Key points are not available for this paper at this time.
Abstract. Glacier monitoring plays a crucial role in understanding the impacts of climate change on these dynamic natural systems. One or more time-lapse cameras are often employed to acquire short-term observations of glacier flow dynamics. However, the lack of multi-camera photogrammetric software packages for multi-temporal 3D scene reconstruction, especially in case of wide camera baselines, hinders the application of Structure-from-Motion techniques to these scenarios. To address this, we present ICEpy4D, a novel Python toolkit designed for 4D monitoring of alpine glaciers using low-cost time-lapse cameras and state-of-the-art computer vision techniques. ICEpy4D leverages deep-learning-based matching algorithms to solve 3D reconstruction with wide camera baselines, making it well-suited for challenging scenarios encountered in mountainous regions. The toolkit offers comprehensive functionalities for multi-epoch monitoring, enabling short-term glacier 3D reconstruction and extraction of relevant information from time-series point clouds, such as volume variations and glacier retreat. In a pilot study on the Belvedere Glacier northern snout (Italian Alps), ICEpy4D estimated glacier volume loss of 63 × 103 m3 of ice and ∼17.5m of retreat. Results showcased the toolkit’s potential for analyzing a glacier ice cliff, with prospects for application to other glaciers with varying characteristics. ICEpy4D is actively being developed as an open-source project at github.com/labmgf-polimi/icepy4d/, promoting ease of extension and customization.
Ioli et al. (Wed,) studied this question.