Abstract. Structural health monitoring of bridge infrastructure is conventionally carried out either through the acquisition of data from embedded sensor arrays or through systematic inspection and documentation of surface-observable changes on the structure. External deterioration, such as crack formation or load-induced structural deformations, is typically identified during these inspections. However, conventional visual inspections are often labor-intensive and costly. Automated, camera-based photogrammetric monitoring methods offer a promising alternative to reduce both effort and expense. For the specific task of detecting vertical deformations in a bridge superstructure, a tripod-stabilized mono camera setup can be utilized. This approach enables the derivation of vertical deformations by tracking the displacement of target points within a predefined measurement field over time. This study investigates the derivation of vertical deformations from image sequences. Using a representative test bridge as a case study, the characteristic vertical deformation response induced by the traversal of a loaded vehicle is analyzed and documented. A measurement point field was affixed to the bridge superstructure and continuously observed. From the acquired image sequences, continuous vertical displacements were successfully derived, with a mean vertical deflection of the bridge superstructure in the range of −1.71 mm to +0.72 mm. The mean measurement noise, averaged across all monitored test cycles, amounted to 0.01 mm. The results demonstrate that high-quality deformation data can be acquired in a cost-efficient manner using a consumer-grade camera in conjunction with a standardized measurement field, thereby offering a viable low-cost alternative to conventional structural monitoring instrumentation.
R. Blaskow, H. Sardemann, Hans-Gerd Maas· The International Archives o...· 0 citations
Abstract. In this paper, we propose a basic workflow for the automatic extraction of glacier contours from high-resolution multi-temporal 3D point clouds. Based on the hypothesis, that glacier movements cause changes in multi-temporal surface models, glacier contours can be detected even in scenarios where glacier margins are not clearly visible, such as in the case of debris-covered glaciers and rock glaciers. After applying a robust registration algorithm, glacier and non-glacier points are filtered in several steps and spatial resolutions based on dense clusters of significantly changed points. Finally, the glacier margin is mapped using a contour extraction algorithm. The method is applied to various datasets in two scenarios: first, to a slightly debris-covered glacier in Norway; second, to a rock glacier in the Austrian Alps. The results clearly demonstrate the basic functionality of the proposed method. However, morphological changes not caused by glacier movement limit the effectiveness of the filtering, which relies exclusively on density-based elevation changes. Against this background, future work will focus on incorporating additional features, such as velocity fields and prior knowledge of rock glacier dynamics, into the point cloud filtering.
S. Isfort, M. Elias, Hans-Gerd Maas· The International Archives o...· 0 citations
Experimental results demonstrate that the adapted SAM2 model achieves stable segmentation under moderate environmental variability, while degrading under severe visibility loss, consistent across model scales and input resolutions.
Bindusara Nagathihalli Lokesh, Laura Camila Duran Vergara, Hans-Gerd Maas et al.· The International Archives o...· 1 citation