Japanese road signs recognition using neural networks

Recently, traffic accidents are increasing because of driver's inattentiveness to the road signs. In this work, in order to assist drivers, we propose a recognition system for road signs using neural networks. We first segment road sign regions using color and shape features to get the location information. The initial sign recognition uses template matching. However, since this did not produce satisfactory results, a neural network is used to learn and recognize the signs. A variety of features are considered for training the neural network. The experimental results show that this recognition method is effective for online road sign recognition.