Benchmarking Local Registration Algorithms on Multi Temporal and Multi Spatial Point Clouds

Tommaso Mainiero, Jad Ghantous, Nives Grasso, Vincenzo Di Pietra

Abstract. Climate change is driving an increase in the frequency and intensity of extreme events in mountainous environments, amplifying geomorphological hazards and the need for accurate multi-temporal topographic monitoring. However, the integration of multi-source datasets remains challenging due to geolocation inconsistencies, heterogeneous data quality, and complex terrain conditions.This study presents a systematic benchmarking framework to evaluate the performance of local point cloud registration algorithms and their impact on geomorphological change detection. Three widely used methods—Iterative Closest Point (ICP), Point-to-Plane ICP, and Generalized ICP (GICP)—were tested across two alpine case studies in Italy (Rio Cucco catchment and Belvedere Glacier), considering different surface types and initial alignment conditions.Results demonstrate that registration performance is strongly controlled by surface morphology, with rocky areas ensuring stable and accurate alignment, while vegetated surfaces introduce significant uncertainties. Point-to-Plane ICP emerges as the most computationally efficient method, whereas GICP provides improved robustness under complex conditions.The study further highlights that integrating robust outlier rejection significantly improves statistical consistency and reduces LoD95. The proposed approach provides a reproducible framework for optimizing co-registration strategies and improving the accuracy of geomorphological monitoring in high-relief environments.