The effect of short ground vegetation on terrestrial laser scans at a local scale

Abstract

Terrestrial laser scanning (TLS) can record a large amount of accurate topographical information with a high spatial accuracy over a relatively short period of time. These features suggest it is a useful tool for topographical survey and surface deformation detection. However, the use of TLS to survey a terrain surface is still challenging in the presence of dense ground vegetation. The bare ground surface may not be illuminated due to signal occlusion caused by vegetation. This paper investigates vegetation-induced elevation error in TLS surveys at a local scale and its spatial pattern. An open, relatively flat area vegetated with dense grass was surveyed repeatedly under several scan conditions. A total station was used to establish an accurate representation of the bare ground surface. Localhighest-point and local-lowest-point filters were applied to the point clouds acquired for deriving vegetation height and vegetation-induced elevation error, respectively. The effects of various factors (for example, vegetation height, edge effects, incidence angle, scan resolution and location) on the error caused by vegetation are discussed. The results are of use in the planning and interpretation of TLS surveys of vegetated areas.

Error in target-based georeferencing and registration in terrestrial laser scanning

“…In addition to the georeferencing/registration-induced uncertainty, the quality of TLS data can be affected by the presence of ground vegetation. The vegetation effects must be taken into account when using TLS in an attempt to survey vegetated ground surfaces (Fan et al, 2014).…”

Terrestrial laser scanning for quantifying small-scale vertical movements of the ground surface in Artic permafrost regions

“…As the scan setup, i.e. the incidence angle of the laser beam, has an impact on occlusion patterns (Fan et al, 2014;Lindenbergh 15 and Pietrzyk, 2015), our method is evaluated for two different approaches: (1) TLS scans are repeated from the same scan positions (2015/08-2016/08) and, thus, the ground surface is scanned with similar incidence angles and (2) the surface is captured from different scan positions compared to the TLS survey before (2015/06-2015/08; 2015/06-2016/08). It has to be noted that with a non-stationary TLS setup it is not feasible to locate the laser scanner at the exact same position and height.…”

“…Thereby we build on the finding that elevation errors due to (short) ground vegetation can be reduced by capturing the area from multiple scan positions (Fan et al, 2014). To get further insights and validate our method, deformation analysis is performed based on (1) all ground surface TLS points and (2) after selecting those ground surface TLS points with neighboring points from other scan positions, i.e.…”

“…The effect of elevation errors due to short ground vegetation is further described by Fan et al (2014). An error of about 65% of the grass height is reported.…”

“…the angle between surface normal and laser beam, for closer ground areas mean that it is easier for laser beams to penetrate through the vegetation to a lower level and to reach the ground surface below vegetation (Fan et al, 2014). Furthermore, Fan et al (2014) conclude that (1) lower incidence angles (greater visibility) reduce the occlusion effects, (2) scanning the same area from multiple locations reduces the vegetation error and (3) greater 10 vegetation errors occur where vegetation is higher/denser. The effect of unfavorable scan geometries on soil erosion measurements is described by Eltner and Baumgart (2015).…”