Traffic Sign Recognition using Multi-class morphological detection

In this paper, a novel method is proposed for the Traffic Sign Recognition (TSR) using the Principle Component Analysis (PCA) and the Multi-Layer Perceptrons (MLPs) network. In particular, the candidate signs are individually detected from two chroma components in the YCbCr space and then classified into three shape classes: circle, square, and triangle based on computing the rotated version correlations. The PCA-based features of these objects will be used for the MLPs as the training system corresponding to previously determined class. This approach not only reduces the time but also increases the performance of the recognition process. In simulation, the proposed method is estimated with over 500 statistic images for the accuracy rate up to 96%.