Lecture 2 Image formation ◼ There are two parts to the image formation process:
◼ The geometry of image formation, which
determines where in the image plane the projection of a point in the scene will be located.
◼ The physics of light, which determines the
brightness of a point in the image plane as a function of illumination and surface properties.
March 23, 2022 Computer Science and Engineering, SUST 2
A Simple model of image formation
◼ The scene is illuminated by a single source.
◼ The scene reflects radiation towards the camera. ◼ The camera senses it via chemicals on film.
March 23, 2022 Computer Science and Engineering, SUST 3
Pinhole Camera ◼ This is the simplest device to form an image of a 3D scene on a 2D surface. ◼ Straight rays of light pass through a “pinhole” and form an inverted image of the object on the image plane.
March 23, 2022 Computer Science and Engineering, SUST 4
CCD (Charged-Coupled Device) cameras ◼ Tiny solid state cells convert light energy into electrical charge. ◼ The image plane acts as a digital memory that can be read row by row by a computer.
March 23, 2022 Computer Science and Engineering, SUST 5
Frame Grabber ◼ Usually, a CCD camera plugs into a computer board (frame grabber). ◼ The frame grabber digitizes the signal and stores it in its memory (frame buffer).
March 23, 2022 Computer Science and Engineering, SUST 6
Image Representation ◼ An image is a function of space ◼ Typically, a 2-D projection of the 3-D space is used, but the image can exist in the 3-D space directly
March 23, 2022 Computer Science and Engineering, SUST 7
Image Representation ◼ Image as a single-valued function, for quantifying the intensity. ◼ Image can be multi-valued function. The multiple values may correspond to different color intensities, for example, R, G, B makes color image.
◼ Notice that we defined images as functions in a
continuous domain. ◼ Images are representations of an analog world. Hence, as with all digital signal processing, we need to digitize our images
March 23, 2022 Computer Science and Engineering, SUST 8
Image Digitization
◼ Digitization of an analog signal involves two operations:
◼ Sampling ◼ Quantization ◼ Both operations correspond to a discretization of a quantity, but in different domains. Sampling ◼ An image may be continuous with respect to the x- and y-coordinates and also in amplitude.
◼ Sampling corresponds to a discretization of the space.
That is, of the domain of the function.
◼ Digitizing the coordinate values is called Sampling.
◼ Digitizing the amplitude values is called Quantization.
March 23, 2022 Computer Science and Engineering, SUST 10
Sampling (Cont’d)
March 23, 2022 Computer Science and Engineering, SUST 11
Sampling (Cont’d) ◼ Image can be seen as a matrix f(0,0) f(0,1) ... f(0, N − 1) f(1,0) f(1,1) ... f(1, N − 1) . . . . . . . . . . . . f(M − 1,0) f(M − 1,1) ... f(M − 1, N − 1) ◼ The smallest element resulting from the discretization of the space is called a pixel (picture element).
◼ For 3-D images, this element is called a voxel
(volumetric pixel)
March 23, 2022 Computer Science and Engineering, SUST 12
Image quantization (example)
March 23, 2022 Computer Science and Engineering, SUST 13
Summary: Sampling and Quantization
March 23, 2022 Computer Science and Engineering, SUST 14
Sampling and Quantization
pixel grid
real image sampled quantized sampled &
quantized
March 23, 2022 Computer Science and Engineering, SUST 15
Digital Image Color images have 3 values per pixel; monochrome images have 1 value per pixel.
a grid of squares, each of which contains a single color
each square is called a pixel (for picture element)
March 23, 2022 Computer Science and Engineering, SUST 16
Pixels
◼ 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).
March 23, 2022 Computer Science and Engineering, SUST 17
Pixels
◼ 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).
March 23, 2022 Computer Science and Engineering, SUST 18
Pixels
Pixel Location: p = (r , c) Pixel Value: I(p) = I(r , c) Pixel : [ p, I(p)]
March 23, 2022 Computer Science and Engineering, SUST 19
A Simple Image Formation Model ◼ An image is a 2-D light intensity function f(x,y) ◼ +ve scalar quantity in spatial domain ◼ Physical meaning is determined by the source of the image
◼ An image is formed through the mapping of 3D scene
onto 2D surface or plane.
◼ The light intensity distribution f(x,y), over that plane
characterizes the image and is considered to be dependent on 2 distributions- illumination and reflectance. ◼ The image is modeled as Illumination-Reflectance model
March 23, 2022 Computer Science and Engineering, SUST 20
The Illumination- Reflectance Model ◼ Illumination- i(x,y): The amount of source illumination incident on the scene being viewed or amount of illumination received by the various objects ◼ Typical values: 9000 foot-candles sunny day, 100 office room, 0.01 moonlight
◼ Reflectivity- r(x,y): Amount of illumination reflected
by the objects in the scene ◼ r(x,y) is the reflectance: 0 < r(x,y) < 1 ◼ r(x,y)=0 implies total absorption ◼ r(x,y)=1 implies total reflectance ◼ Typical values: 0.01 black velvet, 0.80 flat white paint, 0.93 snow
March 23, 2022 Computer Science and Engineering, SUST 21
The Illumination- Reflectance Model ◼ f(x,y) and i(x,y) are +ve for all values of x and y, i.e., ◼ 0 ≤f(x,y)< ∞ ◼ 0 ≤i(x,y)< ∞
◼ r(x,y) characterizes the various objects in the scene
and can be expressed as the product of 2 components
March 23, 2022 Computer Science and Engineering, SUST 22
Simple Imaging Model ◼ The intensity of a monochrome image f at (x,y) is the gray level (l) of the image at that point. Lmin≤ l ≤ Lmax
◼ In practice Lmin=iminrmin and Lmax=imaxrmax
◼ As a guideline Lmin ≈ 10 and Lmax ≈ 1000 for indoor image processing applications
◼ The interval [Lmin, Lmax] is called the gray scale
◼ Common practice is to shift the interval to [0, L-1] where l=0 is
considered black and l=L-1 is considered white. All intermediate values are shades of gray
March 23, 2022 Computer Science and Engineering, SUST 23
Sampling and Quantization
pixel grid
real image sampled quantized sampled &
quantized
March 23, 2022 Computer Science and Engineering, SUST 24
Sampling and Quantization
March 23, 2022 Computer Science and Engineering, SUST 25
Sampling and Quantization
March 23, 2022 Computer Science and Engineering, SUST 26
Representing Digital Image
◼ Result of sampling and quantization
◼ Matrix of real numbers ◼ M x N digital image f(0,0) f(0,1) ... f(0, N − 1) f(1,0) f(1,1) ... f(1, N − 1) . . . . . . . . . . . . f(M − 1,0) f(M − 1,1) ... f(M − 1, N − 1)
◼ What’s the size of M and N and L- gray levels?
March 23, 2022 Computer Science and Engineering, SUST 27
Representing Digital Images ◼ No requirements of M and N other than to be +ve integers ◼ Due to processing, storage and sampling hardware considerations, the number of gray levels L is typically power of 2 ◼ L=2k ◼ [0,L-1] – dynamic range of an image ◼ Image whose gray levels span a significant portion of the gray scale- high dynamic range- high contrast ◼ Low dynamic range- dull washed out gray look
◼ The number of bits, b, necessary to store the image is then
◼ b=M x N x k, when M=N it becomes b=N2k
◼ For example, a 128x128 image with 64 gray levels would require
98,304 bits of storage: b=128x128x6
March 23, 2022 Computer Science and Engineering, SUST 28
Storage Bits Requirements
◼ How many samples and gray levels are required for a “good” approximation?
◼ The resolution (the degree of discernible detail) depends strongly on
these two parameters
March 23, 2022 Computer Science and Engineering, SUST 29
Image Resolution ◼ Digital image implies the discretization of both spatial and intensity values. The notion of resolution is valid in either domain.
◼ Most often it refers to the resolution in sampling-
Spatial resolution.
◼ It also can refer to the number of quantization levels.
March 23, 2022 Computer Science and Engineering, SUST 30
Resolution: How Much Is Enough?
◼The big question with resolution is always how much is
enough? ◼ This all depends on what is in the image and what you would like to do with it ◼ Key questions include ◼ Does the image look aesthetically pleasing? ◼ Can you see what you need to see within the image?
March 23, 2022 Computer Science and Engineering, SUST 31
Resolution: How Much Is Enough?(cont…)
◼The picture on the right is fine for counting the
number of cars, but not for reading the number plate
March 23, 2022 Computer Science and Engineering, SUST 32
Spatial Resolution ◼ Sampling is the principal factor determining the Spatial Resolution of an image ◼ Smallest discernible detail in an image
◼ Example: ◼ Construct a chart with vertical lines of width W, with the space between lines also having width W ◼ One line and its adjacent space- Line pair, width 2W, therefore 1/2W line pairs per unit distance
◼ Resolution: Smallest number of discernible line pairs
per unit distance. Usually 100 line pairs per mm
March 23, 2022 Computer Science and Engineering, SUST 33
Spatial Resolution 1250dpi 300dpi ◼ Dot (pixels) per inch- DPI
March 23, 2022 Computer Science and Engineering, SUST 34
Gray-Level/ Intensity Resolution ◼ Smallest discernible change in gray level- highly subjective process ◼ Discretion regarding the number of samples used to generate a digital image, but this is not true for the number of gray levels. ◼ Due to hardware considerations, # of gray levels usually an integer power of 2. ◼ Most common: 8 bits, 16 bits used in some applications ◼ Terminology: L-level digital image of size M x N ◼ Spatial resolution: M x N ◼ Gray level resolution: L
March 23, 2022 Computer Science and Engineering, SUST 35
Effects of Reducing Spatial Resolution
March 23, 2022 Computer Science and Engineering, SUST 37
Effects of Reducing Gray Level
March 23, 2022 Computer Science and Engineering, SUST 38
Effects of Reducing Gray Level
March 23, 2022 Computer Science and Engineering, SUST 39
Neighbors of a Pixel ◼ An image is denoted by: f(x,y) ◼ Lowercase letters (e.g. p, q) will denote individual pixels ◼ A subset of f(x,y) is denoted by S
◼ Neighbors of a pixel: ◼ A pixel p at (x,y) has 4 horizontal/vertical neighbors at ◼ (x+1,y), (x-1,y), (x, y+1) and (x, y-1) ◼ called the 4-neighbors of p: N4(p) ◼ A pixel p at (x,y) has 4 diagonal neighbors at ◼ (x+1,y+1), (x+1,y-1), (x-1, y+1) and (x-1, y-1) ◼ called the diagonal-neighbors of p: ND(p) ◼ The 4-neighbors and the diagonal-neighbors of p are called the 8-neighbors of p: N8(p)
March 23, 2022 Computer Science and Engineering, SUST 40
Example
4-Neighbors
March 23, 2022 Computer Science and Engineering, SUST 41
Connectivity Between Pixels ◼ Connectivity is an important concept in establishing boundaries of object and components of regions in an image
◼ When are two pixels connected?
◼ If they are adjacent in some sense (say they are 4-neighbors) ◼ and, if their gray levels satisfy a specified criterion of similarity (say they are equal)
◼ Example: given a binary image (e.g. gray scale = [0,1]), two
pixels may be 4-neighbors but are not considered connected unless they have the same value
0 1 1 1
March 23, 2022 Computer Science and Engineering, SUST 42
