Clearence cairnfields forever: combining AI and LiDAR data in the Marcesina upland (northern Italy)

The plateau of Marcesina, in the municipality of Enego (Vicenza, Italy), was selected as a test site to develop an automated protocol for the recognition and quantification of cairns and cairnfields in Alpine environment coupling object-based image analysis (OBIA) and artificial intelligence (AI). The segmentation was implemented starting from a combination of aerial orthoimages and LiDAR-derived visualizations (nDSM and Openness), which were subsequently classified with a knowledge-based approach. Our rule-set was able to identify 491 cairns in an area of approximately 4.5 km2; the average user’s and producer’s accuracy were calculated thanks to an assessment strategy based on drone flights and ground survey and returned values of 81.4% -95%, respectively.