Surface Representation and Area-wise Deformation Analysis

• Applicant: Professorin Dr. Corinna Harmening
• Subject Area: Geodesy, Photogrammetry, Remote Sensing, Geoinformatics, Cartography
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 490989047

Wider research context: The point cloud comparison required for a laser scanner-based deformation analysis can be realized by means of three different strategies: point-to-point, point-to-surface or surface-to-surface. The latter is pursued in this project as it solves two major challenges of a laser scanner-based deformation analysis: Firstly, point correspondences in different point clouds can usually not been taken for granted. Secondly, the magnitude of the deformation may exceed the measurement noise only by a small amount, making the deformation’s detection from raw point clouds impossible. Hence, the basic prerequisite for a successful deformation analysis is the initial modelling and, thus, filtering of the point clouds. Objectives: Within this project, a functional and a stochastic approach to model point clouds and subsequently derive deformations are investigated. Within the functional approach, it is explored whether a refined geometric approximation of the object’s surface provides an added value in terms of localisation of deformations as well as of their estimation. The second approach researches the benefit of a stochastic surface representation in terms of uncertainty and correctness of the deformations’ estimation. Approach and methods: The refinement of the geometric approximation can either be geometrically realized or by regarding geometric details on different coherent spatial scales. Both strategies are going to be investigated within the first subproject employing B-spline based approximation techniques (hierarchical B-Splines, LR B-Splines) and multiresolution analysis (B-spline wavelets). The stochastic approach of the second subproject investigates how typical deformation processes can be represented as realizations of various stochastic processes. Based on the knowledge gained, segmentation and clustering strategies are to be developed. Afterwards, spatial and temporal auto- and cross-covariance functions describing the stochastic process are to be estimated. The estimation of the deformations is realized within an approach similar to least squares collocation. Level of originality and innovation: Terrestrial laser scanning (TLS) has been used only rarely for rigorous deformation analyses so far. In order to use TLS for deformation analysis in such a way that a rigorous statistical assessment of the detected deformation is possible, some challenges have to be overcome. Among other things, a surface representation of the measured object surface is needed that allows for representing object details as well as for introducing smoothness assumptions. Additionally, changes in individual parameters should be connectable to individual – if possible spatially limited – deformations. Primary researchers involved: The Head of Chair for Geodetic Sensorsystems at the KIT (Prof. C. Harmening) and the Head of the Research Division for Engineering Geodesy at the TU Wien (Prof. H. Neuner) cooperate closely within this project.

• DFG Programme: Research Units
• Subproject of: FOR 5455: Deformation analysis based on terrestrial laser scanner measurements (TLS-Defo)
• International Connection: Austria
• Partner Organisation: Fonds zur Förderung der wissenschaftlichen Forschung (FWF)
• Co-Investigator: Professor Dr.-Ing. Frank Neitzel
• Cooperation Partner: Professor Dr.-Ing. Hans-Berndt Neuner