DETECTION OF RESTORATION WORK BY APPLYING THE RANSAC ALGORITHM TO THE POINT CLOUD DATA FROM LASER SCANNING: CASE STUDY AT OSTIA
- Faculty of Human-Environment Studies, Kyushu University, 744, Motoka, Fukuoka, Japan
Keywords: Laser scanning, RANSAC algorithm, Restoration, Roman building, Ostia
Abstract. In Ostia, the huge range of excavation carried out by Guido Calza under Mussolini (1938–1942), the zone of contiguous city blocks unearthed in those massive campaigns. From 2012, new survey by a Japanese team of standing remains using laser scanners formed the basis for an analysis of building history, and for a reconstruction of the original building. There is a considerable amount of undocumented reconstruction work in the upper part of the structure which has been identified from analysis of the surface of the walls. The seam and the absence of coursing between the original walls and the later restored works sometimes including in the Roman phase, and sometimes modern using original part of the walls, make difficult to identify which part of walls were original and which were restorations or re-use 80 years later from the excavation. In this paper, the case that the seams are invisible, but its existence is known from the photographic record of the progress of the excavations. The detection by applying the RANSAC algorithm to the point cloud data from laser scanning relies on several cases of invisible seams running on the surfaces. Additionally, this method allows us without any special knowledge and experience to find detailed characteristics on the surface of the walls, such as slight unevenness or weathering parts, to extrapolate the building history.