| With computer technology, network technology and digital image processing technology, the scope of the application expand continuly in these areas, but also impact on all aspects of the field of education examination. It shows a lot of shortcomings gradually in traditional manual methods of scoring .And a new approach will replace it. Online marking is birth under this complex background. It can reduce the subjectivity of a marking error, and realize a fair, justice principle.Online marking has been gradually used in large-scale selective examination by its own merits, but it has several problems, such as single form of objective questions recognition technology, high accurate printing and cutting of the answer sheet. In this paper, a new recognition technology is proposed, which takes advantage of convenience of OpenCV in VC++. It is able to recognize objective questions automatically according to the different forms of answer sheet without high accurate printing and cutting.At first, the paper introduces the concept of online marking simply, development of online marking at home and abroad and the concept of main functions of OpenCV. Then the paper gives a concrete exposition on work progress of online marking system . It has four steps: the first is collecting image information , the second is image pre-processing , the third is image segmentation, the last is online marking and image recognition .then the paper mainly tell the design method of the new technology and realization steps, as well as how to make use of MFC design of human-computer interface system. Then the paper gives a concrete exposition of the bar-code identification , another important part of online marking system. It detailedly tells the components of one-dimensional bar code EAN.13, basic encoding rules, validation methods and the identification principle in this chapter. In the end, the paper makes a conclusion of the project, and prospects for the development of the system in future. |
PDF Full Text Request