Economical and Efficient Multiple-Choice Question Grading System using Image Processing Technique

Autor: Rangkuti, Abdul Haris, Athalia, Zefany, Thalia, Tara E., Wiharja, Caitlyn E., Rakhmansyah, Abid
Jazyk: angličtina
Rok vydání: 2023
Předmět:
Zdroj: International Journal of Intelligent Systems and Applications in Engineering; Vol. 11 No. 3 (2023): (Articles in Press); 193-198
ISSN: 2147-6799
Popis: A common exam format that has gained popularity, especially in paper-based exams, is the multiple-choice question (MCQ), which is simple to use and quick to grade. MCQs are typically graded using a specialized apparatus that is expensive and inaccessible for the majority of teachers. The goal of this project is to create an image processing-based alternative that is more affordable and effective. The developed system makes use of a web camera to automatically display the students' grade and feedback analysis by capturing the answer key and the answers of the students. The Canny edge detection method was selected for analyzing the position of the answer on each question after system testing and method comparison. By using their phone or webcam, users would be able to grade exams effectively without the hassle of needing a specialized trained operator or machine. This study attempted to further develop previous studies by utilizing a video feed rather than an image feed to improve usability. This study concluded with a successful development of a reliable, effective, and affordable camera-based multiple-choice question grading system which has an accuracy of 94%. This research can be developed to facilitate multiple-choice questions with different formats.
Databáze: OpenAIRE