[Back]

C. Harmening, H.-B. Neuner:
"Evaluating the performance of a space- and time-continuous deformation model";
Talk: 4th Joint International Symposium on Deformation Monitoring (JISDM), Athen (invited); 2019-05-27 - 2019-05-29; in: "4th Joint International Symposium on Deformation Monitoring (JISDM)", V. Gikas (ed.); (2019).

The establishment of terrestrial laser scanning in engineering geodesy provides new possibilities, but also presents new challenges. An appropriate utilization of the huge amount of information contained in laser scanning point clouds requires the development of areal analyzing approaches. Currently, especially the development of a point cloud-based deformation analysis receives much attention.
In this contribution, a space- and time-continuous deformation model is introduced. The basis of this approach is formed by the initial object´s geometry of the first measuring epoch, which is represented by a trend surface in terms of a best-fitting B-spline surface. The deviations of the point clouds measured in the subsequent epochs with respect to the trend surface reflect the deformations. These deviations are interpreted to be caused by a locally stationary and locally homogeneous stochastic process and, thus, they are modelled by means of stationary and homogeneous correlograms as well as slowly varying variances. The combination of the deterministic trend surface with this stochastic signal leads to an adjustment problem similar to a least squares collocation.
The focus of this contribution lies in the application of the deformation model to a deformed surface acting like a first-order system which follows the step response. Simulated data sets of five measuring epochs are used in order to analyze the approach´s performance by investigating the filtering´s residuals and by comparing the filtered data sets to nominal surfaces.

B-spline surfaces; Correlation structure; Laser scanning; Least squares collocation; Locally stationary/homogeneous stochastic process; Space- and time-continuous deformation analysis; Step response

https://publik.tuwien.ac.at/files/publik_280623.pdf