Upload
rakeshsoni
View
230
Download
0
Embed Size (px)
Citation preview
8/7/2019 IP CH 1& CH 2
1/88
Rakesh SoniM.E. (CSE)
Assistant Professor,
PIET
MOTIVATION FOR IP
8/7/2019 IP CH 1& CH 2
2/88
WHYSTUDY IP:
BECAUSE
IP has applications
In all walks of human life
8/7/2019 IP CH 1& CH 2
3/88
What is an image?
a representation, likeness, or imitation of an
object or thing a vivid or graphic description
something introduced to represent something
else
One Picture is worth more than ten
thousand words.
8/7/2019 IP CH 1& CH 2
4/88
An Image is defined as a two-dimensional function, f(x,y),
where x and y are spatial (plane) coordinates, and the amplitude
of f at any pair of coordinates (x,y) is called the intensity or
gray level of the image at that point.
When x, y and the intensity values of f are all finite, discretequantities, we call the image a digital image.
The elements are called picture elements, image elements, pels
and pixels.
8/7/2019 IP CH 1& CH 2
5/88
I. P. APPLICATIONS:
Health Care and Medical diagnostics.
Resources Surveying. Industrial Applications.
Security and Surveillance.
Water/ Irrigation project management. Military combat operations.
Environment and Pollution control.
8/7/2019 IP CH 1& CH 2
6/88
Case-I:
Medical Diagnosis
8/7/2019 IP CH 1& CH 2
7/88
A surgeon is viewing an X-ray plate of
patient suspected to behaving cancerous
growth in chest area. As it is soft-tissue
X-ray, contrast is inadequate to locate the
cancer accurately.
8/7/2019 IP CH 1& CH 2
8/88
8/7/2019 IP CH 1& CH 2
9/88
The surgeon can take one of the two
decisions
1.To go ahead with the operation
2.Not to do the operation.
8/7/2019 IP CH 1& CH 2
10/88
If the surgeon decides on the
first choice and opens up the
body and finds that no cancerous cells at
all. The patient unnecessarily goes through
surgical reghours of medical operation.
8/7/2019 IP CH 1& CH 2
11/88
If the surgeon decides not to operate and if
there is cancerous cell growth, it will
rapidly spread in the entire body and
ultimately kill the patient, in a few weeks
time.
8/7/2019 IP CH 1& CH 2
12/88
Image enhancement techniques in IP
Can compliment the surgeon andassist him to take correctdecision,Know precisely the location
ofCancerous cell growth, thus confinethe operation to limited area.
SAVE THE PATIENT.
WHAT CAN IP DO?:
8/7/2019 IP CH 1& CH 2
13/88
8/7/2019 IP CH 1& CH 2
14/88
Case-II:
Industrial Inspection
8/7/2019 IP CH 1& CH 2
15/88
LSI Devices manufacturing plant receives a
large quantity of raw materials, SiliconWafers, with some impurities, not possible
to detect using normal methods.
8/7/2019 IP CH 1& CH 2
16/88
Manufacturing proceeds and LSI devices are
produced in bulk. Entire batch gets rejectedas it fails to meet the specifications. All
foundry capacity, time, effort gets wasted.
8/7/2019 IP CH 1& CH 2
17/88
WHAT CAN IP DO?:
With proper IP tools, it is possible to
detect impurity levels exceeding limits atraw material stage itself.
I. P. thus saves wastages
And
Boosts productivity.
8/7/2019 IP CH 1& CH 2
18/88
8/7/2019 IP CH 1& CH 2
19/88
8/7/2019 IP CH 1& CH 2
20/88
CASE-III:
MILITARY COMBATS
8/7/2019 IP CH 1& CH 2
21/88
Modern military combats involve
Air Raids with aim of destroying militarybases and thus weaken the enemy.
8/7/2019 IP CH 1& CH 2
22/88
If AR operations are carried out
blindly, it will destroy civil amenities,hospitals, schools etc. Military bases
may remain unaffected.
Waste of AR effort and expenses.
8/7/2019 IP CH 1& CH 2
23/88
Killing innocent civilians raises
Hue & cry at UN bodies and creates a
world sympathy for the enemy.
8/7/2019 IP CH 1& CH 2
24/88
WHAT CAN IP DO?:
With IP one can get precise locations
of Military Bases and weapon storagelocations, through spying ventures.
Thus AR operations can be precisely
targeted to destroy the enemy fully.
8/7/2019 IP CH 1& CH 2
25/88
8/7/2019 IP CH 1& CH 2
26/88
IMAGE PROCESSING
For Life saving, Efficiency and Efficacy .
A common thread in all the above cases is:
Even though only illustrative cases are given
above, IP plays vital roll for variety of applications
namely,R
esourcesS
urveying,
Security &
Surveillance, Water & Irrigation projects,
Astronomy & science search, Environmental &
Pollution control And many many other fields.
8/7/2019 IP CH 1& CH 2
27/88
WHAT is IP ?
It is an engineering science in which we
capture two dimensional pictureinformation and process it using digital
computing facilities.
The information is then compared with the
vast knowledge/data base on the subject,
for effective interpretation and correct
decision making.
8/7/2019 IP CH 1& CH 2
28/88
Engineering Science, Two-D Information,
Digital Computing Interpretation,Decision making and KNOWLEDGE-BASE
KEYWORDS in IP are:
Image Processing consists of:Image Acquisition.Image digitization & sampling.Image Processing.Image Interpretation.
Image Compression.
8/7/2019 IP CH 1& CH 2
29/88
Digital Image Processing Why ?
1. Improvement of pictorial information for human
interpretation
2. Processing of image data for storage ,
transmission and representation for autonomous
machine perception
8/7/2019 IP CH 1& CH 2
30/88
Digital Image Processing
Process digital images by means of computer, it covers
low-, mid-, and high-level processes
low-level: inputs and outputs are images
mid-level: outputs are attributes extracted
from input images
high-level: an ensemble of recognition ofindividual objects
8/7/2019 IP CH 1& CH 2
31/88
8/7/2019 IP CH 1& CH 2
32/88
8/7/2019 IP CH 1& CH 2
33/88
8/7/2019 IP CH 1& CH 2
34/88
8/7/2019 IP CH 1& CH 2
35/88
8/7/2019 IP CH 1& CH 2
36/88
8/7/2019 IP CH 1& CH 2
37/88
8/7/2019 IP CH 1& CH 2
38/88
8/7/2019 IP CH 1& CH 2
39/88
8/7/2019 IP CH 1& CH 2
40/88
8/7/2019 IP CH 1& CH 2
41/88
8/7/2019 IP CH 1& CH 2
42/88
8/7/2019 IP CH 1& CH 2
43/88
8/7/2019 IP CH 1& CH 2
44/88
8/7/2019 IP CH 1& CH 2
45/88
8/7/2019 IP CH 1& CH 2
46/88
Elements of visual perception:
Field of
DIPis built on a foundation of mathematicaland probabilistic formulations.
Human intuition and analysis play central role in the
choice of one technique versus another.
Developing a basic understanding of human visual
perception as a first step.
8/7/2019 IP CH 1& CH 2
47/88
Structure of Human Eye:
8/7/2019 IP CH 1& CH 2
48/88
Fig. shows the density of rods and cones for a cross section of the right eye passing
through the region of emergence of the optic nerve from the eye.
The absence of receptors in this area results in the so-called blind spot.
Except for this region, the distribution of receptors is radially symmetric about fovea.
Receptor density is measured in degrees from the fovea.
8/7/2019 IP CH 1& CH 2
49/88
Image formation in the Eye:
15/100=h/17 orh=2.55 mm
8/7/2019 IP CH 1& CH 2
50/88
Image Sensing & Acquisition:
Image Acquisition
Using
Sensor Arrays:
Transformilluminationenergy into
digital images
8/7/2019 IP CH 1& CH 2
51/88
Image Acquisition Using Single Sensor:
8/7/2019 IP CH 1& CH 2
52/88
Image Acquisition Using
Sensor Strips:
8/7/2019 IP CH 1& CH 2
53/88
A Simple Image formation model:
8/7/2019 IP CH 1& CH 2
54/88
A Simple Image Formation Model
( , ) ( , ) ( , )
( , ) : intensity at the point ( , )
( , ) : illumination at the point ( , )
(the amount of source illumination incident on the scene)
( , ) : reflectance/transmissivity
f x y i x y r x y
f x y x y
i x y x y
r x y
! g
at the point ( , )
(the amount of illumination reflected/transmitted by the object)
where 0 < ( , ) < and 0 < ( , ) < 1
x y
i x y r x yg
8/7/2019 IP CH 1& CH 2
55/88
Some Typical Ranges of illumination
IlluminationLumen A unit of light flow or luminous flux
Lumen per square meter (lm/m2) The metric unit of measure forilluminance of a surface
On a clear day, the sun may produce in excess of 90,000 lm/m2 of illuminationon the surface of the Earth
O
n a cloudy day, the sun may produce less than 10,000 lm/m2
of illuminationon the surface of the Earth
On a clear evening, the moon yields about 0.1 lm/m2 of illumination
The typical illumination level in a commercial office is about 1000 lm/m2
8/7/2019 IP CH 1& CH 2
56/88
Some Typical Ranges of Reflectance
Reflectance
0.01 for black velvet
0.65 for stainless steel
0.80 for flat-white wall paint
0.90 for silver-plated metal
0.93 for snow
8/7/2019 IP CH 1& CH 2
57/88
8/7/2019 IP CH 1& CH 2
58/88
Range of subjective sensations showing a particular adaptationlevel
8/7/2019 IP CH 1& CH 2
59/88
Basic Experimental setup used to characterize brightnessdiscrimination
Weber ratio as a function of intensity
8/7/2019 IP CH 1& CH 2
60/88
Perceived brightness is not a simple function of intensity. The relative verticalpositions betn two profiles in (b) have no special significance; they were chosenfor clarity
8/7/2019 IP CH 1& CH 2
61/88
All inner squares have same intensity, but they appear progressivelydarker as the background becomes lighter
8/7/2019 IP CH 1& CH 2
62/88
Optical illusions
8/7/2019 IP CH 1& CH 2
63/88
Frasers spiral
8/7/2019 IP CH 1& CH 2
64/88
8/7/2019 IP CH 1& CH 2
65/88
8/7/2019 IP CH 1& CH 2
66/88
Image Sampling and Quantization
8/7/2019 IP CH 1& CH 2
67/88
8/7/2019 IP CH 1& CH 2
68/88
8/7/2019 IP CH 1& CH 2
69/88
8/7/2019 IP CH 1& CH 2
70/88
8/7/2019 IP CH 1& CH 2
71/88
8/7/2019 IP CH 1& CH 2
72/88
Effects of varying no. of samples in Digital Image
8/7/2019 IP CH 1& CH 2
73/88
(a) 1024*1024, 8-bit image. (b) 512*512 image resampled into1024*1024 pixels by row andcolumn duplication. (c) through (f) 256*256, 128*128, 64*64, an
32*32 images resampled into1024*1024 pixels.
8/7/2019 IP CH 1& CH 2
74/88
8/7/2019 IP CH 1& CH 2
75/88
8/7/2019 IP CH 1& CH 2
76/88
8/7/2019 IP CH 1& CH 2
77/88
8/7/2019 IP CH 1& CH 2
78/88
Image Zooming
steps: 1) the creation of new pixel locations,
2) the assignment of gray levels to those new
locations.
Methods:
1) Nearest neighbor interpolation2) Pixel replication:
3) Bilinear interpolation:
8/7/2019 IP CH 1& CH 2
79/88
8/7/2019 IP CH 1& CH 2
80/88
Image Shrinking
Methods :
For integer factor row-column deletion
For noninteger factorzooming grid analogy
8/7/2019 IP CH 1& CH 2
81/88
Neighbors of a Pixel
A pixel p at coordinates (x, y) has fourhorizontaland verticalneighbors whose coordinates are given by
(x+1, y), (x-1, y), (x, y+1), (x, y-1)
This set of pixels, called the 4-neighbors of p, is denoted by N4(p).Each pixel is a unit distance from (x, y), and some of neighbors ofp lie outside the digital image if (x, y) is on the border of the image.
The fourdiagonalneighbors of p have coordinates
(x+1, y+1), (x+1, y-1), (x-1, y+1), (x-1, y-1)
and are denoted by ND(p).
These points, together with the 4-neighbors, are called the 8-eightneighbors of p, denoted by N8(p).
As before, some of the points in ND(p) and N8(p) fall outside theimage if (x, y) is on the border of the image.
8/7/2019 IP CH 1& CH 2
82/88
Neighbors of a Pixel
Connectivity
Adjacency
Regions
Boundaries
8/7/2019 IP CH 1& CH 2
83/88
Adjacency
Let Vbe the set of gray-level values
4-adjacency:Two pixels p and q with value from Vare4-adjacent if q is in the set N4(p).
8-adjacency:Two pixels p and q with value from Vare
8-adjacent if q is in the set N8(p).
m-adjacency: (mixed adjacency). Two pixels p and q with
values from Vare m-adjacentIf,
(i) q is in N4(p), or
(ii) q is in ND(p) andthe set N4(p) N4(q) has no pixels whosevalues are from V.
8/7/2019 IP CH 1& CH 2
84/88
(a) Arrangement of pixels; (b) pixels that are 8-adjacent (shown dashed)
to the center pixel; (c) m-adjacency.
8/7/2019 IP CH 1& CH 2
85/88
Distance Measures
For pixels p, q, and z, with coordinates (x, y), (s, t),
and (v, w), respectively, D is a distance function
ormetricif
8/7/2019 IP CH 1& CH 2
86/88
Euclidean distance
F
or this distance measure, the pixels having a distanceless than or equal to some
value rfrom (x, y) are the points contained in a disk of
radius r centered at (x, y).
8/7/2019 IP CH 1& CH 2
87/88
D4 distance (city-block distance)
8/7/2019 IP CH 1& CH 2
88/88
TheD8 distance (chessboard distance)