Multi-sensor fusion target localization methods, electronic devices and storage media

By matching measurement information with trajectory prediction information and constructing pseudo-measurement values, the problems of trajectory splitting, jittering, and pulling caused by measurement gaps in multi-sensor fusion are solved, achieving more accurate target positioning and ensuring the stability and safety of the autonomous driving system.

CN120489149BActive Publication Date: 2026-07-17ECARX (HUBEI) TECHCO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ECARX (HUBEI) TECHCO LTD
Filing Date
2025-06-04
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

In autonomous driving systems, when multi-sensor data is fused, the measurement results are uncertain due to the independence of each sensor and the influence of the environment. This leads to a large difference in ranging between cameras and radar, causing problems such as track splitting, shaking, and pulling, which affect the stability of driving functions.

Method used

By acquiring the measurement information at the current moment and the trajectory information at the previous moment, the trajectory prediction information is determined based on the trajectory information. The measurement information and the trajectory prediction information are matched to construct a pseudo measurement value, and the trajectory value at the current moment is determined. By combining the pseudo measurement value and the trajectory prediction value, the target is located, and the error caused by the measurement difference is reduced.

Benefits of technology

It improves the accuracy of target positioning, solves problems such as track splitting, shaking, and pulling caused by large measurement differences between sensors, and ensures the stability and safety of driving functions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120489149B_ABST
    Figure CN120489149B_ABST
Patent Text Reader

Abstract

This application provides a multi-sensor fusion target localization method, electronic device, and storage medium. The method acquires measurement information at the current moment and track information at the previous moment, determines track prediction information at the current moment based on the track information, and then matches the measurement information and track prediction information to obtain track measurement matching pairs. For each track measurement matching pair, a pseudo measurement value is constructed based on the track prediction value and the measurement value in the track measurement matching pair. Based on the pseudo measurement value and the track prediction value, the track value of the corresponding target at the current moment is determined, thereby achieving target localization. This method can make the final localization result closer to the true position of the target and solve the problems of track splitting, jittering, and pulling caused by large measurement differences between sensors during fusion, so as to ensure the stability of driving functions and driving safety.
Need to check novelty before this filing date? Find Prior Art