Research On The Key Technologies Of Automatic License Plate Recognition Application Platform Based On SaaS

License plate,as the only visible sign of vehicle,is the most important problem of intelligent transportation.At present,license plate recognition has been widely used in road traffic monitoring,intelligent parking management,freeway automatic toll collection,residential intelligent traffic control system and so on.License plate recognition on the market application products emerge in endlessly,the product quality also has its own merits.At present,the main problems of license plate recognition application system are: the accuracy of license plate recognition is difficult to guarantee,the cost of development and operation and maintenance of all kinds of license plate recognition application system is very high,and the function and equipment of various license plate recognition application system are built repeatedly.The data of the license plate recognition application system of all kinds of systems is in the state of information isolated island,which is difficult to be mined and reused.In order to solve the above problems,this paper studies the following contents:In this paper,two key technologies of license plate recognition,namely,license plate region location based on mixed strategy edge detection and defiled license plate character recognition based on BP neural network,are studied.In order to improve the accuracy of license plate recognition algorithm in general environment;Then,this paper establishes a unified license plate recognition service framework based on SOA,which provides uniform license plate recognition service for all kinds of application systems involved in license plate recognition on various platforms,so as to integrate the algorithm and development difficulty of license plate recognition in the whole industry.Improve the accuracy of license plate recognition in the whole industry;Finally,this research establishes the license plate recognition application platform based on SaaS pattern,modularizes and customizes the business function of all kinds of license plate recognition application systems,and establishes a unified running platform.It can greatly reduce the development and running cost of various license plate recognition application systems,improve the development efficiency and product quality of all kinds of license plate recognition application systems,and also provide support for further data sharing and big data analysis.The research of this topic will effectively reduce the development difficulty of all kinds of license plate recognition applications,effectively promote the rapid deployment of various license plate recognition applications,and effectively promote the application depth of intelligent transportation in a larger area.