Description

The dataset is formed by 10 original images of Homer papyri and their corresponding Ground Truth for binarization built manually by D-Scribes project members. The papyri were chosen to reflect the diversity of literary papyri, therefore they come from various places and periods. In addition, they have different states of preservation and restoration. The images are provided by the institutions that own the papyri and are thus also heterogeneous.

I. Pratikakis, K. Zagoris, X. Karagiannis, L. Tsochatzidis, T. Mondal and I. Marthot-Santaniello, "ICDAR 2019 Competition on Document Image Binarization (DIBCO 2019)," 2019 International Conference on Document Analysis and Recognition (ICDAR), Sydney, Australia, 2019, pp. 1547-1556. Access here.

On DIBCO 2019, see https://vc.ee.duth.gr/dibco2019/ 

Dataset available here.