Computer Vision-Based Optical Mark Recognition for Automatic Survey Evaluation

Buleje C.Y.C.; Solis I.S.; Martínez W.X.L. · Lecture Notes in Networks and Systems (2026) · DOI: 10.1007/978-3-032-26217-2_2

Resumen

Optical Mark Recognition (OMR) is a key technology for survey automation, enabling the digitization and automated evaluation of surveys and examinations. This study presents a training-free, computer vision-based OMR system designed for cross-mark detection in educational surveys. The proposed approach complements traditional shaded-circle OMR methods, which are not suitable for cross-shaped marks due to their irregularity and variability. The system can be applied to surveys printed on conventional paper and scanned using standard devices, without requiring specialized hardware or predefined templates. Advanced image processing techniques are employed, including grayscale conversion, noise reduction, contour detection, and geometric filtering, combined with a pixel-density-based mark classification strategy to identify and validate responses. In addition, QR code detection is integrated for automatic form identification and traceability. The system was evaluated in an educational environment using more than 14,500 survey items at a resolution of 300 dpi, with three response options (Never, Sometimes, and Always) and an additional category for invalid responses. Experimental results demonstrate the effectiveness and reliability of the proposed approach, achieving an overall accuracy of 99% and F1-score values between 0.98 and 0.99. These results confirm that the proposed OMR system provides a practical, efficient, and scalable solution for survey automation in real-world educational settings. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

Instituciones: Universidad Nacional José María Arguedas