Search

EP-4027317-B1 - METHOD AND SYSTEM FOR PREDICTING MOTION TRAJECTORY

EP4027317B1EP 4027317 B1EP4027317 B1EP 4027317B1EP-4027317-B1

Inventors

  • SHEN, ZESHU

Dates

Publication Date
20260506
Application Date
20200717

Claims (10)

  1. A motion trajectory prediction method, wherein the method is applied to an intelligent vehicle and comprises: obtaining N motion trajectories of a first target, wherein the N motion trajectories are provided by N devices, and N is a positive integer greater than 1; and determining a to-be-applied motion trajectory of the first target based on types of the N motion trajectories, wherein the to-be-applied motion trajectory of the first target is used for traveling decision of the intelligent vehicle, wherein a type of a motion trajectory comprises a prediction trajectory type and a planning trajectory type, the prediction trajectory type indicates that the motion trajectory is obtained by a device through prediction based on information about the first target that is perceived by the device, the planning trajectory type indicates that the motion trajectory is obtained by the first target by planning traveling of the first target based on a traveling trajectory of a target around the first target, characterized in that , the determining a to-be-applied motion trajectory of the first target based on types of the N motion trajectories comprises a) or b): a) when the types of the N motion trajectories comprise both the planning trajectory type and the prediction trajectory type, determining a motion trajectory whose type is the planning trajectory type in the N motion trajectories as the to-be-applied motion trajectory of the first target; b) when all of the types of the N motion trajectories are the prediction trajectory type, determining a confidence level of each motion trajectory; and fusing the N motion trajectories based on the confidence level of each motion trajectory, to obtain the to-be-applied motion trajectory of the first target.
  2. The method according to claim 1, wherein the N motion trajectories are from N motion trajectory sets provided by the N devices, each motion trajectory set comprises motion trajectories that are of M targets and that are collected by each device, and M is a positive integer.
  3. The method according to any one of claims 1 to 2, wherein before the determining an to-be-applied motion trajectory of the first target based on types of the N motion trajectories, the method further comprises: performing security detection on the N motion trajectories, and discarding an untrusted motion trajectory of the first target.
  4. The method according to any one of claims 1 to 3, wherein the method further comprises: obtaining L motion trajectory sets, wherein L is greater than or equal to N; determining a region of interest, ROI, of the intelligent vehicle based on traveling information of the intelligent vehicle; screening, based on the ROI, out a motion trajectory that is not in the ROI in each motion trajectory set; and obtaining the N motion trajectories of the first target based on L motion trajectory sets obtained after screening.
  5. The method according to option b) of claim 1, wherein the confidence level of each motion trajectory is a confidence level of a motion trajectory set to which each motion trajectory belongs; and the method further comprises: calculating a degree of matching between a motion trajectory of a target in each motion trajectory set in a previous period and a motion trajectory of a corresponding target in an to-be-applied motion trajectory set in the previous period; and determining, based on the matching degree, a confidence level of a motion trajectory set to which each motion trajectory in a current period belongs.
  6. The method according to option b) of claim 1, wherein the confidence level of each motion trajectory is a confidence level of a motion trajectory set to which each motion trajectory belongs, and the confidence level of the motion trajectory set to which each motion trajectory belongs is determined based on an attribute of a device having generated and thus corresponding to the motion trajectory set.
  7. An apparatus, which is configured to perform the method according to any one of claims 1 to 6.
  8. A device, comprising a processor and a memory, wherein the memory stores program code, and the processor executes the program code, so that the device performs the method according to any one of claims 1 to 7.
  9. A computer-readable storage medium, wherein the computer-readable storage medium stores computer program code, and when the computer program code is executed by a computing device, the computing device performs the method according to any one of claims 1 to 6.
  10. A computer program product when executed by a processor, causes the processor to carry out the steps of the method of any one of claims 1 to 6.

Description

TECHNICAL FIELD This application relates to the intelligent driving (intelligent driving) field, and in particular, to a motion trajectory prediction method and a system. BACKGROUND Intelligent driving has great potential in reducing traffic accidents, alleviating traffic congestion, and improving road utilization, and has become a competitive hot spot for many enterprises. The intelligent driving means that a vehicle can intelligently perform perception, determining, inference, decision, and memorization like a human being, and intelligently control or drive the vehicle. A vehicle having an intelligent driving capability is referred to as an intelligent vehicle. The intelligent vehicle can implement active prediction and control, and can perform man-machine interaction and coordination. An important technology for implementing intelligent driving is to predict a motion trajectory of a target in a surrounding environment of the intelligent vehicle in a future time period, and then decide or adjust a motion trajectory, a running speed, and the like of the intelligent vehicle based on the motion trajectory of the surrounding target (for example, based on the motion trajectory of the target in the surrounding environment, determine when to change a lane, determine a traveling speed, and determine whether to perform avoidance). Accuracy of predicting a motion trajectory of another surrounding target is closely related to traveling safety of the intelligent vehicle. In a related method for predicting a motion trajectory of a target in a surrounding environment, information about the target in the surrounding environment is usually perceived based on a sensor system of the intelligent vehicle, and a motion model is constructed based on the information about the target to predict a motion trajectory of the target in a future time period. A prediction result that is of the motion trajectory of the target and that is obtained by using this method has relatively low accuracy, and therefore, traveling safety of the intelligent vehicle cannot be ensured. How to predict a motion trajectory more accurately becomes an urgent problem to be resolved in the intelligent driving field. US 2018/188738 A1 discloses a system and method for dynamic object identification and environmental changes for use with autonomous vehicles. For efficient detection of changes for autonomous, or partially autonomous vehicles, embodiments may use a technique based on background removal and image subtraction which use motion detection rather than full object identification for all objects in an image. Road side units proximate to a road segment or virtual road side units in the cloud, other vehicles or mobile device (e.g., drones) are used to retrieve and store background images for a road segment, to be used by the autonomous vehicle. US 2019/287394 A1 discloses a Roadside Equipment (RSE) located at an intersection of a transportation network. The RSE includes an input to receive data from a sensor oriented to monitor ground transportation entities at or near the intersection. A wireless communication device sends to a device of one of the ground transportation entities, a warning about a dangerous situation at or near the intersection. SUMMARY This application provides a motion trajectory prediction method defined in claim 1. In the method, an to-be-applied motion trajectory of a first target is determined based on types of N motion trajectories of the first target, so that the obtained to-be-applied motion trajectory of the first target is more accurate. This further improves traveling safety of an intelligent vehicle. According to a first aspect, this application provides the motion trajectory prediction method of claim 1. In a possible implementation of the first aspect, the N motion trajectories are from N motion trajectory sets provided by the N devices, each motion trajectory set includes motion trajectories that are of M targets and that are collected by each device, and M is a positive integer. According to claim 1, a type of a motion trajectory includes a prediction trajectory type and a planning trajectory type, the prediction trajectory type indicates that the motion trajectory is a motion trajectory obtained by a device by performing prediction based on information about a target that is perceived by the device, and the planning trajectory type indicates that the motion trajectory is a motion trajectory obtained by the target by planning traveling of the target based on a traveling trajectory of another target around the target. According to the invention, the determining an to-be-applied motion trajectory of the first target based on types of the N motion trajectories includes: when all of the types of the N motion trajectories are the prediction trajectory type, and N is greater than 1, determining a confidence level of each motion trajectory; and fusing the N motion trajectories based on the confidence level of each motion trajec