Object Size Estimation from Digital Images Using ArUco Markers and OpenCV-Based Contour Processing
Journal of Electrical Engineering and Computer Science (JEECS) | E-ISSN : 3089-5952 · 04.07.2026 · 2 (2) · 60-66
DOI: 10.62379/jeecs.v2i2.44 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Electrical Engineering and Computer Science (JEECS) | E-ISSN : 3089-5952, online erschienen am 04.07.2026. Verfasst von Muhammad Ikhsan und 3 weiteren Autor:innen. Die Arbeit nennt 7 Literaturangaben.
Abstract (Originalsprache)
Object size estimation from digital images is an important application in computer vision that enables non-contact measurement for inspection, documentation, and automation tasks. However, most existing systems rely on ideal marker assumptions and limited platform deployment, while the effect of marker-size input error on measurement accuracy is rarely analyzed in a systematic manner. Therefore, improving both robustness and practical implementation remains an important research challenge. Purpose: This study aims to develop and evaluate a dual-platform object size estimation system based on ArUco markers and OpenCV contour processing, while analyzing the impact of marker-size input accuracy on measurement performance. Methods/Study Design/Approach: This research employed an experimental quantitative design. The proposed system integrates ArUco marker detection, grayscale conversion, Gaussian filtering, Canny edge detection, morphological processing, and contour extraction to estimate object dimensions. Experiments were conducted on synthetic data and real images of two physical objects with known ground-truth dimensions captured using a smartphone under controlled indoor lighting conditions. The system was implemented on both Python-based desktop and Java-based Android platforms. Performance evaluation was conducted using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Percentage Error (MPE). Results/Findings: The experimental results show that using corrected marker size significantly improves estimation accuracy. The system achieved MAE = 0.110 cm, RMSE = 0.123 cm, and MPE = 0.77%, compared to MAE = 2.045 cm, RMSE = 2.340 cm, and MPE = 11.98% under nominal marker settings, representing an improvement of more than 94% across all evaluation metrics. The results confirm that marker-size accuracy is the dominant factor influencing measurement precision. Novelty/Originality/Value: The novelty of this study lies in (i) the development of a dual-platform (desktop and Android) object measurement system, (ii) a systematic analysis of marker-size input error on measurement performance, and (iii) the integration of a main-object detection mode to handle textured objects with internal patterns. The findings provide practical insights for improving the reliability of mobile and desktop-based computer vision measurement systems.
Zitieren
APA
Ikhsan, M., Helmina, Ediansa, O., & Akbar, Z. (2026). Object Size Estimation from Digital Images Using ArUco Markers and OpenCV-Based Contour Processing. Journal of Electrical Engineering and Computer Science (JEECS) | E-ISSN : 3089-5952, 2(2), 60-66. https://doi.org/10.62379/jeecs.v2i2.44
Vancouver
Ikhsan M, Helmina, Ediansa O, Akbar Z. Object Size Estimation from Digital Images Using ArUco Markers and OpenCV-Based Contour Processing. Journal of Electrical Engineering and Computer Science (JEECS) | E-ISSN : 3089-5952. 2026;2(2):60-66. doi:10.62379/jeecs.v2i2.44
BibTeX
@article{Ikhsan2026,
author = {Ikhsan, Muhammad and Helmina and Ediansa, Oka and Akbar, Zulfikri},
title = {Object Size Estimation from Digital Images Using ArUco Markers and OpenCV-Based Contour Processing},
journal = {Journal of Electrical Engineering and Computer Science (JEECS) | E-ISSN : 3089-5952},
year = {2026},
volume = {2},
number = {2},
pages = {60--66},
doi = {10.62379/jeecs.v2i2.44}
}
EUVIMED