Published Paper Details:

In this paper, we describe a navigation and control system for autonomous vehicles by using a traffic sign recognition module based on data received from a multi-sensor module disposed in the autonomous vehicle. A computing system of the autonomous vehicle can detect an event based on the traffic signs by retrieving data from the multi-sensor module. The multi-sensor module can recognize the traffic signs and process this information and provide it to the computing unit of the autonomous vehicle. The multi-sensor module may include image sensing unit, global positioning system, traffic detection and/or other sensing units. The traffic signs provide very valuable data about the road to make driving easy and safer. The traffic sign board specifically can be easily recognized by the multi-sensory module as well as humans because of their color and shape that are particularly different from natural environment. Further, it will be useful for the autonomous vehicle to use the traffic sign recognition system along with other sensing units to provide safe control and navigation of the autonomous vehicle. The proposed system first detects the traffic signs using color and shape sign boards and second allows the classification using a neural framework disposed in the computing unit of the autonomous vehicle. Further, the system also focuses on minimizing the rejection rate of the traffic sign boards that are not recognized because of non-standardized traffic sign in different jurisdictions. The system allows standardizing images of the sign boards and advantageously reduces the need for additional unit for standardizing the sign boards. Further, this system allows notifying, the user of the autonomous vehicle, for the traffic sign recognized by the computing system on the user interface of the autonomous vehicle.