Time-Oriented Trajectory Query and Comparison

Line-based trajectory drawings face several challenges for trajectory query scenarios. They produce clutter or small visual differences, impeding the visual comparison between search results and query. They also do not represent the temporal dimension, causing a potential mismatch between visual representation and applied similarity measure. To avoid these limitations, we present a time-oriented approach to trajectory query and comparison, with scarf plots as primary visual representation. We discuss why the simple encoding of scarf plots can be advantageous for the visual comparison, and we present our approach to visual interactive fuzzy area of interest definition to align the visual representation and similarity metric, while simultaneously allowing the intuitive and flexible specification of sequence queries. We also discuss how the backend query engine can be adjusted with the crop & merge algorithm to produce accurate search results from a sliding-window search.