Comparison of pre- and self-calibrated camera calibration models for UAS-derived nadir imagery for a SfM application

Abstract

Structure from Motion (SfM) is a tool being increasingly utilised in geosciences for high-resolution three-dimensional mapping of landscapes. However, a number of authors have demonstrated that broad-scale systematic deformations, in the form of ‘doming’ and ‘bowling’, can occur when applied to linear (low-amplitude, feature-limited) topographies. In such contexts, a more rigorous lens calibration and ground control point acquisition process is required, which means that application of SfM to environments such as tidal flats or desert plains can be challenging. Uncertainties in elevation models generated through SfM were investigated here in the context of the low elevation, micro-topographic environment of saltmarsh. Eight digital surface models (DSMs) were generated for a saltmarsh site in the Deben Estuary (Suffolk, UK) using imagery acquired by a low-cost consumer grade unmanned aerial system (UAS). The results provide clear illustration of the systematic bowling effect following self-calibration during bundle adjustment. This was due to poor estimations of distortion parameters in the camera model. Deformation was most pronounced when UAS-GPS data were used for georeferencing. The use of dGPS-determined ground control points improved the DSM, but did not fully mitigate the deformations. By introducing a pre-calibrated model, derived using a typical checkerboard routine, deformation was significantly mitigated. These results were tested in both the commercial Agisoft PhotoScan® and open-source Micmac software. When self-calibration was used, Micmac generated significantly more accurate DSMs because a more complex lens distortion model could be implemented. The results show that when mapping flat topographies, pre-calibration of the camera model out-performs self-calibration. However, if pre-calibration is not possible, a complex distortion model (such as Micmac’s Four model) can be utilised to limit deformation. The results of the software analysis concluded there is no one-size fits all software solution, and therefore customisable open-source systems offer many potential benefits.

Monitoring Selective Logging in a Pine-Dominated Forest in Central Germany with Repeated Drone Flights Utilizing A Low Cost RTK Quadcopter

“…The 3D positional accuracy of point clouds and products derived from UAS imagery heavily depends on the UAS survey parameters (flight pattern, camera viewing angle, distribution of GCPs) and the SfM processing parameters [10,18,19]. For a weak network of tie points, as commonly achieved with parallel flight tracks and nadir images, systematic errors such as doming of the model frequently occur [18].…”

“…For a weak network of tie points, as commonly achieved with parallel flight tracks and nadir images, systematic errors such as doming of the model frequently occur [18]. This effect can either be mitigated by using sufficient and well-distributed GCPs, or by using more complex flight patterns, including oblique images [10,18,19]. The surveyed GCPs also provide the basis for the absolute orientation and scale of the model within the external (e.g., geographic) coordinate system.…”

“…The final SfM models are highly accurate as well, and thus allow for reliable direct georeferencing results. Additionally, systematic errors, as mentioned above, can be avoided [19].…”

Camera Calibration Using Multiple Unordered Coplanar Chessboards

“…Thus, a 'common' mapping flight, mostly consisting of similar-scale nadir images, does not represent optimal network schemes for camera parameter estimation (James et al, 2017). The problem is further aggravated with flat terrain, corridor configurations and sparse or unevenly distributed ground control (Hastedt & Luhmann, 2015;Griffiths & Burningham, 2018). In such cases, the obvious response is camera pre-calibration (Gerke & Przybilla, 2016;Hastedt et al, 2016;Cramer et al, 2017).…”

“…Thus, pre-calibration with a typical chessboard-based routine and camera settings identical to those of the flight has been reported to outperform self-calibrations with single-scale nadir imagery, the favored choice in environmental mapping (Griffiths & Burningham, 2018); on the other hand, low accuracy in elevations was obtained whenever different focus settings were used for pre-calibration to ensure good frame coverage and focus (Harwin et al, 2015;Han et al, 2016;Gabrlik et al, 2018). This points to an obvious severe limitation of chessboard-based tools, namely the size of the 2D calibration object, either printed or on screen: reported sizes of such patterns do not seem to exceed 1x1 m 2 (Bouros et al, 2015).…”