Sunlight Enabled Vehicle Detection by LED Street Lights

We propose a traffic sensing system that utilizes the photoelectric responses of the LED street lights to ambient light. A support vector machine based learning algorithm is developed to classify whether a vehicle passed by.

I. Introduction

Vehicle detection has become an indispensable part of the intelligent transportation system. Through monitoring and collection of traffic flow data, it is possible to analyze the distributions of the traffic flow density and speed. With the development of the visible light communication (VLC) technology [1], much attention has been devoted to constructing various application scenarios of joint lighting and VLC. Visible light sensing (VLS) has also intrigued increasing interests. The VLS system can directly utilize the illumination LEDs as the sensors [2], [3], and analyzes the response of the LED receiver to the ambient light, thereby inferring the relevant information of the environment.