Abstract: Raw GPS trajectories are usually long and the data volume is large. The transmission， storage and processing of trajectory data are becoming more and more difficult. So， raw GPS trajectories data need to be compressed. However， those problems still exist in the current DPTC (direction-preserving trajectory compression) methods， such as that error tolerance is difficult to determine， computation cost is high， and compression effect is bad. For these problems， the trajectory compression method based on sort tree index was proposed. With the help of the sort tree index， effective pruning in the trajectory compression process improved the efficiency of trajectory compression. Meanwhile， the segment error was redefined， which played a decisive role in determining whether a trajectory point is preserved in the compressed trajectory or not， and the compression effect was enhanced. The extensive experimental results on real data sets show the effectiveness and efficiency of the proposed method.

Key words: trajectory compression, sort tree index, trajectory segment, segment error, compressed trajectory error