Archaeological site identification on aerial imagery using deep learning: ODYSSEY project

• Autor(a) principal: Maia, Leonardo dos Santos
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/41030
• topic: Archaeology; Deep learning; LiDAR; Unet; YOLOv7

This dissertation was developed within the ODYSEY project, which aims to develop a platform intended for archaeologists. Within this context, this dissertation aims to identify archaeological sites from images formed through the data provided by a LiDAR system. The study area is Alto Minho, a Portuguese sub-region belonging to the Northern region, and famous for the preservation of historical structures. This work focuses on the study of tumuli, which are buildings of stone and sand that would have the function of hiding and protecting graves, and the hillforts, which are urban constructions of the Copper Age and Iron Age. In a clear way, the goal is to elaborate a system capable of locating these historical objects from an aerial image. The work ranges from the creation of the database, the implementation of deep learning models, to the inference of the results.