### Abstract

Rock surface roughness is an important property influencing the shear strength of a rock mass and

therefore its stability. For safety reasons and appropriate engineering design on potentially instable

slopes, accurate and reliable measurement of surface roughness is needed. In this research the stateof-the-art terrestrial laser scanning (TLS) technology is used to remotely sense large rock surfaces

and further on estimate the roughness as a function of direction and scale. The research attempts to

answer the following question: what is the smallest scale of surface roughness that can be reliably

extracted from TLS point clouds? To answer the question, TLS capabilities and limitations are studied

in detail. The main limitation of TLS is the range measurement noise, which can result in

overestimation of surface roughness. To reduce the noise effect, different image de-noising methods

(Discrete Wavelet Transform and Non-Local Mean) are applied on TLS meshes and their results are

compared. A second limitation of TLS data is the effective resolution of the point cloud caused by the

divergence of laser beam, which defines the smallest observable surface detail. Therefore, the

effective resolution of TLS data is analyzed. Firstly a theoretical-empirical method based on Average

Modulated Transfer Function is applied. This method implies some assumptions about measurement

procedure in a laser scanner that usually cannot be tested or are unknown to the end-user. Thus, we

aim to develop an empirical method that is time-efficient and simple in order to be performed in-situ

immediately before or after rock surface acquisition.

therefore its stability. For safety reasons and appropriate engineering design on potentially instable

slopes, accurate and reliable measurement of surface roughness is needed. In this research the stateof-the-art terrestrial laser scanning (TLS) technology is used to remotely sense large rock surfaces

and further on estimate the roughness as a function of direction and scale. The research attempts to

answer the following question: what is the smallest scale of surface roughness that can be reliably

extracted from TLS point clouds? To answer the question, TLS capabilities and limitations are studied

in detail. The main limitation of TLS is the range measurement noise, which can result in

overestimation of surface roughness. To reduce the noise effect, different image de-noising methods

(Discrete Wavelet Transform and Non-Local Mean) are applied on TLS meshes and their results are

compared. A second limitation of TLS data is the effective resolution of the point cloud caused by the

divergence of laser beam, which defines the smallest observable surface detail. Therefore, the

effective resolution of TLS data is analyzed. Firstly a theoretical-empirical method based on Average

Modulated Transfer Function is applied. This method implies some assumptions about measurement

procedure in a laser scanner that usually cannot be tested or are unknown to the end-user. Thus, we

aim to develop an empirical method that is time-efficient and simple in order to be performed in-situ

immediately before or after rock surface acquisition.

Originalsprache | englisch |
---|---|

Publikationsstatus | Veröffentlicht - 26 Apr 2017 |

## Fingerprint Untersuchen Sie die Forschungsthemen von „Rock surface roughness estimation using TLS data“. Zusammen bilden sie einen einzigartigen Fingerprint.

## Dieses zitieren

Bitenc, M. (2017).

*Rock surface roughness estimation using TLS data*.