A Trajectory Preprocessing Method Based on Angle and Velocity

Abstract

The fusion of angle and velocity information allows for a more comprehensive description of object motion characteristics, which
holds significant importance for trajectory data analysis and exploration. However, conventional trajectory data processing typically over_x005flooks direction and velocity information between adjacent trajectory points. This paper’s methodology involves initially collecting raw trajectory data, followed by assessing abnormal trajectory points through angle and velocity computations, and ultimately conducting filtering
and refinement for integration. Experimental results demonstrate that preprocessing by integrating angle and velocity effectively enhances the
accuracy and usability of trajectory data, offering richer information for trajectory analysis, behavior identification, and predictive modeling
of motion patterns.