TY - GEN
T1 - Detecting Spiral Text Lines in Aramaic Incantation Bowls
AU - Nammneh, Said
AU - Madi, Boraq
AU - Atamni, Nour
AU - Boardman, Shoshana
AU - Vasyutinsky-Shapira, Daria
AU - Rabaev, Irina
AU - Saabni, Raid
AU - El-Sana, Jihad
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025/1/1
Y1 - 2025/1/1
N2 - Due to irregular spacing, character overlap, and varying orientations, text line detection is especially challenging for unconstrained handwritten and historical documents. These complexities make traditional methods, designed for straight lines, insufficient for detecting spiral text lines from Aramaic incantation bowls used in Sasanian Mesopotamia between the 4th and 7th centuries CE. We introduce a novel learning-based method for extracting spiral text lines inscribed on the surfaces of Aramaic incantation bowls. Our model utilizes an encoder-decoder architecture while leveraging connections among corresponding layers, similar to UNet. It combines high-level and low-level features, enabling precise localization and segmenting spiral text lines. Furthermore, we propose a novel metric to evaluate the dissimilarity between predicted and ground-truth lines. Inspired by the Intersection-over-Union (IoU) metric, We compare our model with the state-of-the-art methods and show that our approach outperforms these methods in terms of the introduced metric.
AB - Due to irregular spacing, character overlap, and varying orientations, text line detection is especially challenging for unconstrained handwritten and historical documents. These complexities make traditional methods, designed for straight lines, insufficient for detecting spiral text lines from Aramaic incantation bowls used in Sasanian Mesopotamia between the 4th and 7th centuries CE. We introduce a novel learning-based method for extracting spiral text lines inscribed on the surfaces of Aramaic incantation bowls. Our model utilizes an encoder-decoder architecture while leveraging connections among corresponding layers, similar to UNet. It combines high-level and low-level features, enabling precise localization and segmenting spiral text lines. Furthermore, we propose a novel metric to evaluate the dissimilarity between predicted and ground-truth lines. Inspired by the Intersection-over-Union (IoU) metric, We compare our model with the state-of-the-art methods and show that our approach outperforms these methods in terms of the introduced metric.
KW - Archaeological Images
KW - Incantation Bowls
KW - Spiral Text Lines
KW - Text Line Detection
KW - UNet
UR - http://www.scopus.com/inward/record.url?scp=85212266262&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-78495-8_16
DO - 10.1007/978-3-031-78495-8_16
M3 - Conference contribution
AN - SCOPUS:85212266262
SN - 9783031784941
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 250
EP - 264
BT - Pattern Recognition - 27th International Conference, ICPR 2024, Proceedings
A2 - Antonacopoulos, Apostolos
A2 - Chaudhuri, Subhasis
A2 - Chellappa, Rama
A2 - Liu, Cheng-Lin
A2 - Bhattacharya, Saumik
A2 - Pal, Umapada
PB - Springer Science and Business Media Deutschland GmbH
T2 - 27th International Conference on Pattern Recognition, ICPR 2024
Y2 - 1 December 2024 through 5 December 2024
ER -