Notice that the pixels on their own, probably noise, are smoothed out.
One problem with this technique is that it tends to blur sharp edges so the picture looks less sharp.
Image Processing
When a digital signal is transmitted it may pick up noise which, even after filtering, may distort it. It would change the occasional 1 to a 0 and vice versa. This usually results in a picture that looks "grainy".
There are ways of minimising the effects of noise. This is one;
|
|
Replace each pixel with one that is the average colour of
its 4 (or 8) nearest neighbours.
Notice that the pixels on their own, probably noise, are smoothed out. One problem with this technique is that it tends to blur sharp edges so the picture looks less sharp. |
Before
smoothing 
After
smoothing
Compression
 |
Large parts of this image are exactly the same colour. This
means that the pixels have exactly the same value. Why transmit exactly
the same 8 bits several thousand times, why not just transmit it once and
say how many times to repeat it.
This is a good way of reducing file sizes. One type of image that does this is a GIF (Graphics Interchange Format). |
Vector Graphics
|
|
Another great way of saving on file size. many graphics are
made up of a combination of simple shapes e.g. circles, lines, squares
etc. Why not just transmit the information to draw these simple shapes.
Shockwave movies and Flash graphics are becoming more and more common on the internet. |
How long does it take to transmit a digital image?
Broadband is becoming very common but many people still have 56k modems. These can receive a maximum of 56,000 bits every second.
A 100kbyte image would therefore take (100,000 x 8) / 56,000 = 14.3 seconds. Pages full of such images would take ages to load. The advantages of the techniques above become obvious.
