| As an important component of the intelligent transportation,License plate recognition technology is playing a more and more role nowadays.License Plate Recognition System through the automatic recognition of license plates realizes the standardization and Scientific of Vehicle Management.So it has a Wide range of applications prospect.License Plate Recognition system mainly consists of license plate location,character segmentation and character recognition of three parts.In this paper,Conducted in-depth study of License Plate Recognition system,solve the following three problems on the basis of summing up experiences and lessons of our predecessors:1.The existing license plate location based on mathematical morphology algorithms,subject to the constraints of selected structural elements,self-adaptive is not high.In response to this situation,this paper proposed an improved algorithm for selecting Structuring element.The algorithm by calculating the image,selecting the appropriate structuring element.In this way,can improve the self-adaptive of morphological algorithm.2.At present,the process of identifying license plate connected domains does not consider the tilt of license plate,and this can easily lead to erroneous results.In response to this situation,this paper proposed An improved algorithm for identifying connected domain.This algorithm Take full account of the tilt of plate,improve the accuracy of the license plate.3.In this paper,based on the experience of predecessors,proposed a new binarization method of license plate.This algorithm has lower complexity and better real-time,could reduce the interference caused by adhesion due to character,anchors and other issues,to increase the success rate of follow-up work.In order to verify the algorithm proposed in this paper,I have multiple pictures of the simulation results.The results show that the algorithm proposed in this paper improved the recognition accuracy,and robust high,could fully meet the needs of vehicle identification. |
PDF Full Text Request