Trajectory Distances in Java

The bachelor thesis of Lisa Fischer was concerned with discussing and understanding some properties of well-known trajectory similarity measures. As part of this work, I am happy that she provided several implementations of different trajectory distance functions and time-series similarities in the Java programming language. All algorithms have been carefully implemented. However, they are not ready to be used in real projects, but intended for research and teaching purposes. You can download the sources from the following link. It contains source code, some binaries and a mixed language (German and English) documentation. Download Link

List of Classes and Algorithms

 Preprocessing

  • Douglas-Peucker Line Simplification
  • Opening Window Algorithm

Distance Measures

  • Point to Trajectory
  • Closest Pairs
  • Sum of Pairs Distance (for equal-length trajectories)
  • Dynamic Time Warping
    • Matrix-Version
    • Sakoe-Chiba
    • FastDTW
  • Longest Common Subsequence (LCSS)
  • Edit Distance on Real Sequences (EDR)
  • Hausdorff distance
  • Discrete Fr√©chet Distance