期刊名称：International Journal of Computer Science and Network
印刷版ISSN：2277-5420
出版年度：2012
卷号：1
期号：5
出版社：IJCSN publisher
摘要：In recent years, there has a lot of research on Indian license platerecognition, and many license plate recognition algorithms havebeen proposed and used. In this paper, a new license platerecognition approach is put forward based on the Radial BasisFunction Neural Networks (RBFNN). On the basis of sharingfeatures of a variety of license plates (LP), the vertical edge wasfirst detected by canny edge detector. Then, some approaches wereadopted to remove the invalid edge regarding the characteristics ofedge grayscale jump and edge density, so that the regions havingfeatures of LP were preserved. Next, for searching LP region weapply horizontal and vertical projections and mathematicalmorphology (MM) operation. Then, color-reversing judgment wasconducted by color analysis, and binarization was done based oncenter region in LP. After that, characters were segmented by meansof prior knowledge and connected components analysis, and applyRadial basis function neural network for character recognition. Withrich samples verified in dark hours and daytime under realconditions, the experiment indicates that it is feasible to adopt thisalgorithm in license plate recognition system (LPRS) to achieveaccuracy.
关键词：License Plate; Plate Recognition; Neural;Network; radial basis function (RBF).
