Automated analysis of Slavic buried settlements based on LiDAR point cloud, UAV and image processing

• The paper proposes a new methodology for the automated analysis of geometric features of fortified buried settlements. • The proposed methodology allows for the detection of archaeologically unidentified settlements using LiDAR data. • High automation degree of the detection of settlements is achieved on the real objects’ datasets acquired from different levels and data types. • The research demonstrates that LiDAR may be applied in the detection of settlements, even in forested areas. In this paper, a methodology for automated analysis of settlements structures is proposed for research in cultural heritage. The proposed automatic, accurate and efficient method for detecting fortified settlements is based on laser scanning data obtained from Airborne and UAV laser scanning systems. Firstly, the method of automated detection of settlement structures based on the Digital Elevation Model generated from point clouds that were obtained from various altitudes is discussed. The index that determined the extraction of settlements was the shape factor determined for the detected objects. This index enabled to eliminate objects that were not fortified settlements. Secondly, the methodology of extracting vertical terrain profiles from the Digital Elevation Model was proposed, for the objects that had been detected at the first stage of detection. Moreover, points that constitute local maxima and create the contour of the settlement were determined for the generated terrain profiles. The effectiveness of the methodology was assessed based on actual historical settlement structures, i.e. settlements consisting of one or two elements, located in Poland.