Real-time Bubble Sheet Detection and Evaluation Using Object Recognition

Abstract

Optical Mark Recognition (OMR) is a technology widely used in education to digitize large amounts of paper data. It is often used for scoring multiple-choice tests. This can be time-consuming when manually grading. OMR technology facilitates this process by providing a controlled format for students to submit their answers. Without the need for manual scoring, OMR is a well-known data entry method and an important technique in human-computer interaction. It is widely used in assessment tests in universities, colleges, questionnaire forms, and competitive examinations. There are many applications for computer image processing and recognition in today's technology-driven world. Existing OMR techniques have limitations, such as the need for special paper, reliance on high quality scanning to achieve accurate results. and to overcome these challenges this research introduces a new image-based OMR method. This reduces the need for special paper and increasing the detection accuracy of the proposed system which is highly reliable by using advanced object recognition techniques that can detect partially filled bubbles with some Marks are slightly blurry or blurred. Additionally, the research includes an automated questionnaire creation system based on specific needs. The generated question paper comes with an answer key. The system will evaluate the completion of the bubble sheet and create the final output file in Excel format. This technique allows for cost-effective optimization of question papers and guarantee effective evaluation in standard reports. Tests have shown the system's accuracy to be more than 99%, effectively scanning and analyzing filled, partially filled bubbles.