A Bimodal Image Dataset for Intelligent Rejoining of Bamboo and Wooden Slip Fragments

Due to long-term burial and environmental deterioration, excavated slips are often recovered in a fragmented state, posing significant challenges to the restoration and study of historical documents. To facilitate research on automated fragment reassembly, we present TailorJiandu, a dedicated multimodal dataset for bamboo and wooden slip fragment matching and retrieval.

The dataset addresses the lack of high-quality annotated resources in this domain and provides a standardized benchmark for developing intelligent reassembly methods. TailorJiandu systematically integrates visible-light and infrared images, enabling comprehensive characterization of fragment appearance, fracture patterns, ink traces, and surface textures. The complementary information provided by the two imaging modalities offers valuable support for fragment matching and retrieval tasks.