A real time license plate detection system based on boosting learning algorithm

Boosting is one of the most well-known and effective techniques in machine learning. The success of using boosting for training a face detector [28] has paved the way of using boosting for training object detectors and made it widely used in computer vision. In this work we present a new framework for fast and automatic detection of vehicle license plate based on boosting learning algorithm. Beside the traditional Haar-like features, we propose to use local binary pattern (LBP) feature for its robust and discriminative power. The boosting classifiers are trained on these features and then combined in an efficient way to achieve high performance. An intensive set of experiments have been conducted. The results show that the classifier with LBP outperform that of Haar-like in the same scenario for the license plate detection problem. By combining them in an reasonable way, our proposed system can perform in real time for detection of license plates with the accuracy up to 100%, outperform state-of-the-art approaches.