An Edge Computing System for Estimating Road Surface Condition on Winter Expressway

This paper proposes an edge computing system for estimation of road surface condition on winter expressway by utilizing in-vehicle camera images. In the field of transportation engineering, the realization of image processing-based methods for road scene analysis has been expected. Therefore, many images obtained by in-vehicle cameras are accumulated in various organizations of transportation engineering. However, since analyzing these accumulated images requires too much time and huge-sized computer, immediate analysis is difficult. Thus, this paper constructs an edge computing system that realizes immediate analysis of road surface condition by utilizing images obtained from in-vehicle camera. The proposed system is constructed on a small-sized computer. Also, an algorithm installed on this system estimates winter road surface condition via Mobile Nets using in-vehicle camera images.