论文部分内容阅读
为了检测车辆目标,提出了一种基于主被动传感器融合的车辆检测方法.将车辆检测分为假设和验证假设两个步骤,在假设阶段,通过被动传感器——毫米波雷达进行目标的检测与跟踪,在对雷达数据进行最邻近法聚类后,在多假设跟踪模型下,将观测目标集与通过卡尔曼滤波器预测的目标集进行数据关联,得到雷达目标.在验证假设阶段,首先通过新出现的雷达目标找出车辆可能存在的区域,然后通过训练好的分类器对这些区域进行验证得到最终的车辆目标.在实验室的无人自主车平台上,本系统在城市道路和乡村道路环境下进行了大量实验,结果表明本文的方法可以有效地检测并跟踪到车辆目标,得到目标的距离和速度信息,从而帮助自主平台实现更多功能.
In order to detect the vehicle target, a vehicle detection method based on the fusion of active and passive sensors is proposed.The vehicle detection is divided into two steps: the hypothesis and the verification hypothesis. In the hypothetical phase, the passive sensor-millimeter-wave radar is used to detect and track the target After the radar data is clustered by the nearest neighbor method, the radar target is obtained by correlating the observation target set with the target set predicted by the Kalman filter under the multi-hypothesis tracking model. In the verification hypothesis phase, The emerging radar target locates the possible areas of the vehicle and then verifies these areas with the trained classifier to get the final vehicle target.Under the laboratory unmanned vehicle platform, the system is used in urban road and country road environment The experimental results show that the proposed method can effectively detect and track the vehicle targets and obtain the target distance and speed information, so as to help the autonomous platform to achieve more functions.