Research On Electoral Statistics System Based On The Technology Of Symbol Recognition

With the rapid development of information society in our country today, the traditional way of statistics in the election has been far from able to meet the needs of society. Previous statistics methods play a significant role in a variety of democratic elections, such as tellers approach based on artificial, electronic elections and based on machine identification even intelligent ballot boxes way. But a variety of ways still has a lot of statistics shortcomings. Artificial tellers carry through manual recording and statistical ballot, however, this approach is not only time-consuming, inefficient and error-prone, but also because of lax supervision,it is difficult to ensure fair democratic of election, therefore, these influence the progress and quality of democratic elections. Because of these technical shortcomings of e-election, its security and reliability have been disputed by the people. Apart from this, the ticket stub of election is not available after the meeting and the form of ballot is single. It has made the promotion and use of electronic voting large limits.The statistics way of Machine identification and intelligent ballot boxes is electoral statistics system of social use, which combines modern computer resources and OMR technology. It is a new real-time electoral statistics system that uses image processing and algorithm of information Card information recognition to achieve the mark of recognition. The system can carry out real-time traffic information, signal control and character recognition, statistics on the ballot, etc. But the system uses the ballot form with Tiantu card ballots, which has been greatly restricted in the promotion and use of the system.The design of" The research of electoral statistics system based on character recognition technology " solve this problem and the system can identify handwritten symbols of voters, such as "√,╳,○,●,╲,╱" and so on, meeting the people who fill out the habit of votes, which has a strong practical.The system design of the "electoral statistics system based on character recognition technology" make available smart boxes as ballot boxes, that has improved image acquisition system and the character recognition system on the ballot box. The system identified star network structure by a number of symbolic devices (smart Boxes), the host and input terminal. It uses different techniques from optical mark readers using character recognition and the algorithm method to complete a variety of hand-written notation on the ballot paper identification. Firstly, applying CMOS camera make the entire image of the vote as input, then algorithm positioning, segmentation, and then locate and cut up using algorithm. The symbol recognition system identified handwritten symbols, "√,╳,○,●,╲,╱" etc. If the handwritten symbols information is in favor of another people or fuzzy or defaced recognition of can not be handled, they will be processed by identify by hand, which are associated with small images and transmitted through the network communication module to the input terminals, displayed on the computer screen, therefore, it improve the character recognition accuracy. Finally, the identification results of the various subsystems issued to the host through the network, the host statistics obtained the election results.The system includes automatic generation votes, handwritten character recognition in the ballot box, implement in favor of another, voting and statistics each subsystem by the host and realization of network architecture of the entire electoral statistics system. In addition, how to test the network host circuit and how to control each character recognition equipment and work status of input terminals are the tasks of the system.