Integrating LIDAR intensity measures and hyperspectral data for extracting of cultural heritage

This paper deals with the integration of LIDaR and hyperspectral data to improve the discovery of new archeological sites. We investigate the possibility of integrating such dataset in order to evaluate irregular behavior of some major ground indexes. While hyperspectral data allow the identification of specific humidity, vegetation and thermal conditions in the target area, accurate geometric information are provided by LIDaR data. In fact, accurate filtering of lasers canning data allows the computation of accurate DTM (Digital Terrain Model), while Lambertian-based calibration of LIDaR intensity enhances the automatic data segmentation and, in this way, the detection of possible sites of interest. So far as hyperspectral data, MIVIS and AISA sensor data were used. In order to fully extract spectral information from these, some specific procedure have been driven, also implementing PCA analysis. Such data processing have been performed using a wellknown commercial software (ENVI® by RSI) on data acquired over the city of Aquileia (UD, North-East of Italy), an area of great historical interest due to many first century B.C. units of roman ruins. Significant results where obtained both in already discovered archaeological sites, and in interesting new areas on the northern side of the old city. When integrated, the resulting datasets showed, with sensible accuracy, the presence of surface/below surface archaeological heritage.