Vehicle trajectory analysis method, device, and storage medium
By clustering and rasterizing historical vehicle trajectories, and combining the actual trajectory of the target vehicle with information from neighboring vehicles, the problems of abnormal vehicle trajectories and reliance on high-precision maps in sensor detection are solved, enabling efficient vehicle trajectory analysis and prediction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG DAHUA TECH CO LTD
- Filing Date
- 2023-12-18
- Publication Date
- 2026-07-24
AI Technical Summary
In roadside intelligent transportation scenarios, the vehicle trajectories detected by sensors are prone to anomalies such as vehicles disappearing or trajectory changes due to the limited detection or tracking algorithm capabilities. Furthermore, existing technologies rely on high-precision maps for vehicle trajectory detection, which makes it difficult to adapt to real-time changes in the road, resulting in low analysis efficiency.
By clustering the starting and ending points of historical vehicle trajectories to form trajectory starting clusters and trajectory ending clusters, and combining them with the actual trajectory of the target vehicle, the target vehicle trajectory is determined using rasterization and fitting techniques. Furthermore, the motion information of neighboring vehicles is used for trajectory prediction and correction, reducing the reliance on high-precision maps.
It improves the efficiency and versatility of vehicle trajectory analysis, enabling accurate analysis and prediction of vehicle trajectories under complex and ever-changing road conditions, reducing reliance on high-precision maps and enhancing analysis efficiency.
Smart Images

Figure CN117911455B_ABST