Abstract:
Multiple choice questions (MCQs) represent a frequently used evaluation method that is employed
by a number of educational institutions, testing services and other organizations. The major
attraction in these evaluations is the ease with which the answer sheets can be scored without human
intervention. The filled answer sheets are automatically scored by specialized scanners equipped
with the Optical Mark Recognition (OMR) software. In most cases, the hardware costs of these
scanners may not be affordable by small organizations, institutes or individuals. Inspired by the
availability of camera in almost all the mobile devices available today, we developedan Android
based Optical Mark Reader (OMR) system.
The developed system relies on image analysis and pattern classification techniques to implement
optical mark recognition and automatically score the answer sheets. An image of the filled answer
sheet is provided to the system. The image is binarized and a set of morphological image processing
operations is applied to extract the individual lines representing a single question. Filled choices are
then identified for each of the questions and responses are compiled. A user may provide an image
of the true solution and validate the processing that automatically converts the responses from
image to text. Later, the images of the filled answer sheets of the candidates are provided to the
system which automatically processes these sheets and compares the responses of the candidates
with the true responses. The overall score of the candidate is determined and is presented to the user
for validation. The developed system evaluated on images of MCQ based answer sheets realized
promising results. It is expected that the developed software could be extended further into a
comprehensive application that could eventually be used by instructors for automatic marking of
MCQ based evaluations.
