Abstract
License plate localization algorithms aim to detect license plates within the scene. In this paper, a new algorithm is discussed where the necessary conditions are imposed into the saliency detection equations. Measures of distance between probability distributions such as negentropy finds the candidate license plates in the image and the Bayesian methodology exploits the a priori information to estimate the highest probability for each candidate. The proposed algorithm has been tested for three datasets, consisting of gray-scale and color images. A detection accuracy of 96% and an average execution time of 80 ms for the first dataset are the marked outcomes. The proposed method outperforms most of the state-of-the-art techniques and it is suitable to use in real-time ALPR applications.