Documente Academic
Documente Profesional
Documente Cultură
http://commons.wikimedia.org/wiki/File:Eye-diagram_bg.svg
Light
Light
Particles known as photons
Act as waves
Two fundamental properties
Amplitude
Wavelength
Frequency is the inverse of wavelength
Relationship between wavelength (lambda) and frequency (f)
c/ f
Where c = speed of light = 299,792,458 m / s
4
What is Digital Image Processing?
Digital image processing focuses on two major tasks
Improvement of pictorial information for human interpretation
Processing of image data for storage, transmission and
representation for autonomous machine perception
Some argument about where image processing ends and
fields such as image analysis and computer vision start
What is DIP? (cont)
The continuum from image processing to computer vision
can be broken up into low-, mid- and high-level processes
paper industry
The Bartlane cable picture
transmission service
Images were transferred by submarine cable between London
Early digital image
and New York
Pictures were coded for cable transfer and reconstructed at
the receiving end on a telegraph printer
History of DIP (cont)
Mid to late 1920s: Improvements to the Bartlane system
resulted in higher quality images
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
New reproduction
processes based
on photographic
techniques
Increased number
of tones in
reproduced images
Improved
digital image Early 15 tone digital image
History of DIP (cont)
1960s: Improvements in computing technology and the
onset of the space race led to a surge of work in digital
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
image processing
1964: Computers used to
improve the quality of
images of the moon taken
by the Ranger 7 probe
Such techniques were used
in other space missions
including the Apollo landings
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Image Aquisition
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Image Enhancement
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Image Restoration
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Morphological Processing
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Segmentation
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Object Recognition
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Representation & Description
Images taken from Gonzalez & Woods, Digital Image Processing (2002)
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Image Compression
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Colour Image Processing
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Object
Acquisition Recognition
Representation
Problem Domain
& Description
Colour Image Image
Processing Compression
Visible Spectrum
22
Conventional Coordinate for Image Representation
RGB components
10 10 16 28
9 656 70 26
56 43
3756 78
32 99 54 96 67
70
15 256013 902296 67
21 54 47 42
85 85 43 92
32 15 87 39
54 65 65 39
32 65 87 99
Image Types : Binary Image
Binary data
0 0 0 0
0 0 0 0
1 1 1 1
1 1 1 1
Image Types : Index Image
Index image
Each pixel contains index number
pointing to a color in a color table
Color Table
29
Anatomy of the Human Eye
30 Source: http://webvision.med.utah.edu/
Human Visual System
Human vision
Cornea acts as a protective lens that roughly focuses
incoming light
Iris controls the amount of light that enters the eye
The lens sharply focuses incoming light onto the retina
Absorbs both infra-red and ultra-violet light which can damage the
lens
The retina is covered by photoreceptors (light sensors)
which measure light
31
Photoreceptors
Rods
Approximately 100-150 million rods
Non-uniform distribution across the retina
Sensitive to low-light levels (scotopic vision)
Lower resolution
Cones
Approximately 6-7 million cones
Sensitive to higher-light levels (photopic vision)
High resolution
Detect color by the use of 3 different kinds of cones each of
which is sensitive to red, green, or blue frequencies
Red (L cone) : 564-580 nm wavelengths (65% of all cones)
Green (M cone) : 534-545 nm wavelengths (30% of all cones)
Blue (S cone) : 420-440 nm wavelengths (5% of all cones)
33
Cone (LMS) and Rod (R) responses
34 http://en.wikipedia.org/wiki/File:Cone-response.svg
Photoreceptor density across retina
35
Comparison between rods and cones
Rods Cones
Used for night vision Used for day vision
Loss causes night blindness Loss causes legal blindness
Low spatial resolution with higher noise High spatial resolution with lower noise
Not present in fovea Concentrated in fovea
Slower time response to light Quicker time response to light
One type of photosensitive pigment Three types of photosensitive pigment
Emphasis on motion detection Emphasis on detecting fine detail
36
Color and Human Perception
Chromatic light
has a color component
Achromatic light
has no color component
has only one property intensity
37
Image Formation in the Human Eye
39
Weber Ratio I/I
Weber Ratio
Human Visual Perception
Light intensity:
The lowest (darkest) perceptible intensity is the scotopic threshold
The highest (brightest) perceptible intensity is the glare limit
The difference between these two levels is on the order of 1010
We cant discriminate all these intensities at the same time! We adjust
to an average value of light intensities and then discriminate around the
average.
Log compression.
Experimental results show that the relationship between the
perceived amount of light and the actual amount of light in a scene
are generally related logarithmically.
The human visual system perceives brightness as the logarithm of the actual
light intensity and interprets the image accordingly.
Consider, for example, a bright light source that is approximately 6times
brighter than another. The eye will perceive the brighter light as
approximately twice the brightness of the darker.
42
Brightness Adaptation and Mach Banding
When viewing any scene:
The eye rapidly scans across the eld of view while coming
to momentary rest at each point of particular interest.
At each of these points the eye adapts to the average
brightness of the local region surrounding the point of
interest.
This phenomena is known as local brightness adaptation.
Mach banding is a visual effect that results, in part, from local
brightness adaptation.
The eye over-shoots/under-shoots at edges where the brightness
changes rapidly. This causes false perception of the intensities
Examples follow.
43
Brightness Adaptation and Mach Banding
44
Brightness Adaptation(Hermann Grid)
45
46
Optical illusion
48
Simultaneous Contrast
49
Image Sensing and acquisition
Single sensor
Line sensor
Array sensor
Gate
Photosites
Vertical Transport Register
Image Sensor: Inside Charge-Coupled Device
Gate
Horizontal Transportation Register
i h g Image pixel
f e d
c b a
i h g
f e d
i h g
Horizontal transport
register c b a
f e d
Horizontal shift
Digital Image Acquisition Process
Original image
Sampled image
1mm
2mm
No detail is lost!
Minimum Nyquist Rate:
Spatial resolution
Period Spatial resolution must be less or equal
(sampling rate)
half of the minimum period of the image
or sampling frequency must be greater or
= Sampling locations
Equal twice of the maximum frequency.
Aliased Frequency
x1 (t ) sin( 2t ), f 1
1 x2 (t ) sin( 12t ), f 6
0.5
-0.5
-1
0 0.5 1 1.5 2
Sampling rate:
1
5 samples/sec
0.5
-0.5
-1
0 0.5 1 1.5 2
Image quantization:
discretize continuous pixel values into discrete numbers
Nc 2 b
Nc-1
Nc-2
Quantization level
1
0
Light intensity
Darkest Brightest
Intensity Resolution
It refers to the smallest discernible change in intensity
level
Number of intensity levels usually is an integer power of
two
Also refers to Number of bits used to quantize intensity
as the intensity resolution
Which intensity resolution is good for human perception
8 bit, 16 bit, or 32 bit
Effect of Quantization Levels or Intensity resolution
64 levels 32 levels
Effect of Quantization Levels (cont.)
or Intensity resolution
16 levels 8 levels
In this image,
it is easy to see
false contour.
4 levels 2 levels
How to select the suitable size and pixel depth of images
V(x, y) = aij xi yj ( i, j = 0 to 3)
Need to solve sixteen equations
Gives better results than other methods
More complex
Used in Adobe Photoshop, and Corel Photopaint
Basic Relationship of Pixels
(0,0) x
(x,y-1) 4-neighbors of p:
(x-1,y)
(x-1,y) p (x+1,y)
(x+1,y)
N4(p) = (x,y-1)
(x,y+1)
(x,y+1)
(x-1,y-1)
(x-1,y) p (x+1,y)
(x,y-1)
(x+1,y-1)
(x-1,y)
(x-1,y+1) (x,y+1) (x+1,y+1) (x+1,y)
N8(p) = (x-1,y+1)
(x,y+1)
(x+1,y+1)
(x-1,y-1)
p
(x+1,y-1)
ND(p) = (x-1,y+1)
(x+1,y+1)
(x-1,y+1) (x+1,y+1)
Connectivity is adapted from neighborhood relation. Two pixels are connected if they
are in the same class (i.e. the same color or the same range of intensity) and they are
neighbors of one another.
S1
S2
q
p
8-path m-path
p p p
q q q
m-path from p to q
8-path from p to q
solves this ambiguity
results in some ambiguity
Distance
w D(p,q) = D(q,p)
De ( p, q) ( x - s )2 + ( y - t ) 2
Distance (cont.)
D4 ( p, q) x - s + y - t
2
2 1 2
2 1 0 1 2
2 1 2
D8 ( p, q) max( x - s , y - t )
2 2 2 2 2
2 1 1 1 2
2 1 0 1 2
2 1 1 1 2
2 2 2 2 2
111
Moire Pattern Effect : Special Case of Sampling