showing the potential airport locations.
of the trajectories associated with the airport locations.
This web page presents the results form our experiments involving the detection, and general classification, of airport locations, and types.
All of this is done in a way that assumes a lack of previous knowledge of airport in the region, and relies exclusively on the analysis of
the trajectory data.
Such methods could have applications in situations requiring the detection of airports in situations where direct access
is not possible, such as remote or hostile environments.
By analyzing aircraft trajectories we are able to extract those which were involved
in a take off or landing maneuver and extract the potential location of the runway where this maneuver occurred.
Once this has been done
for all aircraft, we apply a density based clustering algorithm in an effort to find locations which frequently experience takeoff or landing
events, indicating the presence of an airport.
We can then analyze the trajectories associated with a particular cluster and classify them as
one of three likely aircraft types.
Then, based on the inferred aircraft types present at these airport locations, we can provide a generalized
classification of the airport type.
Additionally, an overview of an index based trajectory search engine is provided, as well as the methods
used to visualize the results.
The following images are examples of the type of capability this method can offer. Clicking on the images will display it in full size.
| |
|
| The result of density clustering, showing the potential airport locations. |
The result of airport classification based on the analysis of the trajectories associated with the airport locations. |
Additional information can be found via these links:
Project Presentation
Project Technical Paper