Cost-Effective Small-Scale Mark Recognition Technique in Evaluation of OMR Sheet

Abstract

In developing countries like Bangladesh, most records are still captured and maintained using paper forms. In every sphere of our life to digitize data, it can be challenging to aggregate, share and analyze the data collected from plain paper. In this research project, an automated system based on digital image processing is intended to develop so that student answer scripts of multiple choices can be evaluated instantly by an instructor. Main target of the paper is to create a program with such perfection that it can perform negative marking, providing grace in case the correct answer is not present in the answer sheet. If answer script contains multiple answer to a question then the system will be able to justify that pattern as well. The system has been developed successfully in MATLAB platform by examining more than thousand exam papers during study. The evaluated result shows that the program possesses the competence of accurately capturing and digitizing data from paper forms.