A statistical method for detecting Move, Stop, and Noise episodes in trajectories

Detecting stops is an important task in trajectory analysis. Stops can reveal interesting aspects of a moving object behavior such as its daily routine, bottlenecks in traffic jams, or visiting times of touristic places. In order to record those traces, trajectories must be sampled and, in some cases, post-processed. This process from collecting raw data to storing them may vary according to the devices and applications that collect the data. Another important characteristic in many trajectories is the presence of noisy segments, a fact is often ignored by most stop detection methods. In this work, we present a method that exploits gaps in time and space to identify episodes of movement, stop, and periods where some classification is inconclusive, which we define as noise. In addition, our method does not rely on contextual information as opposed to some current methods, which makes our proposal also suitable for trajectories recorded in free space.