Image Representation.

advertisement

How Images are Represented

 Bitmap images (Dots used to draw the image)

 Monochrome images

8 bit grey scale images

24 bit colour

 Colour lookup tables

 Vector Graphics (Line art)

 Rasterisation

 Two key concepts

Image resolution

Colour resolution

Digital Images are represented by manipulating this…

And this…

.

Meaning…

Off

And…

.

On

Or rather ..

 0

 And zero

 1 One

 (In fact this is true for everything we see in a computer system!)

1 Bit Images

 Use a single bit to represent each pixel (dot on the screen)

00101

Use more bits and display as a grid…

 The following image is 61 pixels wide by 57 pixels high.

Quick Calculation of File Size

Image width

Image height

800 pixels

600 pixels

1 pixel = 1 bit, so image is 800 * 600 bits which is

480000 bits or divided by 8! (8 bits in a byte) gives us

60,000 bytes

Make the grid larger and pixels smaller…

Here we have increased the image resolution by using more dots per inch (dpi)

 This image is 512 pixels wide by 512 high.

 This type of image may be referred to as a binary image since each pixel is stored as a single zero or one.

 This kind of picture is also called a 1 bit monochrome image since it has no colour content.

8 Bit Grey Scale Images

 Representing shades of grey…

Now 1 pixel = 8 bits

256 Shades of Grey

File Size Calculations

 800 x 600 8 bit image = 480,000 bytes

800 x 600 = 480,000 pixels

1 pixel = 1 byte thus 480,000 bytes

8 x larger than 1 bit system

Colour

RGB Colours (Red Green Blue)

This time a single pixel is made up of 3 bytes or 24 bits.

File Size Calculations

 800 x 600 24bit colour = 1,440,000 bytes

800 x 600 = 480,000 pixels

1 pixel = 3 bytes thus 1,440,000 bytes

24 x larger than our first image

(24 bit colour formats – JPEG, BMP)

24 Bit Colour

 Files getting larger

 We can now represent 16 million colours

 The human eye can only see 10 million of them

 Extra colours useful for image processing and special effects

 Colour resolution is the number of colours a single pixel may display

8 bit Colour Images and Look up Tables

Draw three blue dots

Pixel Red Green Blue

(1 byte) 0 (1 byte) 255 (1 byte) 1

RGB(0,0,255)

2

RGB(0,0,255)

3

RGB(0,0,255)

4

RGB(0,0,255)

0

0

0

0

(1 byte) 0

(1 byte) 0

(1 byte) 255 (1 byte)

(1 byte) 255 (1 byte)

(1 byte) 0 (1 byte)

No of bytes = 12

255 (1 byte)

Pixel

1

2

3

4

Lookup table entry

1

1

1

1

(1 byte)

(1 byte)

(1 byte)

(1 byte)

Lookup Table

Table Entry

1

RGB(0,0,255)

Red

0 (1 byte)

Green

0 (1 byte)

Blue

255 (1 byte)

Reducing the number of bytes in this example from 12 to 7.

The GIF and PNG formats uses a look up table.

Pixel

1

2

3

4

5

Lookup table entry

1

1

1

1

2

(1 byte)

(1 byte)

(1 byte)

(1 byte)

(1 byte)

Lookup Table

Table Entry Red

1 0 (1 byte)

RGB(0,0,255)

2 255 (1 byte)

RGB(255,0,0)

Green

0 (1 byte)

0 (1 byte)

Blue

255 (1 byte)

0(1 byte)

Vector Graphics

800

800

800 x 800 1 bit image = 80,000 bytes

Do we need store every pixel?

Vector Graphics

 (at least) 4 bytes rather than 80,000!

Vector Graphics – Advantages

 They tend to generate small file sizes.

 They resize much better than bitmap graphics.

 Flash uses vector graphics.

Screen capture of a Vector Image…

Zoom in on the eye

Zoom in on a Bitmap image…

Rasterisation

The process of converting a vector shape into a bitmap is called rasterisation.

Photoshop text and shapes are drawn as vector graphics

To Conclude

 Image resolution = the number of pixels that make up the grid of the image per inch (dpi)

 Colour resolution = the maximum number of bits dedicated to a single pixel

 Bitmap graphics = a grid of pixels

 Vector graphics = calculated and drawn – only needs sufficient data to describe the image

 Rasterisation = converting a vector graphic to a bitmap

Download