Abstract
Detection of vehicle license plate is very important in vehicle license plate recognition system. In this paper, a novel classifier fusion-based detection algorithm is proposed. After locating candidate license plate regions, features of these regions are extracted for the optimal feature subset by exhaustive search strategy. Based on the classifier fusion theory, Simple Average (SA) method is compared with two Weighted Average (WA) methods. Experimental results show that after reducing the dependency of features, and SA method works better. The three most important features of the license plate regions are obtained in the experiment and our algorithm can be applied in real-time applications and robust in filtering out false plate regions.