一种基于传感器的目标航迹起始决策算法在摄像头与毫米波雷达融合系统的应用

Application of a Sensor-based Initial Decision Algorithm for Target Track in the Fusion System of Camera and Millimeter Wave Radar

自动驾驶L2级别相关的如ACC、AEB、LKA、ELK、TJA等功能实现的准确度和驾乘舒适度在很大程度上依赖于融合算法的目标的稳定性、正确性和及时性,因此融合算法已成为多传感器融合系统中的最重要的算法之一。一个被传感器量测到的目标是否能够成为系统认可的目标并开始形成被系统认可的航迹,是多传感器融合算法的重要部分。文章介绍了一种摄像头与毫米波雷达的融合系统,并展示了一种基于传感器对量测特点的航迹起始决策算法在该系统中的应用,有效地提高了摄像头与毫米波雷达融合系统中航迹起始的准确率。

The driving comfort and the performance of ADAS functions applied in L2 autonomous driving such as ACC,AEB,LKA,ELK and TJA depend highly on the stability,accuracy and timeliness of fusion targets.Fusion algorithm is one of the most important algorithms in multi-sensor fusion system.Whether a sensor-detect target can be a target confirmed by the fusion system and therefore initiate its track is one of the most important problems in multi-sensor fusion algorithm.In this paper,a single-camera-single-radar fusion system is introduced,in which a track initiation algorithm based on the characters of each sensor is also shown,and the accuracy of track start in the fusion system of camera and millimeter wave radar is improved effectively.