Professional Documents
Culture Documents
GROUP MEMBERS
(Sp10-Bce-011)
(Sp10-Bee-068)
(Sp10-bce-031)
PROJECT SUPERVISOR
&
CO-SUPERVISOR
PRESENTATION LAYOUT
Database (IMBA home public access library &
INOR)
Image Acquisition
Pre-Processing
Gray level slicing
Connected Components and labelling
Morphological operations
Erosion
and dilation
Features
Support vector machine
FLOW CHART
Image
Acquisition
PreProcessing
Segmentation
Feature Extraction
Post Processing
DATABASE
Training
Reference:
https://eddie.via.cornell.edu/cgi-bin/datac/signon.cgi
IMAGE ACQUISITION
PRE-PROCESSING
Gray scale image
(elimination of hue and saturation)
o Histogram Equalization
Overview:
It
Application:
Contrast
Refrence:
Digital image processing by S Jayaraman, S
Esakkirajan , S Veerakumar
Check
Neighbors
Neighbors Not
Labeled
Assign New
Label to pixel
Pixel is
Labeled
Yes
Get Labels Parent
Add to Existing
List
Yes
Parent
is in
Pattern
list
No
on to the next pixel, this one has a neighbour which is already labeled
We increment currentLabelCount
and assign it to the pixel, again its
parent is set to itself
A few more rounds and we should end up with this. Notice the
blue number in the upper right corner, that's the parent
label, the de facto one upon which we aggregate later.
MORPHOLOGICAL OPERATIONS
Erosion
Dilation
Combine to
Openning
Closening
Object
Background
STRUCTURING ELEMENT
EROSION
IMAGE OF EROSION
HOW IT WORKS???
EROSION EXAMPLE
difference
erosion
2.
3.
AB {x Z 2
{x Z 2 x b A for every b B )
Fills in holes.
Smoothes object
boundaries.
Adds an extra outer ring of
pixels onto object
boundary, ie, object
becomes slightly larger.
IMAGE OF DILATION
EXAMPLE OF DILATION
difference
dilation
MATHEMATICAL DEFINATION OF
DILATION
Dilation : x = (x1,x2) such that if we center B on them,
then the so translated B intersects X.
IMAGE OF MORPHOLOGICAL
OPERATIONS
TRAINING
Area (305)
Eccentricity (0.5828)
Perimeter (84.7696)
Mean (7.4599e-4)
Extent (0.6689)
MAXIMUM MARGIN:
FORMALIZATION
w: decision hyperplane normal vector
xi: data point i
Classifier is:
f(xi) = sign(wTxi + b)
yi (wTxi + b)
Sec. 15.1
GEOMETRIC MARGIN
wT x b
Distance from example to the separator is r y
w
Finding r:
Dotted line xx is perpendicular to
decision boundary so parallel to w.
Unit vector is w/|w|, so line is rw/|
w|.
x = x yrw/|w|.
x satisfies wTx+b = 0.
So wT(x yrw/|w|) + b = 0
since |w| = sqrt(wTw).
So wTx yr|w| + b = 0
So, solving for r gives:
43
r = y(wTx + b)/|w|
Sec. 15.1
wTxi + b 1 if yi = 1
wTxi + b 1 if yi = 1
w x b
w
2
w
44
wTxa + b = 1
Hyperplane
wT x + b = 0
Sec. 15.1
wTxb + b = -1
wT x + b = 0
= ||xaxb||2 = 2/||w||2
45
Sec. 15.1
2
w
yi (wTxi + b) 1
46
Sec. 15.1
47
Sec. 15.2.3
NON-LINEAR SVMS
x
48
Sec. 15.2.3
: x (x)
49
Sec. 15.2.3
Sec. 15.2.3
KERNELS
Common kernels
Linear
Polynomial
K(x,z) = (1+xTz)d
Radial
51
TIMELINE
1st
Presentation
2nd
Presentation
3rd
Presentation
4th
Presentation
5th
Presentation
Study of project
Image Acquisition and Pre-Processing
Gray Level Slicing and Connected Components Labeling
Feature Extraction and SVM
Presenting the project to external examiner
52