Project representative:
Sanae Takano(Researcher/The Kyoto University Museum)

Collaborating researchers and co-researchers:
・Hideyuki Uesugi (Researcher/Nara National Research Institute for Cultural Properties)
・Junji Sugiyama(Professor/Graduate School of Agriculture/Kyoto University)

Introducing this project

Research on the Jomon period in the Japanese archipelago has been based on the remains and artifacts left behind by the Jomon people. Jomon pottery, in particular, is part of the time scale for the 13,000-year Jomon period, and research into production techniques and usage methods can help us understand the interactions between people. Jomon pottery, which was made by hand, is not the same as any other piece, and its shapes and patterns are complex and diverse. Researchers have used traditional measuring tools to figure the diverse Jomon pottery, pasting paper on the surface of the pottery and applying ink to it to make rubbings. However, traditional methods of making figures have not been able to fully express the characteristics that allow us to consider human movements, such as the texture of the clay used and the color of the fired pottery, as well as the fine decorations and cord marks on the surface of the pottery. The materials that can be figured are also subject to time constraints and space restrictions, and important materials are not made public and are not used in research, which is also an obstacle to academic development. In this study, we have introduced the digital technology of optical rubbing, which can generate high-resolution two-dimensional images of minute irregularities, to visualize the detailed structure of the cord marks and traces of the cord material. We are also attempting to classify images using machine learning on the optical rubbing images, and have begun work on creating an artificial intelligence (AI) that can automatically identify hundreds of different cord marks.