The Research On Answer Sheets Smart-Identification Based On Image Processing Technology

More and more schools are now using the online marking.The traditional answer sheets recognition system is difficult to meet the exam requirements.Answer sheets identification technique based on image processing technology,simplicity and flexibility in use,lower costs.With the rapid development of electronic technology and the replacement of digital products,the quality of the picture of the answer sheets has been further improved.Using the answer sheets picture shoot by those devices is the current needs.And the answer sheets recognition method put forward higher requirements.Image processing is the key technology used in the answer sheets identification process,the main application to students after the exam information,identification and subjective images of objective questions such as segmentation work.Image processing technology has been applied to multiple areas of the image recognition,and very mature.This article is the research on smart identification of research answer sheets based on image processing theory and technology.It makes easier to answer sheets recognition system in use and flexibility.At present answer sheets variety from one category to another.In order to study the answer sheets recognition method,this thesis do a more comprehensive understanding and analysis based on answer sheet technology in recent years and the associated theory and algorithm,and discusses the answer sheets recognition application characteristic and difficulty.Firstly,the thesis do some research on image recognition application process,answer sheets type,answer sheets structure,answer sheets intelligent recognition of the need to address the problem,answer sheets identification process,image to image processing,denoising,thresholding and image graphics operation and so on.Secondly,the paper studies the segmentation and correction of the picture of the answer card,which uses the method of projection to segment the student test number and objective problem options,and corrects the image of the inclined or deformed answer card using the methods of Hough line detection,Fourier transform,edge detection and image geometric transformation.Thirdly,the thesis study the answer sheet information recognition methods in student test number,objective problem options and handwriting test number,using methods including the region of statistics,template matching and neural network convolution and so on.Finally,the thesis study the realization of answer sheets image processing system in application design and function,include answer sheets recognition system requirements analysis,problem identification system development environment and architecture,system architecture,software development using patterns,answer sheets and other relationships between entities and the realization of image processing functions.The content and the developed modules in this paper reach the requirement of practicability and stability,and enhance the adaptability to various examinations,making the system more convenient and flexible to use.