A novel trajectory segmentation and compression approach for moving objects based on arc and polyline integration

Objects moving in free space typically have trajectories that contain one or more approximate arc segments. To achieve better fitting between the compressed trajectory and original trajectory, widely used trajectory segmentation and compression methods based on polyline simplification require more segmented feature points to be retained for arc segments. Consequently, the compression rate cannot be effectively reduced. In this study, we propose a novel trajectory segmentation and compression approach based on arc and polyline integration, which improves global trajectory simplification quality by accurately extracting approximate arc segments from the trajectory and optimizing their compression. First, the compressed representation of directional arc segment is defined, enabling unique expression and reconstruction of an arc segment. Then, two methods for detecting arc segments contained in the trajectory are presented: One based on median filtering corner deviation, and the other based on approximately circle deviation caused by three points forward pushing and point adding. Finally, the detected arc segments are compressed and represented separately, while the remaining trajectory segments are simplified by polylines. By integrating compressed arcs and polylines, better trajectory simplification is achieved. To evaluate the compression effect, two indexes applicable for arc and line integrated compression are established based on perpendicular Euclidean distance and geometric transformation cost. Experiments with both the real and simulated trajectory data are implemented. Experimental results demonstrate that compared to typical polyline simplification methods, the proposed method further reduces the compression rate while maintaining good compressed trajectory accuracy.