1 Definition
- In trajectory data analysis, "Staypoint" (Staypoint) is a very key concept, which reflects the stay behavior of an individual or object at a certain location.
- Typically, a small geographic area is considered a geographic area if, over a period of time, a person or object moves below a certain threshold and stays longer than a set time threshold. A stopping point.
- As shown in the figure above, starting from p5, the spatial distance between subsequent points p6~p8 and p5 is relatively small, and the time span from p5 to p8 is large.
- ——>p5~p8 are considered as a dwell point
2. Track dwell point detection
- The key is how to define the "small" space range and the "large" time span.
The main idea comes from Mining User Similarity Based on Location History
2.1 preliminary
- GPS track
- A string of GPS points, P={p1,p2,...pn}
- Each trajectory point is represented by the corresponding position and time (Lat, Lngt, T)
- Residency point: represented by S
- Dwell point P3: long stay
- Stay point [P5~P8]: Stay in an area
2.2 Algorithm idea
- Starting from the first trajectory point pi, use this as the anchor point, and find all points whose distance does not exceed the distance threshold distThreh backward.
- If the total duration of these points is greater than the time threshold timeThreh, they are classified as stay points, and the next point in the stay point set is used as the new anchor point.
- Otherwise, move the anchor point back one position and recall and judge again.
2.3 Pseudocode
2.4 Advantages and Disadvantages
- advantage
- Implementation is simple, fast and easy to understand
- shortcoming
- After fixing the starting point, the upper limit of the threshold may be reached in the stay area, causing the stay point to be separated inside and become two consecutive stay point sets.
- The parameters (thresholds) are difficult to determine, and the thresholds in different areas should not be fixed.