Trembilski, Andrzej (2003)
Naturalistische Methoden zur Visualisierung meteorologischer Daten in Augmented Video.
Technische Universität Darmstadt
Dissertation, Bibliographie

Kurzbeschreibung (Abstract)

Numerical weather simulation models, like these of the German Weather Service (DWD), produce twice a day large quantities of simulation data, which must be visualized for the presentation on television. Special weather phenomena, which occur only locally, would disturb and are not desired or implemented in such a general view in today TV shows. The requirements change as soon as one tries to make a local weather presentation. They result from the combination of the techniques of Augmented Video and Scientific Visualisation. If weather visualisation should be integrated into a video film, the schematic, global representation of the clouds is not longer acceptable. On the one hand the spectator can compare the artificial clouds with the genuine world presented in the video. Any break in the representation diverts from the actual information and irritates the spectator. On the other hand only a realistic cloud representation yields understandable weather forecasts. The actuality of the data in the presentation is an important restriction for the selection of the visualisation methods. This work presents efficient methods to generate Augmented Video visualisations from weather forecast data. Several presented algorithms include the naturalistic generation of clouds and sky for the automatic insertion into the video sequence. Another part of the work deals with manipulation of the video sequences themselves for a better integration of the computer generated parts of the image and the mapping of the sky colours at the video sequence. The introduction of a general model for the visualization in augmented video closes the work.

Frage zum Eintrag

Redaktionelle Details anzeigen