Research On Military Vehicle License Plate Recognition System

With the deepening of the reform of national defense and the army,the new organizational system has begun to operate,and a large number of new vehicles and equipment are installed in the army,which puts forward new challenges and requirements for the management of the army in the new era,especially for the management of military vehicles.Military vehicle license plate is the valid identity certificate of military vehicle.The realization and application of military vehicle license plate recognition system technology has brought good news to military construction,especially the informatization construction of the army,and has become an important part of the intelligent barracks.By identifying the license plates of military vehicles and monitoring the operation of military vehicles,the effective management and control of military vehicles can be achieved.In this paper,the realization and application of key technologies of license plate recognizing technology are studied,and the related theories of license plate recognition system are explained.On this basis,the concrete measures for the realization and application of the key technologies of license plate recognition system are put forward.Hopefully,the research on this aspect will be carried out.It contributes a little to the development and application of the key technologies of license plate recognition system.In this paper,a military vehicle license plate recognition system based on image processing technology is studied.Firstly,from the technical theory point of view,according to the four steps of image preprocessing,location,segmentation and recognition,the related technical theory is introduced,and the character characteristics of military vehicle license plate are analyzed.The mathematical morphology principle of the image,Roberts algorithm,Sobel algorithm and Canny algorithm used in edge detection are introduced in detail.Temp late sample matching method,analysis feature matching method and BP neural network recognition method are used in character recognition.The simulation results of corresponding MATLAB software are given.A picture of the fruit.Secondly,from the perspective of practice test,the implementation and application of the key technologies of license plate recognition are studied.In the part of license plate pretreatment,image grayscale and image noise removal are introduced.The technologies of image processing such as grayscale image,image information enhancement,noise removal filtering and binarization of Otsu algorithm are mainly studied.The image pretreatment of military vehicle license plate is carried out.In the part of license plate location,two methods,rough location and precise location,are introduced.The two methods,which are based on mathematical morphology and color space,are mainly studied.In the vehicle license plate tilt correction technology,because the intercepted vehicle image may be affected by the angle of photography or the slope of road surface,the obtained license plate image has a certain angle,which will affect the segmentation of license plate characters.The processing technology of Hough transform,rotation projection and so on are emphatically studied.In the part of license plate segmentation,we focus on the license plate frame and screw removal,license plate character segmentation,character image unification.In the part of license plate recognition,BP neural network is used to train the character template library to get the learning model,then the model is used to recognize the characters and output the recognition results.Finally,from the perspective of system integration,the design and Simulation of intellige nt license plate recognition system for military vehicles are studied.In this paper,MATLAB R2013 A programming is used to realize the verification of image recognition algorithm based on BP neural network.Through the above analysis and operation,the abo ve algorithm is integrated into an executable operation interface and edited by selecting the GUI(Graphic user interface)interface that general software will use.This graphical-based interface is more visually acceptable and easy to operate for users.The final result shows that the recognition accuracy of the military vehicle license plate recognition system studied in this paper is relatively high,which shows that the military vehicle license plate character recognition based on MATLAB image processing technology can achieve its effectiveness.This paper studies the BP neural network algorithm for license plate recognition.If the algorithm can be applied reasonably,the recognition speed and efficiency can be greatly improved.Moreover,through the study of this paper,it is found that the method of character recognition based on template BP neural network can recognize the license plate information quickly and accurately in complex background.It can satisfy the real-time requirement of military vehicle license plate recognition system,which is also the greatest advantage of this algorithm.