Advances in Automobile Engineering
Open Access

Abstract

Pothole Detection using LiDAR

Varsha Prasad*, Shivani Kumari, Suryatej IP

Advanced Driver Assistance System (ADAS) are systems developed to enhance, automate and adapt with vehicle systems for safety and better driving experience.
Our ADAS system proposes to detect potholes on the road which are beyond the driver’s visibility range in real-time, allowing the driver with sufficient time to react. This pothole detection is accomplished using a LiDAR sensor and detection algorithm that is running on the computer. This system uses a LiDAR sensor to detect a pothole. LiDAR is a laser sensor used to detect the relative distance between the target and the sensor. We are using a LiDAR sensor package that provides us 360’ distance data by having the sensor on a rotating platform. By using a 2D LiDAR angle and distance, information about the roads is obtained.
An algorithm is designed to define if pothole is present and the location of the pothole relative to the vehicle. The algorithm of pothole detection includes clustering, line segment extraction, generation of data function, gradient of data function.
After pothole is detected, the GPS location of the pothole is obtained using a GPS sensor. The pothole data including the location of the pothole is then sent to a cloud database. A web application designed to mark potholes on Google Maps embedded in a custom webpage using Maps JavaScript API.

Published Date: 2021-06-23; Received Date: 2021-06-02