Digital Image Processing
● Processing of images which are Digital in nature by digital Computers.
● An image is a spatial representation of a two-dimensional or three-dimensional scene.
A digital image is an image composed of picture elements, also known as pixels.
Image is an array, or a matrix pixels arranged in columns and rows.
● Pixels are small little dots that
make up the images on
computer displays.
Image Processing Motivation & Applications
● Improvement of pictorial information for human perception
● Image processing for autonomous machine applications
● Efficient storage and transmission.
Applications:
• Noise filtering (spotted image to clear image)
• Content enhancement (low contrast to enhanced contrast)
• De-blurring (defocused/motion-blurred to de-blurred)
• Medical Images (Tumor Detection, Cancer Detection, Ultra sonogram)
• Remote sensing (Ariel images from satellite; map, weather, star, galaxy)
Components of a Computer Vision System
Camera
Lighting
Computer
Scene
Scene Interpretation
Computer Vision
Extract the image information for further computer processing
Applications:
● Industrial machine vision for product assembly and inspection (Boiling Plant
Automation)
● Automated target detection (Fingerprint Recognition) and tracking (Number Plate
Detection)
● Machine processing of aerial and satellite imagery for weather prediction and crop
assessment.
● Boundary Information, Surface Information
Video Sequence Processing
To detect moving objects in image sequence for further processing
Applications:
● Detection and tracking of moving targets for security surveillance
● Finding the trajectory of moving targets
● Monitoring the movements of organ boundaries in medical applications
Compression
An image contains redundant information which can be exploited for achieving
compression
Types of redundancy:
● Pixel redundancy (Storage space reduction, Bandwidth reduction)
● Coding redundancy
● Psycho visual redundancy
Lossy compression
Remove redundancy as well as also remove some of the information so that after removing
those information, the quality of the reconstructed image is still acceptable.
There will be some loss or some distortion and this is taken care by what is called rate
distortion theorem.
Image Representation
At a particular point X Y in the image, conventionally the X coordinate is taken vertically
downwards and the Y axis is taken horizontally towards right.
If I look at this image, this image is nothing but a 2 dimensional intensity function which
is represented by f (x, y).
f (x, y) is represented by product of 2 terms;
- r (x, y) is the refractivity of the surface of the
corresponding image point from where the light
gets reflected.
- i (x, y) represents the intensity of the incident light.
Image Representation
m number of rows x n number of columns
Each of these elements in this matrix
representation is called a pixel or
a pale / picture elements.
Each pixel stores a value.
Pixel values are most often grey levels in
the range 0 - 255 (black - white)
Image Digitization
There are infinite number of points with infinite possible intensity values.
Take samples of the image on a regular grid. First level representation of an image in a
digital computer is spatial discretization by grids: called SAMPLING.
Once we get these sample values, at every point, the value of sample is again continuous.
So, it can assume any of the infinite possible values.
For that, the second operation that we have to do is discretization of the intensity values of
different samples: called QUANTIZATION.
Quantization
Values of sample
Each of this sample values are quantized and
typically for image processing applications,
The quantization is done using 8 bits for
black and white image
24 bits for color image, there are 3 color planes
- red, green and blue.
For each of the planes, if used 8 bits for quantization, then it gives us 24 bits which is
used for representation of digital color image.
Fundamental Steps in Digital Image
Processing
Relationship with other Fields
Key Stages in Digital Image
Processing
Image Morphological
Restoration P
rocessing
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain recognition
Colour Image Image
Compression
Processing
Key Stages in Digital Image
Processing: Image Aquisition
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Example: Take a picture Object
Problem Domain recognition
Colour Image Image
Processing Compression
Examples
18
Key Stages in Digital Image
Processing: Image Enhancement
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Example: Change contrast
Object
Problem Domain
recognition
Colour Image Image
Processing Compression
Examples
20
Key Stages in Digital Image
Processing: Image Restoration
Image Morphological
Restoration P
rocessing
Example: Remove
Image Noise
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Problem Domain Object
recognition
Colour Image Image
Compression
Processing
Example
22
Key Stages in Digital Image
Processing: Morphological
Processing
Extract
Image Morphological attributes
Restoration Processing useful for
describing
image
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain recognition
Colour Image Image
Compression
Processing
Example
24
Key Stages in Digital Image
Processing: Segmentation
Divide
Image Morphological image into
Restoration Processing constituent
parts
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain recognition
Colour Image Image
Compression
Processing
Example
26
Key Stages in Digital Image
Processing: Object Recognition
Image
regions
Image Morphological transformed
Restoration Processing suitable for
computer
processing
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain recognition
Colour Image Image
Compression
Processing
Example
28
Key Stages in Digital Image
Processing: Representation
&
Finds &
Description Image
Restoration
Morphological
Processing objects in
Labels
scene (e.g.
motorbike)
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain recognition
Colour Image Image
Compression
Processing
Example
30
Key Stages in Digital Image
Processing: Image
Compression
Reduce
Image Morphological
image size
Restoration Processing
(e.g. JPEG)
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain
recognition
Colour Image Image
Processing Compression
Example
32
Key Stages in Digital Image
Processing: Colour Image
Processing
Image Morphological
Restoration Processing
Image
Segmentation
Enhancement
Image Representation
Acquisition &
Description
Object
Problem Domain recognition
Consider color
Colour Image Image
images (color
Processing Compression
models, etc)
Example
34
Fundamental Steps in Digital Image
EssentialProcessing
steps when processing digital images:
Acquisition
Enhancement
Outputs are
Restoration
digital
images
Color image restoration
Wavelets
Morphological proces
Outputs are
Segmentation attributes of the
Representation image
38
Recognition
Mathematics for Image
Processing
Calculus
Linear algebra
Probability and statistics
Differential Equations (PDEs and ODEs)
Differential Geometry
Harmonic Analysis (Fourier, wavelet, etc)
Lecture No. 3
• Human Vision System
37
Human Vision System
The best vision model we have !
Knowledge of images form in the eye can help us with
digital image processing.
Lets take a tour of human visual system
Reflected Light
The colours humans perceive are determined by nature of light reflected from
an object
For example, if white light (contains all wavelengths) is shone onto green object
it absorbs most wavelengths absorbed except green wavelength (color)
Colours Absorbed
Structure Of Human Eye
The lens focusses light from objects onto the retina.
The retina is covered with light receptors called
cones (6-7 millions) and rods (75-150 millions)
Rods are used for distinguishing illumination of
color.
40
Image Formation In The Eye
Muscles in eye can change the shape of the lens allowing us focus on
near or far objects
An image is focused onto retina exciting the rods and cones and send signals
to the brain
Problem of Human Visual Image
System
42
Example_1
43
Human Eye Perception (Overlap between layers)
44
Brightness Adaption & Discrimination
45
Simultaneous Contrast
46
Optical Illusions
• Some tricky stuff by our vision
system
47
Light and The Electromagnatic
Spectrum
48
Light And The Electromagnetic Spectrum
Light: just a particular part of electromagnetic spectrum that can be
sensed by the human eye
The electromagnetic spectrum is split up according to the wavelengths of
different forms of energy
Electromagnetic Spectrum and IP
Images can be made from any form of EM radiation
Images from Different EM Radiation
Radar imaging (radio waves)
Magnetic Resonance Imaging (MRI) (Radio waves)
Microwave imaging
Infrared imaging
Photographs
Ultraviolet imaging telescopes
X-rays and Computed tomography
Positron emission tomography (gamma rays)
Ultrasound (not EM waves)
Example of Digital Images
a) Natural landscape
b) Synthetically generated scene
c) Poster graphic
d) Computer screenshot
e) Black and white illustration
f) Barcode
g) Fingerprint
h) X-
ray
i) Microscope slide
j) Satellite Image
k) Radar image
l) Astronomical object
Brightness Adaptation & Discrimination
The human visual system can perceive approximately 1010 different light
intensity levels
However, at any one time we can only discriminate between a much smaller
number - brightness adaptation
Similarly, perceived intensity of a region is related to the light intensities of the
regions surrounding it
Representing Images
Image data structure is 2D array of pixel values
Pixel values are gray levels in range 0-255 or RGB colors
Array values can be any data type (bit, byte, int, float,
double, etc.)
Image (Spatial) Sampling
A digital sensor can only measure a limited number of samples at a discrete
set of energy levels
Sampling can be thought of as: Continuous signal x
comb function
Image
Quantization
Quantization: process of converting continuous analog signal into its digital
representation
Discretize image I(u,v) values
Limit values image can take
Image Sampling And Quantization
Sampling and quantization generates approximation of a real
world scene
Image Sampling And Quantization
A digital sensor can only measure a limited number
of samples at a discrete set of energy levels
Quantisation is the process of converting a
continuous analogue signal into a digital
representation of this signal
Image Sampling And Quantization
Remember that a digital image is always only an
approximation of a real world scene.
Spatial Resolution
The spatial resolution of an image is determined by
how fine/coarse sampling was carried out
Spatial resolution: smallest discernable image detail
Vision specialists
talk about image resolution
Graphic designers
talk about dots per
inch (DPI)
Spatial Resolution
62 Vision specialists will often talk about pixel size
Spatial Resolution
1024 * 1024 512 * 512 256 * 256
128 * 128 64 * 64 32 * 32
63
Graphic designers will talk about dots per inch
Intensity Level
Resolution
Intensity level resolution: number of intensity levels
used to represent the image
The more intensity levels used, the finer the level of detail
discernable in an image
Intensity level resolution usually given in terms of number
of bits used to store each intensity level
Number of
Number of Bits Examples
Intensit
y
Levels
1 2 0, 1
2 4 00, 01, 10, 11
4 16 0000, 0101, 1111
8 256 00110011, 01010101
16 65,536 1010101010101010
Saturation & Noise
Saturation: highest intensity
value above which color is
washed out
Noise: grainy texture pattern
Resolution: How Much Is
Enough?
The big question with resolution is always
how much is enough?
Depends on what is in the image (details) and
what you would like to do with it (applications)
Key questions:
Does image look aesthetically pleasing?
Can you see what you need to see in image?
Resolution: How Much Is Enough?
Example: Picture on right okay for counting number
of cars, but not for reading the number plate
Intensity Level
Resolution
Low Detail Medium Detail High Detail
Image File Formats
Hundreds of image file formats. Examples
Tagged Image File Format (TIFF)
Graphics Interchange Format (GIF)
Portable Network Graphics (PNG)
JPEG, BMP, Portable Bitmap Format (PBM), etc
Image pixel values can be
Grayscale: 0 - 255 range
Binary: 0 or 1
Color: RGB colors in 0-255 range (or other color model)
Application specific (e.g. floating point values in astronomy)
How many Bits Per
Image
Element?
