResample tcheckerboa
one sample
In the case
board, new
is reasonab
Top right als
reasonable
Bottom left
(dubious) a
has checks
big.
Aliasing in Videos
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
59/92
Aliasing in Videos
Aliasing in Graphics
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
60/92
Aliasing in Graphics
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
61/92
Sampling and Aliasing
Top row shows the images, sampled at every second pixel to ge
bottom row shows the magnitude of frequency spectrum of thes
Anti aliasing
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
62/92
Anti-aliasing
1. Sample more often
2. Get rid of all frequencies that are greater tthe new sampling frequency(Nyquistfrequency)
Will lose information
But its better than aliasing
Apply a smoothing filter
S li d Ali i
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
63/92
Sampling and Aliasing
Sampling with smoothing. Top row shows the images. We get the nex
smoothing the image with a Gaussian with sigma 1 pixel, then samplin
pixel to get the next; bottom row shows the magnitude spectrum of the
Subsampling without Pre-filt
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
64/92
Subsampling without Pre-filt
1/4 (2x zoom) 1/8 (41/2
Subsampling with Gaussian Prfiltering
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
65/92
filtering
G 1/4 GGaussian 1/2
Slide by
Image Pyramids
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
66/92
Image Pyramids
Key component of many
high level computer vision
tasks
How to create an image
pyramid?
Image Pyramids
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
67/92
Image Pyramids
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
68/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
69/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
70/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
71/92
The Correspondence Proble
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
72/92
The Correspondence Proble
Basic assumptions:
Most scene points are visible in both ima
Corresponding image regions are similar
These assumptions hold if:
The distance of points from the cameras larger than the distance between camera
The Correspondence Proble
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
73/92
The Correspondence Proble
Is a search problem:
Given an element in the left image, search focorresponding element in the right image.
We will typically need geometric constraints the size of the search space
We must choose: Elements to match
A similarity measure to compare elements
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
74/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
75/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
76/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
77/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
78/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
79/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
80/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
81/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
82/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
83/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
84/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
85/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
86/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
87/92
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
88/92
Summary
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
89/92
Single Pixel Operations
Histogram Equalization
Filtering in the Spatial Domain
Subsampling and Anti-aliasing
Template Matching
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
90/92
Lab Work 4
Next Week: Edge detection
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
91/92
Template Matching
Lab Work 3
Tugas 3
-
8/10/2019 2014 Visikom 04 Grayscale Image Analysis (1)
92/92
Buatlah dua fungsi Matlab berikut:
1. function score = ssd(t, im)
2. function score = ncc(t, im)untuk gambar input im dan template t.
Setiap fungsi menghasilkan matriks score yang ukurannya sim.
Buatlah script untuk menguji kedua fungsi tsb yang menampinput, gambar template, dan matriks score. Beri keterangan
bagaimana menginterpretasikan matriks score yang ditampilKumpulkan dua file yang menguji dua pasang gambar dan teberbeda
Contoh gambar input dan template bisa didownload di elearn