Changed detection area file format to polygon based + enabled editing

[hekkup]

• Fixes UFOID/UFO-Detector#46
• Detection area file is now in polygon format which means it is easy to edit manually in CLI usage. It's also up to 100 times faster than the point file. ActualDetector is using the same type of point vector for detection area as before so nothing else changed in that regard but the file format.
• Added possibility to edit detection area in the editor dialog. Existing area is shown first.
• CLI version has an option to reset the detection area file (--reset-detfile --reset-area-file) in order to have a skeleton file for editing.

The new file format looks like this:

<?xml version="1.0" encoding="UTF-8"?>
<detectionarealist>
  <detectionarea>
    <camera id="0" width="352" height="288"/>
    <polygon>
      <point x="0" y="0"/>
      <point x="0" y="287"/>
      <point x="351" y="287"/>
      <point x="351" y="0"/>
    </polygon>
  </detectionarea>
</detectionarealist>

The format makes it possible to have multiple cameras and have a detection area defined for each of them. Also, it's possible to have multiple polygons in a detection area. However, these are not implemented in the code so far.

BTW: Mark, did you compare which one is faster/handier, to use the current point vector detection area, or to define it as a mask image?

It seems this pull request is including everything from the point my master branch forks from your master branch... Hope I'm not screwing anything here. :)

[UFOID]

The way the selection works with this pull request seems ideal. Obviously a massive improvement in file size and time it takes to save/load! I think we can keep it as a single polygon for now. Give me one more day to look at the code properly.