KOM3212 Image Processing in Industrial Systems Week 2 DI

KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
1
DIGITAL IMAGE FUNDAMENTALS
KOM3212 Image Processing in Industrial
Systems
Some of the contents are adopted from
R. C. Gonzalez, R. E. Woods, Digital Image Processing, 2nd edition, Prentice Hall, 2008
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Today’s lecture
• Elements of visual perception
• Visible Light Spectrum
• Imaging with Sensors
• Image Models
2
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
3
Elements of visual perception
• Although the digital image processing field is built on a foundation of
mathematical and probabilistic formulations mimicking the human vision
system plays an essential role.
Simplified Diagram of
Human Eye
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
4
Elements of visual perception
Three membranes enclose the eye
1.
The cornea and sclera outer cover : Cornea is a transparent tissue. Sclera
is an opaque membrane encloses the remainder of the optic globe
2.
The choroid : directly below Sclera. Blood vessels. Divided into 3 sections,
-Ciliary body: fibers and muscles
-Iris diaphragm: contracts or expands to balance the amount of light entering
-Lens : suspended by fibers that attach to the ciliary body. Helps focusing.
3.
The retina : the innermost membrane
-Cones vision (photopic / bright-light vision): centered at fovea, highly
sensitive to color. ~7 million
-Rods (scotopic /dim-light vision): general view. 75 to 150 million
-Blind spot
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
5
Elements of visual perception
Distribution of rods and cones on the retina
• The fovea itself is a circular indentation in the retina about 1.5mm in diameter.
• It can be regarded as 1.5mm x 1.5mm square sensor.
• For a CCD sensor the same number of elements can be on a 5mm x 5mm chip
• When the intelligence and light adaptivity is added human eye is outweighs the man-made
sensors
6
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Elements of visual perception
Cone distribution on
the fovea (200,000
cones/mm2)
•Models human visual system more precisely
•Hexagonal sampling requires fewer samples
than rectangular sampling
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
7
Image formation in the eye
• The distance between the center of the lens and the retina (focal length)
varies from 17mm to 14 mm as the refractive power of the lens increases
from its max to its min
• The shape of the lens is controlled by tension of the fibers in the ciliary body.
• To focus on distant objects lens is flattened.
• Lens is thickened to focus on near objects
8
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Image formation in the eye
electrical
impulses
radiant
energy
Light receptor
Brain
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Brightness adaptation
• Subjective brightness is what human
perceives.
• Dynamic range of human visual system
• 10-6 ~ 104
• Cannot accomplish this range
simultaneously
• For a given set of conditions current
sensitivity level is called brightness adaption
• Total adaption range is to large when
compared to brightness adaption level.
• The current sensitivity level of the visual
system is called the brightness adaptation
level
9
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Brightness discrimination
• Weber ratio (the experiment) DIc/I
• I: the background illumination to the opaque glass
• DI increment of illumination in the center
• DIc : the increment of illumination discriminable 50% of the time with
background illumination
• Small Weber ratio indicates good discrimination
• Larger Weber ratio indicates poor discrimination
10
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
11
Perceived brightness
The perceived brightness is not a
simple function of intensity
This side of the boundary is
perceived darker than it is
This side of the boundary is
perceived brighter than it is
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
12
Perceived brightness
All inner squares have the same intensity
but they appear to be brighter from left to right when background is set darker
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
13
Light and Electromagnetic Spectrum
• Electromagnetic spectrum can be expressed in terms of wavelength and frequency
𝑐
• Wavelength and frequency are related by: 𝜆 = 𝜈 𝑐 is the speed of the light,
𝜆 is the wavelength,𝜈 is the frequency.
• The energy of EM waves is : 𝐸 = ℎ𝜈 (ℎ is Planck constant, 𝐸 is the energy)
• Visible light represent a very small portion of the electromagnetic spectrum
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
14
Light and Electromagnetic Spectrum
• Electromagnetic waves can be regarded as a stream of mass-less particles each
travelling in a wavelike pattern and moving at the speed of light
• Each particle contains a certain amount of energy
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
15
Light and Electromagnetic Spectrum
• Light is a particular type of electromagnetic radiation
• 0.43m-violet to about 0.79m-red
• For convenience 6 bands are used to name the light in the visible spectrum
• The colors that human perceive on an object is determined by the light it is
reflected from it.
• A body that reflects all of the light bands will be perceived as white to the
user. If a body favors specific band and reflects it that we see it having that
color.
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
16
Color Perception of Human Eye
• Humans do not have a sensor for each frequency of light
• Instead, rods and cones are employed.
• Cones focus on recognizing brighter light (photopic vision) with more
sensitivity about individual colors,
• Rods on dim lighting scenarios (scotopic vision) with less color sensitivity, and
a combination of the two for in-between lighting (mesopic vision).
• cones detect light along something resembling a gaussian curve. There are
three wavelengths at which the cones are centered, as shown
blog.triplelift.com
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
17
Imaging Sensors
An image sensor converts a optical image
into an electronic signal. It is used mostly
in imaging devices.
Today, most digital still cameras use either
a CCD image sensor or a CMOS sensor
Neither technology has a clear advantage
in image quality
CMOS sensors can potentially be
implemented with fewer components, use
less power
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Imaging Sensors
18
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
19
Color Images in Cameras
There are several main types of color
image sensors, differing by the type of
color-separation mechanism:
One of them is
Bayes filter sensor: An array of color filters
is used. Original light is separated to its
bands and the resulting patterns are
registered to form a single image,
en.wikipedia.org
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
A simple image formation model
• f(x,y): the intensity is called the gray level for monochrome image
• f(x, y) = i(x, y).r(x, y)
• 0 < i(x, y) < inf, the illumination (lm/m2)
• 0< r(x, y) < 1, the reflectance
• Some illumination figures (lm/m2)
• 90,000: full sun
• 10,000: cloudy day
• 0.1: full moon
• 1,000: commercial office
- 0.01: black velvet
- 0.93: snow
20
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
A simple image formation model
21
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Sampling and Quantization
22
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Sampling and Quantization
23
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Sampling and Quantization
24
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
An effect of under-sampling, aliasing
 Aliasing (the Moire effect)
www.wfu.edu/~matthews/misc/DigPhotog/alias/
25
26
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
1x
1/2 x
1/4 x
1/8 x
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Uniform quantization
• Digitized in amplitude (or pixel value)
• PGM – 256 levels  4 levels
255
192
3
2
128
64
0
1
0
27
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
28
Uniform quantization
Original
8 bits
4 levels (2 bits)
16 levels (4 bits)
2 levels (1 bit)
KOM3212 Image Processing in Industrial Systems | Dr Muharrem Mercimek
Image resolution
• Spatial resolution
• Line pairs per unit distance
• Dots/pixels per unit distance
• dots per inch - dpi
• Intensity resolution
• Smallest discernible change in intensity level
• The more samples in a fixed range, the higher the resolution
• The more bits, the higher the resolution
29