Autonomous driving and driver assistance require a continuous and reliable perception of the road boundaries, namely curbs and berms, including also other minor, or not so minor, obstacles in the neighborhood of the car. This paper proposes to use a 4-layer LIDAR placed close to the ground to capture measurements of the road ahead of the car and allow the detection of the boundaries. This setup provides a special point of view that allows the accumulation of points on vertical surfaces on the road as the car moves, which increases the point density in vertical surfaces but keeps it limited in horizontal surfaces. This technique allows to successfully distinguish curbs from the flat parts of the road. However, this approach has some limitations, namely to detect berms, and another approach had to be developed using the gradient of point density, which extends the detection capabilities to berms and negative obstacles. This is achieved by flattening the point clouds to 2D and use traditional computer vision gradient and edge detection techniques, which also improves the processing speed. Results are obtained on the ATLASCAR real system, at different velocities, and a good performance is reached when comparing to a manually created ground truth.

自主驾驶和驾驶员协助需要对道路边界（即路缘和护堤，包括在汽车附近的其他较小或不太轻微的障碍物）的持续且可靠的感知。本文建议使用靠近地面放置的4层激光雷达，以捕获汽车前方道路的测量数据并允许对边界进行检测。这种设置提供了一种特殊的视角，允许在汽车行驶时在道路上的垂直表面上积累点，从而增加了垂直表面上的点密度，但又限制了水平表面上的点密度。这种技术可以成功地将路缘石与道路的平坦部分区分开。但是，这种方法有一些局限性，即检测边缘，必须使用点密度梯度开发另一种方法，从而将检测能力扩展到边缘和负面障碍物。这是通过将点云展平为2D并使用传统的计算机视觉梯度和边缘检测技术来实现的，这也提高了处理速度。在ATLASCAR真实系统上以不同的速度获得结果，并且与手动创建的地面真相进行比较可获得良好的性能。