Empirical analysis of AdaBoost algorithms on license plate detection

AdaBoost algorithm is an effective license plate detection method in the field of license plate recognition technology. A through analysis of three boosting algorithms (namely Discrete, Real and Gentle AdaBoost) is presented for license plate detection, including the algorithm details and experiment comparisons. The experimental results show the Gentle AdaBoost algorithm obtains an overall better results in terms of high detection rate and low false positive rate than the discrete AdaBoost algorithm or real AdaBoost algorithm.