Use of LiDAR for Negative Obstacle Detection: A Thorough Review

Abstract

Abstract Negative obstacles for field autonomous land vehicles refer to potholes, ditches, cliffs, pits, or any type of obstacles that are on the road but not in a visible way, which bring risks both to the vehicles or potential passengers and even to the environments where autonomous land vehicles are moving along. These negative obstacles can cause severe damage to autonomous land vehicles, from damage to the vehicle’s suspension, rollovers, or even loss of the autonomous vehicle. Obstacle detection is the first step to avoid any risk, and it is extremely important to be able to warn of nearby obstacles, to avoid any type of danger that may arise. In recent years several articles have been published on the different types of hazards or obstacles and different methods of avoiding them. The authors report on ways to better understand the environment so that autonomous land vehicles stay out of harm’s way. Some of the dangers could be, the collision with another vehicle, hitting a pedestrian, not recognizing the traffic signs on the road, leaving the planned routes or lanes, or not braking when they detect an obstacle. The use of image recognition using different types of cameras, as well as the use of ultrasonic sensors and Light Detection and Ranging (LiDAR) sensors, or the combination of some or all of the sensors is widely used in the field of autonomous land vehicles. Most of this research work has been carried out in positive obstacle detection; that is, obstacles that are above the road. However, the detection of negative obstacles and their avoidance remains less explored. Due to this, an exhaustive review is made of the state of the art of use of LiDAR sensors in the detection of negative obstacles applied in autonomous land vehicles and the different techniques that have been used to recognize and classify the different types of potholes, ditches, pits.