Multi-scaled license plate detection based on the label-moveable maximal MSER clique
In this paper, we consider a robust vehicle license plate detection problem for intelligent transportation systems in the presence of various illumination situations. We propose a robust and fast multi-scaled license plate detection and location algorithm, which exploits a Label-Moveable Maximal MSER clique. Specifically, first, we extract the candidate character regions using the Maximally Stable Extremal Region (MSER) features. Second, we divide each candidate character region into four types and extract the suspected initial node (the top-left character) based on its neighbor MSER distribution characteristic. Third, we label each candidate character region to accomplish license detection and location based on the detected suspected initial node and the corresponding label-moveable maximal MSER clique. The robust of license plate detection, the accuracy of character labeling for license location, and the improvement of calculation efficiency are evaluated via the real data.
KeywordsIntelligent transportation system License detection Maximally stable extremal region
- 8.Anakavej, T., Kawewong, A., Patanukhom, K.: License plate localization using MSERs and vehicle frontal mask localization using visual saliency for vehicle recognition. In: Fourth International Conference on Digital Information and Communication Technology and its Applications (DICTAP), pp. 31 (2014)Google Scholar
- 9.Wei, W., Jiang, Q.J., Zhou, X., Wan, W.Y.: Car license plate detection based on MSER. In: Proceedings of the International Conference on Consumer Electronics, Communication and Networks (CECNet), pp. 3973 (2011)Google Scholar
- 10.Li, B., Tian, B., Yao, Q.M., Wang, K.F.: A vehicle license plate recognition system based on analysis of maximally stable extremal regions. In: Proceedings of the 9th IEEE International Conference on Networking, Sensing and Control (ICNSC), pp. 399 (2012)Google Scholar