本文为美国大峡谷州立大学（作者：Milan Aryal）的硕士论文，共46页。

自动驾驶车辆周围物体的检测和跟踪对车辆的安全运行至关重要。本文提出了一种检测、分类和跟踪目标的算法。所有物体按类型（如车辆、行人或其他）分为移动或静止目标。提出的方法使用最先进的深度学习网络YOLO（You Only Look Once）结合激光扫描仪数据来检测和分类物体，并估计物体在汽车周围的位置。面向FAST和旋转BRIEF（ORB）的特征描述符用于将同一对象从一个图像帧匹配到另一个图像帧，利用扩展卡尔曼滤波器将该信息与GPS/INS耦合测量数据融合。由此得到的解决方案有助于定位汽车本身及其环境中的对象，从而使汽车能够在道路上安全地自主驾驶。该算法已经开发完成并使用牛津机器人汽车收集的数据集进行了测试，该机器人汽车配备了摄像头、激光雷达、全球定位系统和惯性导航系统，并在牛津市中心拥挤的城市环境中采集了数据。

The detection and tracking of objectsaround an autonomous vehicle is essential to operate safely. This paperpresents an algorithm to detect, classify, and track objects. All objects areclassified as moving or stationary as well as by type (e.g. vehicle,pedestrian, or other). The proposed approach uses state of the artdeep-learning network YOLO (You Only Look Once) combined with data from a laserscanner to detect and classify the objects and estimate the position of objectsaround the car. The Oriented FAST and Rotated BRIEF (ORB) feature descriptor isused to match the same object from one image frame to another. This informationfused with measurements from a coupled GPS/INS using an Extended Kalman Filter.The resultant solution aids in the localization of the car itself and theobjects within its environment so that it can safely navigate the roadsautonomously. The algorithm has been developed and tested using the datasetcollected by Oxford Robotcar. The Robotcar is equipped with cameras, LiDAR, GPSand INS collected data traversing a route through the crowded urban environmentof central Oxford.

1. 引言

2. 相关论文

3. 文献回顾、扩展方法与结果

更多精彩文章请关注公众号：