Traffic sign recognition model on mobile device

Abstract

Incorporation of a good machine vision into mobile device can create a powerful application. With such advancement, mobile users can use their existing smartphone as their third-eye to perceive the world. This paper introduces a framework using the smartphone to increase ambient intelligence and road safety of moving vehicles through traffic sign alert application. The driver is alerted of the incoming traffic signs in different modes depending on the user's preference. However, the competency of the application is still bounded with the capability of a normal human vision. Unlike conventional research work, this paper emphasizes on portability and expandability. Portability means there is no additional installation or placement of electronic hardware required in the vehicle in order to make the system work. Expandability focuses on the transparency of the traffic sign recognition output to other applications within same hardware device or external devices. This provides an option to other developers to use or expand their own applications by using the result of the proposed application. An interaction model of the traffic sign recognition system on mobile device is introduced.