Detailed recording of gully morphology in 3D through image-based modelling

Abstract

15The ability to understand gully erosion development is closely related to our ability to quantify 16 the morphology of gullies. At present, various technologies are at hand to collect data at 17 increasing levels of detail. However, many of the developed technologies are time-consuming, 18 difficult to apply or expensive. As an alternative, image-based modelling offers a cost-efficient, 19 flexible and rapid method to quantify gully morphology from photographs taken in the field. In 20 this study, the use of image-based modelling was tested and fine-tuned to quantify the 21 CITE AS Frankl, A., Stal, C., Abraha, A., Nyssen, J., Rieke-Zapp, D., De Wulf, A., Poesen, J., 2015. Detailed recording of gully morphology in 3D through image-based modelling. Catena 127, 92-101. 2 morphology of four gully heads in contrasting biophysical environments prone to gully erosion: 22 two bank gullies in Central Belgium and two permanent gullies in Northern Ethiopia. Ground 23 photographs (n = 88-235) were taken with a reflex Canon EOS 450D camera having a 20 mm 24 wide-angle lens with a fixed focal length. The data collection occurred during days of 30-100% 25 cloud cover and after removing excessive vegetation in the gullies. Processing of the 26 photographs occurred in PhotoScan 1.0.2. software using the semi-automated Structure from 27Motion-Multi View Stereo (SfM-MVS) workflow, and allowed to produce 3D Digital Elevation 28Models with accuracies that range from millimetres to centimetres. In addition, for the same 29 surface, 2.5D models were created in ArcGIS. Gully morphological properties were derived and 30 include cross-sections, total volume and volume of undercut walls and soil pipe inlets. For the 31 volume calculation, OPTOCAT software was used. Cross-sections were also quantified by tape 32 meter measurements. When compared to 3D models, cross-sections quantified from tape meter 33 measurements and from 2.5D models underestimate the cross-sectional area by <1-14% and 0-34 2.5% respectively. Considering gully volume, 2.5D and 3D approximations show only small 35 differences, related to the volume of soil pipe inlets and undercutting areas. These differences, 36 however, highlight the erosive activity of the gullies, and are important to understand gully 37 dynamics in detail. Geomorphologically spoken, areas where undercutting or soil piping occurs 38 are among the most dynamic and reveal where important morphologic changes are about to occur. 39The accuracies reported in this study are similar to those obtained in other studies that consider 40 surfaces of similar scales. In sum, image based modelling is a promising tool to study in detail 41 gully morphology in 3D, which is the closest approximation of the surface morphology. 42 Frankl, A., Stal, C., Abraha, A., Nyssen, J., Rieke-Zapp, D., De Wulf, A., Poesen, J., 2015. Detailed recording of gully morphology in 3D through image-based modelling. Catena 127, 92-101. 3 43 CITE AS

Assessment of UAV and Ground-Based Structure from Motion with Multi-View Stereo Photogrammetry in a Gullied Savanna Catchment

“…These results agree with other studies that have used aerial platforms to capture images across large areas in complex terrain (RMSE values ranging from 0.05 m to~1 m [18,36,47,[101][102][103]), and ground-based approaches to model gully systems (RMSE values ranging from 0.025 m to 0.155 m [40][41][42]46,47,49,104,105]). While the overall elevation errors of the UAV topographic models were large relative to a pre-existing airborne LiDAR dataset and RTK validation points, cross-sections extracted from the UAV DSM broadly match the shape of gully profiles, although are vertically offset.…”

“…In the point-to-point comparison, elevation values were extracted from the dense point cloud (prior to rasterization) and compared to concordant RTK validation points, using the 'compute cloud/cloud distance' tool in CloudCompare. Point-to-point comparison is useful in topographically complex environments [101] such as gullies, where steep sides and overhangs are common [46], as a given set of x and y coordinates can have multiple z values [42,102]. In the raster-to-raster comparison, the UAV DSM was subtracted from the airborne LiDAR DEM, allowing examination of the spatial distribution of error.…”

Image-based surface reconstruction in geomorphometry – merits, limits and developments

“…4a and 5a-b), the study of Snapir et al (2014) due to a very high reference accuracy of Lego bricks (excluded from Figs. 4c and 5b), and the study of Frankl et al (2015) due to a high measured error as the study focus was rather on feasibility than accuracy (excluded from Fig. 5c).…”

Predicting plot soil loss by empirical and process-oriented approaches. A review

“…Terrestrial digital photogrammetry is characterised by high spatial resolution (centimetres to millimetres) and minimal impact of the field activity on both the ground measurements and farming operations and it requires a time-consuming work for post-processing the digital photos. Image-based modelling creates a digital terrain model using a set of photographs taken from the same surface (Frankl et al, 2015). The procedure involves a solution with camera model parameters and scene geometry simultaneously using redundant information coming from oblique images and without using control points during the composition of the model (Gómez-Gutiérrez et al, 2014).…”