Communicating Quantitative Information
Download
Report
Transcript Communicating Quantitative Information
Communicating Quantitative
Information
Is a picture worth 1000 words?
Digital images. Number bases
Standards, Compression
Will [your] images last?
Homework: (Post project proposal). Work on
Project. Postings. Vote on Election Day.
Digital images
• Basic encoding
– Is a picture worth 1000 words?
• Standards
– bmp, gif, jpg
• Preservation issue
– will the files last on storage medium
– will there be appropriate software to display
and manipulate the image
Basic encoding
• Divide image into picture elements (pixels)
• For each pixel (cell in the grid), record
color
– Different color spaces:
• palette-based
• RGB (redness, greenness, blueness)
• YUV (Y=brightness, U and V together are hue and
saturation)
• other
Digital cameras marketing
• More pixels (increased resolution) means
that pictures can be 'blown up' more
without showing pixelations
• If you are NOT planning to make poster
size pictures, you don't need more than
3megalpel? 5 megalpel?
• cameras now competing on features like
modes
Palette based
• Think of painter's palette
• Can only use those colors
• If no match:
– painter mixes up something (creates a new color
on palette)
– computer software dithers
• produces spots of different colors that, hopefully, our
eyes mix together to see desired color.
The software constructs palette
when saving this image using a format with
limited palette
Image file
• Simple bit mapped, palette
– (may include the palette, with general
representation of the colors)
– string of numbers, one for each pixel,
indicating how that pixel is to be painted
Contrast
• Painter (in smock) starts with palette
• Computer software (PhotoShop, Paint
Shop Pro, etc. doing conversions) can
construct the best palette for a given photo
Encoding
• Image file is all numbers!!
• Actually, all bits (sequences of 1s and 0s)
• Suppose, palette is black and white
– 0 could stand for white
– 1 could stand for black
• Suppose, palette holds 4 colors
– 00, 01, 10, 11 represent 4 distinct colors
• Suppose, palette holds 256 colors
– 00000000, 00000001, …. 11111111
There are 256 distinct patterns of 8 1s and 0s.
Quick exercise
• How many different patterns can be made
using 0s and 1s, 3 bits long? Make a list
000
General formula
• Space N bits can hold 2N distinct patterns
• 1 bit can hold 2 distinct patterns (represent
2 colors) 21
• 2 bits can hold 4 distinct patterns
(represent 4 colors) 22
• 3 bits can hold 8 distinct patterns 23
• 8 bits can hold 28
Image [file] size
• Say image is 300 by 400 pixels (300 wide,
400 high)
• Each pixel is 8 bits (so the picture can
have 256 colors—not especially big)
• Say there is no compression (will get to
this later)
• Size is 300 * 400 * 8 bits
– 960000 bits
1000 words
• (Plain text file, not Word document)
• How big is a word? Assume it averages
out to 6 letters.
• How many bits does a letter require?
– In the standard encoding, 8 bits.
• 1000 words occupies
1000* 6 * 8 = 48000 bits
Is a picture worth 1000 words?
• Answer: it better because it costs/takes
considerably more space.
• compression reduces image sizes, but
only so much
What is a K?
• Kilobyte of data
– Byte is 8 bits of information: a piece of
memory holding 8 slots, each a 1 or a 0.
• Kilo
– Sometimes means 1000
– Sometimes means 1024 (2 raised to the 10th
power = 2*2*2*2*2*2*2*2*2*2
For
purple
hat
Compression
• 519 * 653 * 200 is 67781400 . This is
roughly 100 times either
• 677 * 1000 is 677000
• 677 * 1024 is 683008
• So… purple hat image is compressed
(encoded) to take advantage of similar line
sections of color. The gif format does this.
Digital camera
• Uses a (large) palette
• Records a number for what will be each
pixel
• There may be settings for color depth
(amount of bits ~ numbers of colors) and
resolution (size of pixel)
Standards (briefly)
• gif: uses common strings, so repeated patterns
do get compressed
– compression is lossless. Can restore full detail.
– restricted to 256 colors
– Good for line art
• jpg: uses common areas, but in a different color
space and a different way (approximates
changes in 8 by 8 blocks). For example, detects
common levels of brightness.
– compression is lossy. Cannot restore full detail.
– can hold millions of colors: 8 bits for each of YUV.
– Good for photographs
Number bases
• Refresher: we use the decimal system:
base 10
– 10 distinct symbols: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9
– Base system
• first column on the left is the ones/unit place (100)
• second column starting from the left is the 10s
place (101)
• third column is the 100th place (102)
• and so on…
Base 2
• Binary number system
– have 2 symbols: 0 and 1. These are called
bits!
– Base system
•
•
•
•
first column on the left is the ones/unit place (20)
second column is the 2-place (21)
third column is the 4-place (22)
and so on
• Fewer symbols (simpler circuitry), longer
strings of symbols needed
Warning
• bit is either a 0 or a 1. Abbreviation b
• byte is 8 bits. Abbreviation B
• k may mean 1000 or 1024 (a power of 2)
• Mega-bit, Mega-byte, Mega-pel
1000000 bits, bytes or pixels…
• Gigabyte: 1000000000
Base 16
• Hexadecimal
– 16 symbols: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C,
D, E, F.
– first column is the 1s place
– second column is the 16th place
– third column is the 256th place
– and so on.
Exercise
• Write your age in
– decimal
– binary
– hexadecimal
• How many 16s are there (probably just 1)?
• What is left over?
Use of Hexadecimal
• RGB colors
– Each of red, green,
blue is allocated
8 bits takes up 2
hexadecimal digits
(bad term)
– Deep red is FF0000
– Deep blue is 0000FF
– white is FFFFFF
– black is 000000
Will digital images last
Depends on
• storage medium. Hard disks can fail. CDs
and DVDs can get damaged. Memory
sticks fairly robust, but ends can get
damaged (and they can get lost)
• software: Need software that 'knows' the
standard used for the picture
– Software for display, manipulation, printing
Sound
• Sound files also come in different formats.
– wav (different parameters)
– mpeg
– ogg
• Browsers recognize different formats
• Same issue regarding archiving
– Storage needs to remain
– Software to manipulate and play needs to be
available
• New formats still under development.
HTML5 fragment
<audio autobuffer>
<source src="hithard.ogg" />
<source src="hithard.mp3" />
</audio>
Video
• Video (animation) is sequence of still
images shown in rapid succession.
• LARGE data files
• Compression possible per frame and from
frame to frame.
• Multiple compression schemes exist and
still others under development.
HTML5 fragment
<video id="vid" controls="controls" preload="auto">
<source src="sfire3.mp4" type='video/mp4;
codecs="avc1.42E01E, mp4a.40.2"'>
<source src="sfire3.theora.ogv" type='video/ogg;
codecs="theora, vorbis"'>
<source src="sfire3.webmvp8.webm"
type="video/webm; codec="vp8, vorbis"'">
Your browser does not accept the video tag.
</video>
Homework
• Project I
• Postings
– Posting topic: problems in archiving digital
images
– Old issue: major patent dispute concerning gif