The Research On Licence Plate Recognition Algorithm Base On Neural Network

With the rapid development of China's economy,the number of automobiles has been increasing year by year.The intelligent traffic management system plays an important role in traffic control,and the license plate recognition is the core of the intelligent traffic management system.In this dissertation,we study the license plate recognition algorithm,applying convolutional neural network LeNet-5 to the license plate character recognition.And we improved the license plate character recognition algorithm especially.License plate recognition processing steps are divided into three steps,license plate location,license plate character segmentation and character recognition.Firstly,the image of the license plate is pre-processed by gray-scale and smoothing filter,and then the license plate localization is implemented based on edge detection and mathematical morphology algorithm.Secondly,the vertical projection algorithm is improved based on the specific characteristics of the license plate to segment the license plate characters.The segmented license plate characters are used as the data set for the next step of character recognition.Finally,on the basis of in-depth research on the working principle and advantages of CNN,the network with the characteristics of weight sharing and local receptive field is applied to license plate character recognition.Based on the classical LeNet-5 network structure,the structure of the input layer,the hidden layer and the output layer is improved,and the parameters of the convolutional layer and the lower sampling layer are optimized to construct the DLeNet-5 network.Compared with the original network,the improved network is applied to the license plate character recognition with a higher recognition rate.License plate recognition not only plays an indispensable role in the public safety management system and various traffic occasions,but also will play an important role in the fields of unmanned driving and driverless cars in the future,which will bring a lot of convenience for the healthy development of cities.