An Interactive Region-Based Filter for Moving Objects Datasets: Making Sense of Chaos

Moving objects databases are able to store and retrieve the movements of objects such as cars, people, animals, and vehicles. This data can be collected from GPS and Wi-Fi tracking systems, computer vision, or even the movements of virtual objects in simulations. Once we have this data, the next question is how can it be meaningfully interpreted, manipulated or queried? Moving objects datasets can appear chaotic when viewed in their entirety. This work aims to make moving object data accessible to analysts by way of a simple but powerful visual interface. To this end, we have formally defined a region-based filter that partitions moving object datasets. In practical terms, a filter is constructed, in real-time, by drawing and manipulating regions on a visualisation of a dataset. The partitions are then visualised to identify meaningful subsets of the data. We evaluate our prototype in terms of its ability to provide real-time feedback for varying types and quantities of filter regions. Finally, we evaluate the usefulness of the approach by exploring a possible scenario.