Development of an Intelligent Road Anomaly Detection System for Autonomous Vehicles

Abstract

Globally, road transportation has become one of the most reliable means of moving goods and services from one place to the other. It has contributed immensely to the standard of living and modern civilization. However, this means of transportation is characterised by some issues which are poised to be harmful to the human population if not properly addressed. One of such issues is the presence of potholes, bumps, and other road anomalies. Unfortunately, the late identification of road anomalies (Speedbumps and Potholes) and the inability of drivers to detect and slow down while approaching such road anomalies has also been a big challenge faced by many nations. Therefore, there is a need for an automatic and intelligent approaches to be built into vehicles to mitigate the number of road accidents caused by these anomalies. In this work, the development of an intelligent road anomaly identification and manoeuvring system for autonomous vehicle is presented. The developed system focuses on the detection of road anomalies specifically speedbumps and potholes; and the regulation of the vehicular speed when these anomalies are detected. A modified Histogram Oriented Gradient (HOG) and Fuzzy Logic Control (FLC) have been proposed in this work. Furthermore, promising results have been obtained and presented which depicts the proposed HOG algorithm outweigh other techniques in the identification and detection of speedbumps and potholes. In addition, the developed FLC was able to regulate the speed of the vehicle in the presence of speedbumps as well as navigate the vehicle accordingly in the presence of potholes.