Computer Vision and Image
processing (CoSc4151)
2. Digital Image Fundamentals
Jimma University
Jimma Institute of Technology
For Computer Science students
Abel W.
Concept of image
What is an image?
Signal = function (variable with physical meaning)
I. one-dimensional (e.g. dependent on time)
II. two-dimensional (e.g. dependent on two co-ordinates in a
plane)
III. three-dimensional (e.g. describing an object in space)
In general an image is a signal and as with all signals they
are functions of physical things.
CV & IP 02 18-Nov-16 2
Concept of image
Image function value = brightness at image
points
Brightness is dependent on several factors
1. object surface reflectance properties
surface material, microstructure and marking
2. Illumination properties
provide their own light as in glow in the dark stickers
3. object surface orientation with respect to a viewer and
light source
temperature, distance from the observer
CV & IP 02 18-Nov-16 3
Concept of image
• Image capturing
CV & IP 02 18-Nov-16 4
Concept of image
Image intensity, f(x, y), is proportional to EM
energy radiated/received 0 f ( x, y )
Required f(x, y) is nonzero and finite
f ( x , y ) i ( x, y ) r ( x, y )
where 0 i( x, y ) , incident illumination
0 r(x, y) 1, reflected illumination
Typical values: 90,000 lm/m 2 , clear sunny day
10,000 lm/m 2 , cloudy day
i(x,y) 2
0.01 lm/m , full moon, clear night
2
1,000 lm/m , commercial office
0.01, black material
r(x, y) 0.8, flat white wall paint
CV & IP 02
0.93, snow
18-Nov-16 5
Concept of image
Thus, the intensity image lies in the
range
i r f ( x, y ) i r
min min max max
0 f ( x, y ) L 1
CV & IP 02 18-Nov-16 6
Digital image representation
A digital image can be Where
represented as a two- f(x, y) =monochrome
dimensional (2D) matrix images of size M×N
of real numbers. X = row number (from 0
to M−1)
Y = column number
(between 0 and N−1)
CV & IP 02 18-Nov-16 7
Digital image representation
The value of the two-dimensional function f(x, y) at
any given pixel of coordinates (x0,y0), denoted by
f(x0,y0), is called the intensity or gray level of the
image at that pixel.
Maximum and Minimum values that a pixel intensity
can vary based on
data type
CV & IP 02 18-Nov-16 8
Digital image representation
1. Data Type
Logical: 1.0 for white and 0.0 for Black
uint8( unsigned integer, 8 bits): 0(Black) and 255(White)
2. Convection
f( p, q)to refer to the MATLAB representation of f(x, y) ( where p
denotes row and q denotes column )
CV & IP 02 18-Nov-16 9
Concept of pixel
A digital image, I, is a mapping from a 2D grid of
uniformly spaced discrete points, {p = (r , c)}, into
a set of positive integer values, {I( p)}, or a set of
vector values, e.g., {[R G B]T( p)}.
At each column location in each row of I there is a
value.
The pair ( p, I( p) ) is called a “pixel” (for picture
element).
CV & IP 02 18-Nov-16 10
Concept of pixel
p = ( r, c) is the pixel location indexed by row, r,
and column, c.
I( p) = I( r, c) is the value of the pixel at location p.
If I( p) is a single number then I is monochrome.
If I( p) is a vector (ordered list of numbers) then I
has multiple bands (e.g., a color image).
CV & IP 02 18-Nov-16 11
Concept of pixel
Pixel : [ p, I(p)] Pixel Location: p = (r , c)
Pixel Value: I(p) = I(r , c)
p = (r , c)
= (row # , col # )
=CV(&272,
IP 02
277) 18-Nov-16 12
Digital image acquisition process
CV & IP 02 18-Nov-16 13
Image sampling and quantization
An image may be continuous with respect to the x- and
y-coordinates, and also in amplitude.
When x, y, and the amplitude values are all finite,
discrete quantities, we call the image a digital image.
To convert it to digital form, we have to sample the
function in both coordinates and in amplitude.
1. Digitizing the coordinate values is called sampling.
2. Digitizing the amplitude values is called quantization.
CV & IP 02 18-Nov-16 14
Image sampling
Continuous image
A scan line from A to B in
the continuous image
To sample this take
equally spaced samples
along
CV & IP 02line AB. 18-Nov-16 15
Image sampling
The samples are shown as
small white squares
superimposed on the
function.
The set of these discrete
locations gives the
sampled function
CV & IP 02 18-Nov-16 16
Image quantization
In order to form a
digital function, the
gray-level values
also must be
converted
(quantized) into
discrete quantities.
CV & IP 02 18-Nov-16 17
Concepts of sampling and quantization
CV & IP 02 18-Nov-16 18
Types of Digital Image
Binary (1-Bit) Images
Gray-Level (8-Bit) Images
Color Images
CV & IP 02 18-Nov-16 19
Types of Digital Image
1. Binary (1-Bit) Images: encoded as a 2D array, typically
using 1 bit per pixel, where a 0 usually means “black” and a 1
means “white”
CV & IP 02 18-Nov-16 20
Types of Digital Image
2. Gray-Level (8-Bit) Images: (also referred to as
monochrome) encoded as a 2D array of pixels, usually with 8
bits per pixel, where a pixel value of 0 corresponds to “
black,” a pixel value of 255 means “white,” and intermediate
values indicate varying shades of gray
CV & IP 02 18-Nov-16 21
Gray-level images
Continuous image function f(x,y)
arguments - two co-ordinates (x,y) can have any value
Digital image functions - represented by matrices
co-ordinates = integer numbers
Cartesian (horizontal x axis, vertical y axis)
Monochromatic image function range
lowest value - black
highest value - white
Limited brightness values = grey levels
CV & IP 02 18-Nov-16 22
Gray level images
CV & IP 02 18-Nov-16 23
Representation of gray level images
a digital 8-bit, 8x8 images is shown as an array of numbers on the
left and brightness levels on the right.
CV & IP 02 18-Nov-16 24
Gray level images
• The number of distinct gray levels is usually a power
of 2 that is, 𝑳 = 𝟐𝑩
where B is the number of bits in the binary representation of
the brightness levels.
• When B>1 we speak of a gray-level image; when B=1
we speak of a binary image.
• In a binary image there are just two gray levels which
can be referred as “black” and “white” or “0” and “1”.
CV & IP 02 18-Nov-16 25
Types of Digital Image
3. Color Images: Representation of color images is more
complex and varied.
The two most common ways of storing color image
contents are
RGB representation in which each pixel is usually represented by a
24-bit number containing the amount of its red (R), green (G),
and blue (B) components
indexed representation—where a 2D array contains indices to a
color palette (or lookup table-(LUT)).
CV & IP 02 18-Nov-16 26
Cont.. Color Image
24-Bit (RGB) Color images, represented using three
2D arrays of same size, one for each color channel:
red (R), green (G), and blue (B)
Each array element contains an 8-bit value, indicating
the amount of red, green, or blue at that point in a
[0, 255] scale. The combination of the three 8-bit
values into a 24-bit number allows 2 raised to
24 (16,777,216, usually referred to as 16 million or 16
M) color combinations.
CV & IP 02 18-Nov-16 27
Cont’d… color images
Represented by vector not scalar
• Red, Green, Blue (RGB)
• Hue, Saturation, Value (HSV)
• Cyan, Magenta, Yellow (CMY)
Solid color cube
((28)3 = 16,777,216 colors)
CV & IP 02 18-Nov-16 28
Cont.. Color Image
• 24-Bit (RGB) Color
Original Image
b. Red Plane
c. Green Plane
d. Blue Plane
CV & IP 02 18-Nov-16 29
Properties that define image format
• Pixel (or spatial) resolution (e.g. 640x480 pixels)
• Pixel bit-depth (8-bit)
• Number of planes (1 for gray scale images, 3 for color)
• Colorspace (RGB, CMYK etc.)
CV & IP 02 18-Nov-16 30
Binary and Gray scale images
CV & IP 02 18-Nov-16 31
Intensity Resolution
Smallest discernible change in intensity level
8-bit 4-bit 3-bit
7-bit
6-bit CV & IP 02 5-bit 2-bit 18-Nov-16
1-bit
32
Image of size
CV & IP 02 18-Nov-16 33
Image of size
fewer Pixels Mean Lower Spatial Resolution
CV & IP 02 18-Nov-16 34
Image interpolation
Introduction
• An image f(x,y) tells us the intensity values at the
integral lattice locations, i.e., when x and y are both
integers
• Image interpolation refers to the “guess” of intensity values
at missing locations, i.e., x and y can be arbitrary
• Note that it is just a guess (Note that all sensors have finite
sampling distance)
• image interpolation is about D-A conversion
CV & IP 02 18-Nov-16 35
Image interpolation:
Image re-sampling or re-sizing operations:-
zooming, shrinking, rotating, and geometric corrections
Uses known image data to estimate at the new locations.
Different methods:
1. Nearest neighbor interpolation:-
2. Linear interpolation
3. Bilinear interpolation
4. Cubic interpolation and others
CV & IP 02 18-Nov-16 36
Image Interpolation
Example:
f(1) = 1, f(2) = 10, f(3) = 5 , f(4) = 16, f(5) = 20
CV & IP 02 18-Nov-16 37
Image Interpolation
How can we find f(1.5)?
One approach: take the average of f(1) and f(2)
f (1.5) = 5.5
CV & IP 02 18-Nov-16 38
Linear interpolation
Fit a line between each pair of data points
CV & IP 02 18-Nov-16 39
Linear interpolation
• What is f(1.8)?
Answer: 0.2 f(1) + 0.8 f(2)
f (1.8) = 0.2 (1) + 0.8 (10) = 8.2
CV & IP 02 18-Nov-16 40
Linear interpolation
To compute f(x), find the two points xleft and xright that x lies
between
f (x) = (xright - x) / (xleft – xright) f (xleft) +
(x - xleft) / (xleft – xright) f (xright)
f (xleft)
f (xright)
xleft x xright
CV & IP 02 18-Nov-16 41
Bilinear interpolation
What about in 2D?
Interpolate in x, then in y
Example
We know the red values
Linear interpolation in x between red
values gives us the blue values
Linear interpolation in y between the
blue values gives us the answer
http://en.wikipedia.org/wiki/
Bilinear_interpolation
CV & IP 02 42
18-Nov-16
Bilinear interpolation
Bilinear interpolation: - uses the
four nearest neighbors to estimate the
intensity at a the new location
http://en.wikipedia.org/wiki/
Bilinear_interpolation
CV & IP 02 43
18-Nov-16
Cont’d…
Nearest neighbor interpolation
Direct zooming
Bilinear interpolation
CV & IP 02 18-Nov-16 44
Extrapolation
Suppose you only know the values f(1), f(2), f(3), f(4)
of a function
What is f(5)?
This problem is called extrapolation
Much harder than interpolation: what is outside the image?
For the particular case of temporal data, extrapolation is
called prediction (what will the value of MSFT stock be
tomorrow?)
If you have a good model, this can work well
CV & IP 02 18-Nov-16 45
Image extrapolation
http://graphics.cs.cmu.edu/projects/scene-completion/
Computed using a database of millions of photos
CV & IP 02 18-Nov-16 46
Mathematical Tools used in Digital Image
Processing
1. Array versus Matrix Operations:
Digital image viewed as an array
Array operation is pixel-by-pixel
Consider image arrays A and B, where
a11 a12 b11 b12
A , B
21 22
a a 21 22
b b
then a11 a12
a 2
a 2
A b11 b12
A 2
2 11 12
2
, C
a21 a22 B a21 a22
b21 b22
CV & IP 02 18-Nov-16 47
Mathematical Tools used in Digital Image
Processing cont’d…
2. Linear versus Nonlinear Operations:
most image processing operations are linear.
Consider an input image f(x, y), an operator H
and output g(x, y), i.e.,
H[f(x, y)] = g(x, y)
H is a linear operator if the following is true.
H [ai f i ( x, y ) a j f j ( x, y )] ai g i ( x, y ) a j f j ( x, )
Mathematical Tools used in Digital Image
Processing cont’d…
3. Arithmetic Operations:
These are operations between corresponding pixels
Same size images, M x N, and the basic arithmetic
operations are
s ( x , y ) f ( x , y ) g ( x, y )
d ( x , y ) f ( x, y ) g ( x , y )
p ( x, y ) f ( x, y ) g ( x, y )
v ( x, y ) f ( x, y ) g ( x, y )
Thank you!!!
CV & IP 02 18-Nov-16 50
