Images

Last updated May 26, 2023 Edit

• Compsci
• Theory

Images consist of tiny pixels (picture elements). The higher the number of pixels, the higher the resolution of the image and the higher the storage space.

• Coordinate based, mathematical
• Does not lose resolution when scaled
• Does not store a binary value for each pixel, much more efficient than a bitmap
• Geometrical shapes such as lines and curves are used to represent an image
• A representation of mathematical data—not a direct mapping
• Consists of a drawing list in the file header including:
  • A command that describes the shape of the object
  • Attributes of each object (e.g. position)
  • Line colour
  • Line thickness
  • Fill colour
  • etc
• Scalable Vector Graphics (SVG) is an example of a vector image. It is an open standard.
• Vector graphics are widely used in animated movies, PDFs, etc.
• Vector images are more efficient with larger images, with very small images a bitmap may be more efficient.
• Example:
  • `Circle(centre = 0,0; radius= 5; fill = green; bordercolour = black; weight = 2px;``)
  • Rectangle(position= 3,10;width = 10;height= 5;fill = blue;bordercolour = None)
• An vector image cannot be printed, so must be converted to a bitmap before printing

• Raster Images or Pixel Maps
• Each pixel is stored on a grid—directly representing each element.
• Blurs when zoomed into
• JPG, PNG, GIF are examples of bitmap images
• Used by digital cameras and smartphones
• Each pixel colour is stored as a binary value
• More realistic than a vector image

A simple 2 colour image can be stored using 1 bit per pixel. So a 0 could be black and a 1 could be white.

1 bit => 2 possible colours (42K) 2 bits => 4 possible colours 4 bits => 16 possible colours 8 bits => 256 possible colours (420KB)

24 bits => 16 million possible colours (1.2MB)

As the number of bits increases, more colours can be used.

An image with colour depth n can represent 2^n colours.

Luminosity: Red: 24 bits Green: 24 bits Blue: 24 bits

Each channel has 24 bits, to allow very precise colours.

Screen resolution is horizontal pixels * vertical pixels.

• The pixel density of an image is measured in dots per inch or pixels per inch. It is the number of pixels or dots in a unit.
• Magazines and books have higher resolution compared to the images on computer screen.
• An image on a website is usually 72dpi. An image in a book has a resolution of 300 or even up to 600 dpi.

• Pixel density for a screen is calculated using the following steps.
• Calculating the pixel density of a Samsung Galaxy S10 phone which has a resolution of 1440 x 3040 pixels and a 6.1 inch display.
  • Add the squares of resolution sizes, 1440^2 + 3040^2 = 11315200
  • Take the square root of the result = 3363.81
  • Divide by the screen size, 3363.81 / 6.1 = 551ppi (pixels per square inch)

• Multiply the width and height of the image by the colour depth.
• Example:
  • 1010 x 562px
  • Colour Depth = 8
  • 1010 x 562 x 8 = 45409060 bits
  • 45409060 / 8 = 567620 bytes
  • 567620 / 8 = 0.568 megabytes

600 x 800 pixels 16 bit colour

600 x 800 = 567620 x 16 = 9081920 bits = 0.96 MB

• RLE is lossless compression
• We count the number of bits in a row that have the same value, and then store them as one entry rather than by storing each pixel individually.

“Data about data”

• Some examples of metadata for an image include:
  • Filename, format
  • Dimensions resolution, and colour depth of the image
  • Date and place the photo was taken
  • Time and date when the photo was changed
  • Camera settings when the photo was taken

Next: Sound

Theory