Calibration of Laser Scanners

点击翻译按钮获取中文摘要

The accuracy of terrestrial laser scanners (TLS) is mitigated by internal scanner misalignments (calibration parameters) whose magnitude mostly exceeds the one of random deviations. This has mainly two consequences in the context of a deformation analysis: On the one hand, the differences d between the measurements at two epochs are biased by the systematic deviations. On the other hand, systematic deviations affect adjacent points in a similar way, which causes correlations between the measurements that need to be incorporated into the stochastic model Σ_dd. To guarantee an unbiased deformation analysis, the systematic deviations must be calibrated. Thus, the goal of this project as part of the whole research unit with the projects is to - reduce systematic deviations to get unbiased point cloud differences d - and to derive a fully populated variance-covariance matrix for the calibration parameters of the scanner Σ_cc, that is an essential part of Σ_dd. A further development of TLS calibration procedures is one of the linchpins for the unbiased laser scanner-based deformation analysis that aims at detecting small geometric changes. Namely, the calibration is of high relevance for several points raised in this proposal for a research unit: a) not-accounting for systematic deviations causing artifacts in data that can bias the surface representation; b) removing instrument-related systematic deviations increases the sensitivity; c) correct quantification of the calibration uncertainty is necessary for the comprehensive measurement uncertainty estimation; d) a correct functional model of systematic errors is necessary for unbiased planning of optimal instrument positioning.