Information Retrieval From Damage OMR Sheet by Using Image Processing

In this dynamic era, people want to save his or her time as much as possible. They sometimes want the result just after the experiment. The motive of this paper is to scan the OMR sheets and collect data by image processing and matching the current data with the previous one to find out the latest or current results. 1. In today's world many exams are being conducted which offer Multiple Choice Questions to the students. These exams offer Multiple Choice Questions(MCQs) to the examines. Generally in examinations with MCQs, the OMR sheet is used. 2. There are many significant usages of this technology in different sectors such as institutional research, community and consumer surveys, data compilation, product evaluation and so on. It detects the specific pencil marked round bubble from the optical answer sheet of the multiple-choice questions. As it is a manual process, so many errors may occur here. In this paper, such a technique is invented which will overcome the limitations of the OMR machine. This system also works for the damaged sheets which the OMR machine does not at all. By scanning the hard copy of the OMR sheet and created a JPEG format database and a data set is used as the input. Different types of errors and image problems are solved to complete the process. Here the system is build up in MatLab and python language by image processing.

I. Introduction

OMR technology has been increased vastly within the past few years. It is used in educational institutions such as schools, colleges, universities. It is also a very popular scheme in survey works. This system is mostly used in the exams as it is very easy to demonstrate and strongly secured and obviously cheaper. This sheet can be designed according to one’s desire and this is the main reason for this technology to be this much popular. The sheet is designed according to the exam or the data collection type. But this OMR system requires special paper, ink, special input reader and environment. Besides, if there occurs any stain or scratches in the paper or if it gets folded, the OMR machine can not perform the operation on those papers. Generally, these paper, ink, and reader are very costly. Besides managing these components is also a tough job to do. Basically, It consumes too much time to check the kind of the paper, ink, used by the OMR sheet reader. OMR makes the data collection process very complicated, if the user wants to gathers large number of texts. There are also other complications like missing data, incorrect ordering due to unnumbered pages. If there is no safeguard in the first place, a page might het rescheduled, providing duplicate and skewed data.