The invention relates to a vehicle track prediction method based on heterogeneous node time-space domain perception, and the method comprises the steps: building a vehicle track prediction task which extracts an online perception map generated based on a driving vehicle visual angle perception result, and employs a time-space domain attention mechanism to complete the self understanding and mutual fusion of different types of information; according to the method, map information in a local range of the predicted target vehicle can be effectively obtained without using a high-precision map, high robustness and estimation precision are ensured while track prediction is completed, and the method is more in line with actual application requirements.
本发明涉及一种基于异构节点时空域感知的车辆轨迹预测方法,搭建了一个提取基于行驶车辆视角感知结果生成的在线感知地图并使用时空域注意力机制来完成不同种类信息自身的理解和相互的融合的车辆轨迹预测任务,可以在不使用高精地图的同时有效获取预测目标车辆的局部范围内的地图信息,完成其轨迹预测的同时保证较高的鲁棒性和估计精度,更加符合实际应用的需求。
Vehicle track prediction method based on heterogeneous node time-space domain perception
一种基于异构节点时空域感知的车辆轨迹预测方法
2023-05-30
Patent
Electronic Resource
Chinese
IPC: | B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion |
Vehicle track prediction method based on intention perception space-time attention network
European Patent Office | 2023
|European Patent Office | 2024
|Other-party vehicle track prediction method based on visual perception
European Patent Office | 2021
|Vehicle track prediction method based on multi-vehicle space-time interaction relationship
European Patent Office | 2024
|Real-time measurement of track curvature based on vehicle attitude perception
Taylor & Francis Verlag | 2023
|