Sistem Pengecekan Lembar Jawaban Komputer Dengan Optical Mark Recognition (OMR) Berbasis Open Computer Vision Python

Abstract

The computer answer sheet checking system is no stranger to being used during exams at educational institutions or at a non-educational institution. Generally, the method used when checking computer answer sheets is to use the OMR (Optical Mark Recognition) method. Generally, OMR technology is expensive and makes OMR use limited for certain institution. An alternative that we can use is computer vision technology which we can use from several open source libraries and OpenCV is one of them. The author tries to build an alternative OMR LJK checking system using OpenCV in Python. In addition to building the LJK checking system, the author also conducted tests to determine the consistency of the accuracy of the values ​​generated by the system with several test scenarios. The test uses the way the results are compared with manual checking by humans. As a result, the LJK checking system is able to consistently produce accurate values ​​as long as the LJK conditions are not too tilted, not too bright or too dark for the lighting.