HOW TO EXTRACT USEFUL INFORMATION ABOUT THE DECAY OF BASS RELIEVES IN ARCHAEOLOGICAL AREA
- 1Department of Construction, Civil Engineering and Architecture, Università Politecnica delle Marche, Ancona, Italy
- 2Department of Information Engineering, Università Politecnica delle Marche, Ancona, Italy
- 3Institute for technologies applied to cultural heritage, Consiglio Nazionale delle Ricerche, Rome, Italy
Keywords: Change Detection, Color Decay, 3D Reconstruction, Archaeology, Photogrammetric application, k-means clustering
Abstract. Cultural Heritage goods represent the memory and the history of the civilization. Notwithstanding, there are not sufficient public resources to guarantee their preservation and maintenance. Nowadays between several geomatic techniques available, the pillar for the preservation of mankinds heritage is the low cost close photogrammetric acquisition. The advantages of virtual reconstructions based on Multi View Stereo (MVS) and Structure from Motion (SfM) algorithms is extended from the heritage documentation to its virtualization or modelling. The digital preservation of archaeological sites is committed in more agile and friendly procedures that give automatic extraction of information to perform in depth analysis over ancient artefacts. In the field of CH research, the characterization and classification of the conservation state of the materials composing the surface of the artefacts are essential to study their damage. The first step for conservation state of a goods is the study of the changes in different times. The possibility to automatically study this time modification due to different factor represents a key point for the archaeologists’ work. With this in mind, the aim of this work is to propose a completely automatic methods for change detection between three data set acquired in different époques. The work flow applied is based on the unsupervised clustering techniques applied on a combination of two type of differences images. The results, unlike the objective, demonstrate that the unsupervised methods are not effectiveness in the CH study, instead of the supervised methods that outperforms in terms of reliability of results.