Accelerating the discovery of new Nasca geoglyphs using deep learning


World Heritage Site: Lines and Geoglyphs of Nasca and Palpa

Publication Year: 2023

Publication Type: Article

Publication Identifier: WHE3F16CA1AEC

Summary

Researchers have successfully uncovered four previously unknown Nasca geoglyphs—depicting a humanoid, legs, fish, and bird—using deep learning (DL) techniques on high-resolution aerial imagery. This breakthrough, verified through on-site surveys, demonstrates that DL can accelerate geoglyph discovery by approximately 21 times compared to manual inspection alone. The study addresses the challenges posed by limited ground truth data and variability in geoglyph design, offering a novel pipeline for integrating AI with archaeological fieldwork. These findings not only expand our knowledge of the Nasca Pampa UNESCO World Heritage Site but also establish a new paradigm for combining artificial intelligence with traditional survey methods.

Citation

Sakai, M., Lai, Y., Olano Canales, J., Hayashi, M., & Nomura, K. (2023). Accelerating the discovery of new Nasca geoglyphs using deep learning. Journal of Archaeological Science, 155, 105777. https://doi.org/10.1016/j.jas.2023.105777

You are here:
World Heritage Explorer > World Heritage Research > Publications > Accelerating the discovery of new Nasca geoglyphs using deep learning

This website includes data sources licensed under CC BY-SA 4.0. Additional original content by World Heritage Explorer, also licensed under CC BY-SA 4.0. WHE is not affiliated with UNESCO or the World Heritage Committee. Legal Notice. Privacy Policy.

Open Data for an Open World