IT101 Section 001

1
IT-101Section 001
Introduction to Information Technology

• Lecture 7

2

• Overview

• Chapter 5
• From the real world to Images and Video
• Introduction to visual representation and display
• Converting images to gray scale
• Color representation
• Video
• Image Compression

3

• Introduction to Visual Representation and Display

• Images play a fundamental role in the
  representation, storage and transmission of
  information
• In the previous chapters we learned how to
  represent information that was in the form of
  numbers and text
• In this chapter we will learn how to represent
  still and time varying images with binary digits
• A picture is worth ten thousand words
• But it takes a whole lot more than that!

4
Image Issues

• The world we live in is analog, or continuously
  varying
• There are problems involved in digitizing, or
  making discrete, so we make some approximations
  and determine tradeoffs involved
• While digitizing, we need to consider the
  following facts
• We are producing information for human use
• Human vision has limitations
• Take advantage of this fact
• Produce displays that are good enough

5
Digital Information for Humans

• Many digital systems take advantage of human
  limitations (visual, aural, etc)
• Human gray scale acuity is 2 of full brightness
• Or Most people can detect at most 50 gray levels
  (6 bits)
• The human eye can resolve about 60 lines per
  degree of visual arc- a measure of the ability
  of the eye to resolve fine detail
• When we look at a 8.5 x 11 sheet of paper at 1
  foot (landscape) the viewing angles are 49.25
  degrees for the horizontal dimension, and 39
  degrees for the vertical dimension
• We can therefore distinguish
• 49.25 degrees x 60 2955 horizontal lines
• 39 degrees x 60 2340 vertical lines
• These numbers give us a clue about the length of
  the code needed to capture images

6

• lines per degree of visual arc
• Image brought closer to the eye, we can resolve
  more detail
• Humans can resolve 60 lines per degree of visual
  arc
• A line requires two strings of pixels one
  black, one white
• Pixel The smallest unit of representation for
  visual information

Visual Arc
7
Pixels

• A pixel is the smallest unit of representation
  for visual information
• Each pixel in a digitized image represents one
  intensity (brightness) level (gray scale or
  color)

13 x 13 grid 169 pixels
Gray scale
Color
8

• To form a black line, you need two rows of pixels
  (one black and one white) to give a visual clue
  of the transition from black to white
• For our paper example
• Number of pixels needed to represent total image
  on a page (2 x 2955) x (2 x 2340) 27,658,800
  pixels per page
• This number of pixels would be sufficient to
  represent any image on the page with no visible
  degradation compared to a perfect (unpixelized)
  image at a distance of one foot
• As the number of pixels that form an image
  (spatial resolution) decreases, the amount of
  data that needs to be transmitted, stored or
  processed decreases as well. However, the
  tradeoff is that the quality of the image
  degrades as a result

9
A note about printer resolution

• When dealing with printers we often quote the
  resolution in terms of dots per inch (dpi), which
  corresponds to pixels per inch in our example
• It is popular to set laser or ink-jet printer
  settings to 600 dpi of resolution
• However, if we hold the paper closer, we would
  need a greater resolution printer for example
  720 dpi, 1200 dpi or greater

10
How many pixels should be used

• If too few pixels used, image appears coarse

16 x16 (256 pixels)

64 x 64 (4096 pixels)
11
Digitizing Images (gray scale)

• The first step to digitize a black and white
  image composed of an array of gray shades, is to
  divide the image into a number of pixels,
  depending on the required spatial resolution
• The number of brightness levels to be represented
  by each pixel is assigned next
• If we wish to use for example, 6 bits for the
  brightness level of each pixel, then each pixel
  can represent 64 different brightness levels
  (shades of gray, from black to white)
• Then, each pixel would have a 6-bit number
  associated with it, representing the brightness
  level (shade) that is closest to the actual
  brightness level at that pixel

12

• This process is known as quantization (we will
  learn more about this later in the course)-It is
  the process of rounding off actual continuous
  values so that they can be represented by a fixed
  number of binary digits
• As a result of the operations just described, the
  analog image is digitized and represented by a
  string of binary digits

1010010101010101010
13
6-bit image (64 gray levels)

• In the figures below, each pixel in the image is
  represented by 6 bits, 3 bits and 1 bit. The
  effect of varying the number of bits used to
  represent each pixel is evident

3-bit image (8 gray levels)
14
1-bit image (black and white)
15
How much storage is needed?

• Total number of bits required for storage total
  number of pixels number of bits used per pixel
• For example Black and white photo
• 64 x 64 pixels
• Use 32 gray levels (5 bits)
• 64 x 64 x 5 20,480 bits 2560/1024 bytes
  2.5KB
• Remember data storage is in bytes
• KB represents 210 or 1024 bytes

16
Another example

• Black and White photo
• 256 x 256 pixel
• 6 bits (64 gray levels)
• How much storage is needed?
• 256 x 256 x 6 393,216 bits
• 393,216/8 49,152 bytes
• 49,152/1024 48 KB

17
A note about resolution

• Since the total number of bits required for
  storage total number of pixels number of bits
  used per pixel, there are two ways to reduce the
  number of bits needed to represent an image
• Reduce total number of pixels
• Reduce number of bits used per pixel
• Applying these however, reduces the quality of
  the image. The first results in low spatial
  resolution (image appears coarse). The second
  results in poor brightness resolution, as seen by
  the previous couple of slides.
• The amount of storage can, however be reduced by
  applying Image Compression

18

• Digitizing Images (color)

• Recall that any color can be created by adding
  the right proportions of red, green and blue
  light
• If we wish to digitize a color image, we must
  first divide the image into pixels
• We must then determine the amount of red, green
  and blue (RGB) that comprises the color at each
  pixel location
• Finally, we must convert these three levels to a
  binary number of a predefined length
• For example
• If we use 3 bits for each color value, we would
  be able to represent 8 intensity levels each of
  red, green and blue
• This representation would require 9 bits per
  pixel
• This would give us 512 different colors per pixel

19
Example

• Color photo
• 256x256 pixel
• 9 bits per pixel (3 bits each for red, green and
  blue)
• 256x256x9589,826 bits
• 589,826/873,728 bytes
• 73,728/102472 KB of storage is needed to store
  this color photo

20
Hue Luminance Saturation

• Another approach to color representation of
  images is Hue, Luminance and Saturation (HLS)
• This system does not represent colors by
  combinations of other colors, but it still uses 3
  numerical values
• Hue Represents where the pure color component
  falls on a scale that extends across the visible
  light spectrum
• Luminance Represents how light or dark a pixel
  is
• Saturation How pure the color is, i.e. how
  much it is diluted by the addition of white (100
  saturation means no dilution with white )
• Let us see at how this system works with the
  power point color palette on this box

21

• Video

• Human perception of movement is slow
• Studies show that humans can only take in 20
  different images per second before they begin to
  blur together
• If these images are sufficiently similar, then
  the blurring which takes place appears to the eye
  to resemble motion, in the same way we discern it
  when an object moves smoothly in the real world.
• We can detect higher rates of flicker, but only
  to about 50 per second
• This phenomenon has been used since the beginning
  of the 20th century to produce moving
  pictures,'' or movies.
• Movies show 24 frames per second
• TV works similarly, but instead of a frame, TV
  refreshes in lines across the tube
• This same phenomenon can be used to create
  digitized video--a video signal stored in binary
  form

22
Video

• We have already discussed how individual images
  are digitized digital video simply consists of a
  sequence of digitized still images, displayed at
  a rate sufficiently high to appear as continuous
  motion to the human visual system. The individual
  images are obtained by a digital camera that
  acquires a new image at a fast enough rate
  (say,60 times per second), to create a
  time-sampled version of the scene in motion
• Because of human visual latency, these samples at
  certain instants in time are sufficient to
  capture all of the information that we are
  capable of taking in!

23
Adding up the bits

• Assume a screen that is 512x512 pixels--about
  the same resolution as a good TV set.
• Assume 3 bits per color per pixel, for a total of
  9 bits per pixel
• Let's say we want the scene to change 60 times
  per second, so that we don't see any flicker or
  choppiness. This means we will need 512 x 512
  pixels x 9 bits per pixel x 60 frames per second
  x 3600 seconds 500 billion bits per hour--just
  for the video. Francis Ford Coppola's The
  Godfather, at over 3 hours, would require nearly
  191 GB--over 191 billion bytes--of memory using
  this approach. This almost sounds like an offer
  we can refuse. But, do films actually require
  this much storage? Fortunately, no..
• The reason we can represent video with
  significantly fewer bits than in this example is
  due to compression techniques, which take
  advantage of certain predict abilities and
  redundancies in video information to reduce the
  amount of information to be stored.

24
Image Compression

• Near Photographic Quality Image
• 1,280 Rows of 800 pixels each, with 24 bits of
  color information per pixel
• Total 24,576,000 bits
• 56 Kbps modem
• 56,000 bits/sec
• How long does it take to download?
  24,576,000/56,000 439 seconds/60 7.31
  minutes

Obviously image compression is essential.
25

• Images are well suited for compression
• Images have more redundancy than other types of
  data.
• Images contain a large amount of structure.
• Human eye is very tolerant of approximation
  error.
• 2 types of image compression
• Lossless coding
• Every detail of original data is restored upon
  decoding
• Examples Run Length Encoding, JPEG, GIF
• Lossy coding
• Portion of original data is lost but undetectable
  to human eye
• Good for images and audio
• Examples - JPEG

26
The two compressed image formats most often
encountered on the Web are JPEG and GIF.

• JPEG -Joint Photographic Experts Group
• 29 distinct coding systems for compression, 2 for
  Lossless compression
• Lossless JPEG uses a technique called predictive
  coding to attempt to identify pixels later in the
  image in terms of previous pixels in that same
  image
• Lossy JPEG consists of image simplification,
  removing image complexity at some loss of
  fidelity
• GIF Graphics Interchange Format
• Developed by CompuServe
• Lossless image compression system.
• Application of Lempel-Ziv-Welch (LZW)

27
Digital Video Compression (MPEG)
Motion Picture Expert Group (MPEG) standard for
video compression.

• MPEG is a series of techniques for compressing
  streaming digital information
• DVDs use MPEG coding
• MPEG achieves compression results on the order of
  1/35 of original
• If we examine two still images from a video
  sequence of images, we will almost always find
  that they are similar
• This fact can be exploited by transmitting only
  the changes from one image to the next
• Many pixels will not change from one image to the
  next.

Called IMAGE DIFFERENCE CODING