Principal component analysis for speed limit Traffic Sign Recognition

Traffic Sign Recognition has recently become a popular research field. The main applications include Autonomous Driving Systems, Driver Assistance Systems, road sign inventory, to name a few. In this paper, a speed limit traffic sign recognition system is proposed based on Principal Component Analysis (PCA) with preprocessing steps, that help on perspective and extreme luminance variation correction. The classification is performed by a feed-forward neural network, or Multi-Layer Perceptron (MLP). Experimental results present a classification accuracy similar to some state of the art systems, but with a more compact scheme.