ViLDAR: a novel speed estimation system using visible light in vehicles

Abstract

In this paper, we introduce Visible Light Detection and Ranging (ViLDAR): a novel speed estimation system that utilizes visible light variation during a known time difference. The use of light emitting diodes (LED) in vehicles lighting systems has become increasingly common in the automotive industry. Concurrently, monitoring drivers' speeds constitutes a critical issue for safety regulation agencies. Most of the speed monitoring systems are based on measuring the speed of vehicles using wellknown RADAR or LiDAR systems that use radio frequency or laser signals, respectively. However, these systems have certain limitations, e.g., the requirement of line-of-sight for accurate results in RADAR and LiDAR, and narrow beam-width and deficient performance in curved roads in RADAR. We present a simple as well as novel estimation method for the speed of an approaching vehicle given only the received light intensity information. Assuming a Lambertian channel model for the light wave propagation, the performance of the estimator is compared in different angle of incidences between the vehicle and the detector. The proposed method outperforms RADAR/LiDAR systems in terms of speed estimation accuracy for a wide range of incidence angles.