Multi-dimensional Pattern Discovery of Trajectories Using Contextual Information Mohammad Sharif Ali Asghar Alesheikh Faculty of Geomatics Engineering K. N. Toosi University of Technology, Tehran, Iran
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Multi-dimensional Pattern Discovery of Trajectories Using Contextual Information
Mohammad SharifAli Asghar Alesheikh
Faculty of Geomatics EngineeringK. N. Toosi University of Technology, Tehran, Iran
Introduction
2
} Movement: moving objects / processes
} Movement point object: tracking / positioning systems
} Trajectory (x, y, t)
} Movement is taking place in different “contexts”
} Contextualized trajectories (x, y, t, c)
} Movement analysis: GKD | similarity measure | pattern discovery
( time, traffic, weather condition, etc.)
Methodology
3
} Similarity measure function: Dynamic Time Warping (DTW)- Speech recognition- Suitable for trajectories with missing information - Addresses parametric data very well- Handles trajectory of different sizes- ....
Methodology (Cont.)
4
} Contextualizing similarity measure and pattern discovery for trajectories
Dataset
5www.aviationweather.govwww.flightaware.com
Implementation
6
} Mining Similar PatternsLatitude, Longitude, Time
Altitude, Airplane’s ground speed,Time
- Reference Trajectory (in red)- All trajectories (in light green)- Discovered trajectories (in dark green)
Implementation (Cont.)
7
} Mining Patterns
Boeing
Airbus
McDonnell Douglas
Embraer
Implementation (Cont.)
8
} Multi-dimensional pattern discovery of trajectories
Conclusion
9
} new insights into the similarity analysis and pattern discovery of trajectories based on spatial and contextual information
} the DTW distance function is shown sensitive to small alterations in context variations
} movements of point objects are highly affected by both internal and external contexts
} added values of context information is significant in discovering patterns among trajectories
10
Thank you
Example
11
Relative similarity values of four trajectories: (a) Latitude and longitude (2D); (b) Latitude, longitude, and altitude (3D); (c) Latitude, longitude, altitude, airplane speed, and airplane heading; (d) Wind speed and wind direction; (e) All the previous dimensions together
(a) (b) (c) (d) (e)
Related Documents
