Design And Implementation Of Intelligent Scoring System Based On Mobile Platform

In recent years,computer image recognition technology has been developed rapidly and has been applied in many fields.Mobile phone,as popular equipment,has gradual become an indispensable good teacher and helpful friend in learning and working.In the past,many works rely on computers to complete,are now gradually replaced by mobile phones.The examination is an indispensable means of education,and always need to do a marking after having a test,based on the full understanding of the school's daily teaching and teachers' current marking methods,to improve teaching quality and reduce the burden on teachers,developed an intelligent marking system based on Mobile platform suitable for the majority of teachers.The working mode of the marking system is to use the camera as the input device,automatically identify the answer sheet,and then read the information on the answer sheet,then upload the information to the server for processing,the server compares the answers?summary,stored in the database,generate reports for teachers to print and access.The information collection process in this system is: open the phone camera,the system capture real-time image,and then grayscale processing,determine the position of the answer sheet and its tilt rotation in the image,so as to correct the image and intercept the seeming answer sheet for normalization,and then according to the standard answer sheet template feature for micro matching is not the real answer sheet.When it is determined that was the real answer sheet,and then refers to the standard answer sheet template to identify the information area to get the answer.In this paper,point at the characteristics of the mobile platform marking,some related technology research to be done,focusing on solving the following problems:(1)How to determine whether there is an answer sheet inside the image;(2)How to correct distorted images;(3)How to separate the answer sheet from the background in a complex background;(4)Answer sheet template production;(5)Identification of answer sheet information.The system can be divided into two stages: rough matching and micro matching,the SIFT algorithm is applied to the rough matching stage,solved the problem of image positioning under non stationary camera,the method of geometric moment and automatic threshold method was used in the micro matching stage,it can not only improve the matching efficiency,but also improve the matching accuracy and intelligence.An automatic feature extraction method is proposed in the design phase of the answer sheet template,it reduced manual labor.This system is developed under the environment of Qt,finally ported to the mobile platform,the marking cost is greatly reduced,and it is very convenient to carry and use,can significantly increase the flexibility and improve work efficiency,the system can also be used as a survey system.Practice has proved that the system has the value to promote the use of society.