Proceedings of the MULTINOVA: First International Conference on Artificial Intelligence in Engineering, Healthcare and Sciences (ICAIEHS- 2025)

<Previous Article In Volume
Next Article In Volume>

T.A.R.S: Table Analysis via Regression and Signal Processing for Robust Table Structure Recognition

Authors
Siddharth Vinod Kaduskar1, *, Naveen Vaswani1
1Thadomal Shahani Engineering College, Bandra, Mumbai, 400050, India
*Corresponding author. Email: sidkaduskar@gmail.com
Corresponding Author
Siddharth Vinod Kaduskar
Available Online 7 October 2025.
DOI
10.2991/978-94-6463-852-3_4How to use a DOI?
Keywords
Table Structure Recognition (TSR); Computational Geometry; Signal Processing; Wavelet Transform; Skew Correction; Contour Regression; Document Digitization; Resource-Constrained Deployment; Attendance Sheet Processing
Abstract

Extracting tabular data from scanned documents remains a challenge in CPU-limited environments, despite advances in deep learning. Table structure recognition (TSR) faces challenges due to geometric distortions such as skewness, curvature, and noise from crumbled pages or poor printing. Although deep learning (DL) models excel in accuracy, they demand significant resources; rule-based methods are lightweight but fragile. We propose T.A.R.S., a lightweight TSR pipeline combining contour-based geometric regression and multiscale wavelet detection to robustly extract cells from skewed/noisy documents without GPU dependency. Experiments on curved/skewed tables show T.A.R.S. achieves 90.8% F1-score with 0.4–0.6s/image latency on CPUs, generalizing from bordered attendance sheets to broader usecases. The framework bridges computational efficiency and structural robustness, offering a practical solution for resource-constrained deployments.

Copyright
© 2025 The Author(s)
Open Access
Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

<Previous Article In Volume
Next Article In Volume>
Volume Title
Proceedings of the MULTINOVA: First International Conference on Artificial Intelligence in Engineering, Healthcare and Sciences (ICAIEHS- 2025)
Series
Advances in Intelligent Systems Research
Publication Date
7 October 2025
ISBN
978-94-6463-852-3
ISSN
1951-6851
DOI
10.2991/978-94-6463-852-3_4How to use a DOI?
Copyright
© 2025 The Author(s)
Open Access
Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Cite this article

risenwbib
TY  - CONF
AU  - Siddharth Vinod Kaduskar
AU  - Naveen Vaswani
PY  - 2025
DA  - 2025/10/07
TI  - T.A.R.S: Table Analysis via Regression and Signal Processing for Robust Table Structure Recognition
BT  - Proceedings of the MULTINOVA: First International Conference on Artificial Intelligence in Engineering, Healthcare and Sciences (ICAIEHS- 2025)
PB  - Atlantis Press
SP  - 44
EP  - 61
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-852-3_4
DO  - 10.2991/978-94-6463-852-3_4
ID  - Kaduskar2025
ER  -