Emptyscapes initiative Mapping under canopy past landscapes by high resolution dronebased LiDAR

>> YOUR LINK HERE: ___ http://youtube.com/watch?v=RJe8qA8yAWw

#aerialarchaeology #archaeology #aarg #LiDAR #deeplearning #Tuscany • #italy • From the Aerial Archaeology Research Group 2024 Annual Meeting • https://aargonline.com/wp/events/aarg... • SESSION 5: TECHNICAL DEVELOPMENTS • Emptyscapes initiative: Mapping under canopy past landscapes by high resolution drone-based LiDAR by Giuseppe Prospero Cirigliano, Gabriele Mazzacca, Fabio Remondino, Herbert Maschner and Stefano Campana • This paper presents the results of Emptyscapes an ongoing research project which, in the past two decades, handles a wide variety of datasets collected with multidisciplinary methodologies integrating different survey techniques and targeted/open area archaeological excavation. This contribution aims to show a methodology and workflow applicable in woodland Mediterranean areas aimed to identify and map under canopy archaeology. The study area is a hill chain characterised by dense vegetation typical of the Mediterranean scrub, which is composed of both high and low trees, and located in the southern part of Tuscany (Italy), within the territory of the ancient city of Roselle located along the Tyrrhenian seaside next to the modern city of Grosseto. In February 2021, a Lidar acquisition was conducted over an area of 550 hectares, with an average of ca 600 points/m². The area, which shows an abundance of archaeological evidence from various historical periods, allowed for a diachronic approach to identifying archaeological features. Specifically, the combination of aerial drones and Lidar sensors enabled us to obtain high-resolution and large-quantity data, requiring significant processing efforts facilitated by collaboration among various expertise in different fields, such as archaeologists, computer scientists, and geomatics. The challenges associated with large datasets and semantic segmentation of point clouds have been addressed through the application of deep learning by enabling the development of an efficient workflow. This has yielded excellent results that add crucial elements to the reconstruction of the historical narrative of the ancient landscape of Roselle/Tuscany. • Video by ‪@openpast‬

#############################

New on site