Automatic Vehicle Type Classification Using Strain Gauge Sensors

Abstract

In this paper we describe the use of machine learning algorithms (Naive Bayesian, neural network, and support vector machine) on data collected from strain gauge sensors to automatically classify vehicles into classes, ranging from small vehicles to combination trucks, along the lines of Federal Highway Administration vehicle classification guide. Knowing the types of vehicles can help reduce operating costs and improve the health monitoring of infrastructure and would help to make transportation safer and personalized; use of such non-image-based data permits user privacy. Our results indicate that the support vector machine technique outperforms the rest with an accuracy of 94.8%