Skip to content

Latest commit

 

History

History
95 lines (69 loc) · 2.88 KB

Manual.md

File metadata and controls

95 lines (69 loc) · 2.88 KB

A Short Manual on subimtrack

It is assumed that you have started the image tracking in a ipythonshell (see Getting Started in Readme).

Explanation of Visualization

Example Track

The example shows you an natural color Meteosat image. The ground is shown in greenish colors, the sea would appear dark, and clouds are either white or cyanish depending on the amount of cloud ice in the top levels.

You also see the current track point as red circle.

Moreover, previous and subsequent track points are provided as white circles.

Typical Workflow

You will typically choose the following workflow:

  1. create a manual track
  2. store the track
  3. continue with item 1. (make the next track and store it)

Creating a track

The following steps are reasonable:

  • Zoom into the region of Interest (Strg + Mouse Window)

  • Create a Track Point (Left Mouse Button)

  • ( Delete Wrong Track (Backspace) )

  • Proceed to next image (Arrow Down)

  • ( Or go the previous one (Arrow Up)

  • Create a Track Point

  • Finally store the track (Right Mouse Button)

Key- and Mouse-Bindings for Tracking

The subimtrack.runcommand starts an interactive matplotlib window. You will interact with that window using mouse and keyboard.

Mouse-Bindings

Mouse Buttons Action
Left Mouse Button Sets the Track Point
Middle Mouse Button Deletes the Full Track
Right Mouse Button Saves the Track in a netcdf File / Starts a new Track

Key-Bindings

  • Tracking
Shortcut Keys Action
Arrow Down Next Image
Arrow Up Previous Image
Backspace Remove Track Point
  • Zoom
Shortcut Keys Action
Strg + Hold Left Mouse Button Create a Zoom Window
Esc Reset Zoom to Full Image
  • Experimental (default nstep = 5)
Shortcut Keys Action
Page Down +nstep Next Image
Page Up -nstep Previous Image
+ double nstep
- half nstep

Throughts on Post-Processing

Finally, you will have a simplistic set of track data stored. For instance,

> ncdump -h SubImTrack/test/tracks/track_20120608T1050Z_P222x397.nc 
netcdf track_20120608T1050Z_P222x397 {
dimensions:
	time = 3 ;
	string14 = 14 ;
variables:
	char time(time, string14) ;
		time:_Encoding = "utf-8" ;
	double column_index(time) ;
		column_index:_FillValue = NaN ;
	double row_index(time) ;
		row_index:_FillValue = NaN ;
}
  • the filename contains the start point information
  • time : contains your time identifier string which you might need to parse
  • column_indexand row_index : contain the respective row and column indices as float values

You need to read a geo-reference information that exactly fits with your image to translate the indices into longitude and latitude values.