Classifying road surfaces using smarphone accelerometers for informed road transportation
Road navigation applications such as Google Maps and Apple Maps provide routing information to their continuously increasing number of users, enabling them to get from one destination to another. These applications provide information such as routes and traffic conditions which influence the time taken to travel by the user of the information. However, these navigation services are lacking in providing road surface quality information. Road surface quality information of a route not only influences the time taken to travel the route, but also provide salient information on the comfort of travel for the passenger and the effect the terrain will have on the vehicle. This work builds on previous work by further developing and characterizing a Logistic Regression (LR) algorithm for classifying road surface quality using accelerometer data sourced from mobile devices in moving vehicles along four different types of roads: very good, good, bad, and very bad roads.
Undergraduate thesis submitted to the Department of Computer Science, Ashesi University, in partial fulfillment of Bachelor of Science degree in / Computer Science, May 2020
road surface quality, accelerometer, Convolutional Neural Networks (CNN), Logistic Regression Algorithm