Comparative Study of Two Approaches for Application of Terrestrial

Laser Scanner in Structural Health Monitoring and Damage

Assessment

Shakhzod M. Takhirov

Department of Civil and Environmental Engineering, University of California,

Berkeley, 337 Davis Hall, Berkeley, California, U.S.A.

Keywords: Laser Scanning, Displacement Tracking, Laser Targets, Structural Health Monitoring, Corner Detection and

Tracking, Structural Damage Assessment.

Abstract: The main objective of the paper was to evaluate two approaches aimed at tracking small displacements. The

first approach is based on the usage of laser targets commonly used for stitching point clouds together. The

second approach is based on the estimation of a corner of a prismatic shape and utilizes thin horizontal slices

of the shape’s point cloud. The corner’s location is estimated as an intersection of two straight lines best fitted

to the point clouds before and after the corner. It was shown that for both approaches a sub-millimetre

accuracy can be achieved. The first approach requires the installation of two laser targets in order to measure

the change of the distance between them. The second approach offers more flexibility because it does not

require the installation of a laser target. Hence it can be used in the quantitative assessment of structural

damage in the aftermath of natural disasters such as earthquakes, fires, tsunamis, landslides and hurricanes,

to name a few.

1 INTRODUCTION

The application of terrestrial laser scanners and

drones in structural health monitoring and structural

assessment in the aftermath of a natural disaster is

steadily increasing. Terrestrial laser scanners usually

acquire point clouds with a better accuracy than those

collected by the drones. It is quite common that the

laser scanners deliver a few millimetres accuracy.

While this accuracy might be sufficient for most

applications such as tracking large surfaces, it is not

adequate for monitoring small displacements. The

option of utilizing laser scanning targets can increase

the accuracy of tracking. This is related to the fact that

their vertices can be acquired with a better accuracy

based on complex manipulations of the target’s point

cloud. Because of that, they are commonly used as

reference points for stitching the point clouds to each

other. Based on the specifics of the target’s shape and

colouring pattern, their vertices can be acquired with

much greater accuracy and their displacement can be

tracked with a sub-millimetre accuracy. This high

https://orcid.org/0000-0002-4396-7946

accuracy was reported earlier (Takhirov, 2009) for a

single field measurement comparing the

displacement of the target from laser scans acquired

by ScanStation 2 (Leica GeoSystems AG, 2007) to

that obtained by accurate position transducers. This

adequate accuracy for the Trimble laser scanner

(Trimble, 2016) was confirmed for a series of

measurements conducted in the laboratory

environment (Takhirov, Gilani, and Allen, 2021).

This paper is focused on the evaluation of this

approach for the Faro Focus

(Faro, 2021) laser

scanner. In addition, this approach was compared to

another approach based on tracking the corner points

of the prismatic structural elements or components.

This approach was developed earlier (Takhirov and

Mosalam, 2014) and evaluated for ScanStation 2 in

the reconnaissance effort following the 2010 Haiti

Earthquake (Mosalam, Park, and Takhirov, 2014). It

was also evaluated for a cost-effective scanner

(Takhirov, Gilani, and Allen, 2020). Recently this

approach was developed further for applications in

quality control of construction (Takhirov, 2021).

166

Takhirov, S.

Comparative Study of Two Approaches for Application of Terrestrial Laser Scanner in Structural Health Monitoring and Damage Assessment.

DOI: 10.5220/0010992000003121

In Proceedings of the 10th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2022), pages 166-171

ISBN: 978-989-758-554-8; ISSN: 2184-4364

Point clouds collected by the C10 laser scanner (Leica

GeoSystems AG, 2010) was used in the latter study.

This paper is focused on the evaluation of both

approaches for point clouds collected by the Faro

Focus

laser scanner (Faro, 2021).

2 EXPERIMENTAL SETUP

The laser scanning project was conducted in a

laboratory environment and a special experimental

setup was assembled for the project. The setup

consisted of a linear bearing rail with two carriages

that can slide along the rail’s axis. One of the

carriages was fixed and the other one was displaced

in respect to the fixed one. A rigid block was mounted

on top of each carriage. To imitate the typical texture

and colouring of the material commonly used in

construction, two concrete blocks were used. As

noted earlier, both were placed on the carriages of a

linear bearing system as presented in Figure 1. One of

the blocks was larger in size with overall dimensions

of 394 mm by 197 mm by 89 mm. Two laser scanning

targets (LT and LB) were installed on this block as

shown in Figure 1. The second block was smaller

with overall dimensions of 292 mm by 292 mm by 51

mm. This block had only one laser scanning target

(RB) and the block and the target were used as a

reference. For this purpose, the block was installed on

a fixed carriage of the linear bearing system that did

not move during the experiments.

Figure 1: Setup: two concrete blocks on a linear bearing

system with a few targets.

Two additional fixed laser scanning targets were

used in the study. They were mounted on the wall as

shown in Figure 1 and they were labelled W1 and W2,

respectively.

The tall block was left free to move in respect to

the fixed one, but it was restrained to only move along

the linear bearing system by its attachment to the

movable carriage in the system. The displacements

were imposed in an incremental way by inserting

high-precision blocks (Mitutoyo, 2016) between the

carriages as presented in Figure 2.

Figure 2: Displacement between carriages set by high-

precision blocks.

As shown in Figure 3, the terrestrial laser scanner,

Faro Focus

(Faro, 2021), was used to collect point

clouds of both blocks at each displacement.

Figure 3: Terrestrial laser scanner used in the setup.

Comparative Study of Two Approaches for Application of Terrestrial Laser Scanner in Structural Health Monitoring and Damage

Assessment

167

The laser scanner was placed at about 3.8 m away

from the blocks and it was not moved during the

experiment. The scanner was installed at about 45

degrees to the planes of the concrete blocks. It had a

setting of 1/4 resolution and 3x quality (in the scanner

specific options selectable for a scanning process).

The laser scanner was levelled during the scans

ensuring that the vertical axis of all scans coincided

with the true gravitational axis.

3 TWO APPROACHES

Two approaches were evaluated in this study. The

first was based on tracking the centres of laser

scanning targets and the second was based on

monitoring the displacements of the concrete block’s

corner in respect to the corner of the fixed block.

3.1 First Approach

The first approach for tracking the displacements of

the blocks was based on utilization of the laser

scanning targets. In each laser scan the centres of the

targets were estimated by using Cyclone software

from Leica GeoSystems (Leica GeoSystems AG,

2018). A result showing all the targets for one of the

typical scans is presented in Figure 4. The target

vertices are indicated by crosses. In this paper, the

point clouds are presented in a local coordinate

system ensuring that the X-axis is parallel to the

direction of displacements and the Z-axis is parallel

to the true gravitational vertical axis.

Figure 4: Vertices of laser scanning targets and the point

cloud.

The acquired target centres were exported in

ASCII format and all remaining data reduction was

conducted in the Matlab environment (MathWorks,

2016). Estimates of the distances between the vertices

are presented in Figure 5 for LB to RB targets as a

typical example.

It is worth noting that the point cloud data and the

location of vertices does need to be transformed into

a new coordinate system because the distances

between the vertices are not dependent on a selection

of a coordinate system.

Figure 5: Approach 1: distance between vertices of targets.

This procedure was applied to all five targets in

the study. The targets fixed to the wall were used for

estimation of accuracy of the approach since the

targets were not moving in respect to each other.

This approach was evaluated earlier (Takhirov,

Gilani, and Allen, 2021) for a terrestrial laser scanner

Trimble TX6 (Trimble, 2016) and it was shown to be

an adequate correlation with displacements measured

by accurate position transducers.

3.2 Second Approach

The second approach for tracking the displacements

of the blocks was based on finding a corner of the

moving block at the elevation of the target’s centre,

and tracking its displacement from scan to scan. In

this case, a thin horizontal slice of the point cloud at

the elevation of the target’s centre was analysed as

shown in Figure 6. In this case, the corner of the fixed

block was used as a reference.

A plan view of a typical horizontal slice is shown

in Figure 7. The point cloud corresponding to the

front and the side surfaces of the block are separated

from each other. These subsets of point clouds were

individually best fitted to straight lines by the least

square method. An intersection point of these two

straight lines was considered as a corner of the block.

PHOTOPTICS 2022 - 10th International Conference on Photonics, Optics and Laser Technology

168

Figure 6: Horizontal slices of point clouds at elevations of

target centres (LB and RB).

To increase the accuracy, the points right next to

the corner were removed to minimize the effects of

the corner’s imperfections. The points in shaded

boxes shown in Figure 7 were used for best fitting to

the straight lines.

Figure 7: Typical result for a corner estimation.

Figure 8: Approach 2: distance between the corners is

estimated.

A corner of the fixed block was estimated in the

same way. The distance between two corners was

estimated as shown in Figure 8 and it was tracked

from scan to scan.

This approach was introduced earlier for the

estimation of residual drifts of columns in a

reconnaissance effort conducted soon after the 2010

Haiti Earthquake (Mosalam, Park, and Takhirov,

2014). This study was conducted by means of a

ScanStation 2 laser scanner (Leica GeoSystems AG,

2007). The application of this approach for the quality

control in construction was recently studied

(Takhirov, 2021). In the latter case, a point cloud

collected by the C10 laser scanner (Leica

GeoSystems AG, 2011) was used.

As mentioned earlier this paper is focused on the

evaluation of both approaches by the Faro Focus

laser scanner (Faro, 2021).

4 RESULTS AND DISCUSSION

The approaches described in the previous section of

the paper were compared to each other.

The distance between two targets mounted on the

walls of the laboratory (W1 and W2) must remain the

same. The distances between those targets measured

from the laser scans using the first approach provided

information about its accuracy. A variability of the

estimates of distances between wall-mounted targets

is presented in Figure 9. This variability is shown in

respect to the average of all distances. This result

shows that the variability stays within ±0.6 mm for all

eleven scans.

Figure 9: Variability of estimates of distances between wall

fixed targets (W1 and W2): in respect to the mean.

The distances between the targets mounted on the

movable block (LT and LB) must also remain the

same. The variability of the distance measurements is

, mm

Comparative Study of Two Approaches for Application of Terrestrial Laser Scanner in Structural Health Monitoring and Damage

Assessment

169

presented in Figure 10. The variability of distances

between those targets is also very close to ±0.6 mm

for all eleven scans. The slight difference between the

two variabilities can be related to the accuracy of the

scanner or the spacing differences in the horizontal

direction (wall-mounted targets) and vertical

direction (left block mounted targets).

Figure 10: Variability of distances between LB and LT: in

respect to the mean.

Figure 11 shows the displacements measured by

both approaches compared to the displacements set

by the high-precision blocks (noted as HP-blocks in

the plot).

Figure 11: Displacement estimates: two approaches.

The error between the two approaches is

presented in Figure 12. As shown in the plot, the error

of the displacement estimate stays within ±0.55 mm.

As shown in Figure 12, both approaches have

about the same sub-millimetre accuracy. The

approach using the laser scanning targets has slightly

better accuracy with the error remaining within a

±0.40 mm range around its average. In the approach

based on tracking corners, the error varies within a

±0.55 mm range around its average.

Figure 12: Errors for two approaches (in respect to the

mean).

5 CONCLUSIONS

The main objective of the paper was to evaluate two

approaches aimed at tracking small displacements.

The first approach is based on tracking displacements

of laser targets commonly used for stitching point

clouds to each other. The second approach is based

on the estimation of corners of thin horizontal slices

of a prismatic shape’s point cloud. The corner’s

location is estimated as an intersection of two straight

lines best fitted to the point clouds beyond the corner.

It was shown that for both approaches a sub-

millimetre accuracy can be achieved. The first

approach requires the installation of two laser targets

to measure the change of the distance between them

but offers slightly better accuracy of ±0.4 mm. The

second approach offers more flexibility because it

does not require the installation of a laser target.

Hence it can be used in the quantitative assessment of

structural damage in aftermath of natural disasters

such as earthquakes, fires, tsunamis, landslides, and

hurricanes, to name a few. Based on the results of this

work, the error of this approach is about ±0.55 mm

greater than that of the first approach.

ACKNOWLEDGEMENTS

Special thanks are due to Faro for providing an access

to the Faro Focus

laser scanner. Special thanks are

due to Holly Halligan of UC Berkeley for editing the

paper.

, mm

X coordinate, mm

, mm

PHOTOPTICS 2022 - 10th International Conference on Photonics, Optics and Laser Technology

170

REFERENCES

Faro (2021). https://media.faro.com/-/media/Project/

FARO/FARO/FARO/Resources/TechSheet_Focus_La

ser_Scanner/TechSheet_Focus_Laser_Scanner_EN.pd

f?rev=1291d7f5ef814fe99d471dde60321833.

Leica Geosystems AG (2007). https://w3.leica-

geosystems.com/downloads123/hds/hds/ScanStation/b

rochures/LeicaScanStation%202_brochure_en.pdf.

Leica Geosystems AG (2011). Leica ScanStation C10.

http://w3.leica-geosystems.com/downloads123/hds/

hds/ScanStation%20C10/brochures-datasheet/Leica_

ScanStation_C10_DS_us.pdf.

Leica Geosystems AG (2018). Cyclone Version 9.2.1.

MathWorks (2016). Matlab Version R2016b.

Mitutoyo (2016). https://www.mitutoyo.com/wp-content/

uploads/2016/09/E-section-Gage-Blocks.pdf last

retrieved on 11/23/20.

Mosalam, K.M., Takhirov, S.M., and Park, S. (2014).

Applications of Laser Scanning to Structures in

Laboratory Tests and Field Surveys. Journal of

Structural Control and Health Monitoring, Volume 21,

Issue 1, pages 115–134, January 2014.

Takhirov, S.M. (2009). Video, Still, and Laser Imaging

at nees@berkeley: Case Studies, Presentation at

Bi-annual Hybrd Simulation Workshop, August 2009.

Takhirov, S.M., and Mosalam K.M. (2014). Applications of

Laser Scanning to Structural Testing and Field Survey

in Earthquake Engineering. 10th US National

Conference on Earthquake Engineering, Frontiers of

Earthquake Engineering, Quake Summit 2012, July 21-

25, 2014, Anchorage, Alaska.

Takhirov, S.M., Gilani, A., and Allen, J. (2020).

Experimental Evaluation of Novel On-Demand and

Smart Real-Time Structural Monitoring of Buckling

Restrained Brace’s Deformation with Laser Scanners.

EWSHM2020. The 10th European Workshop on

Structural Health Monitoring. Palermo, Italy, July 6-9,

2020.

Takhirov, S.M. (2021). Control of Construction Quality by

a Terrestrial Laser Scanner: Example of Steel Frame

Building. HORA-2021, the 3rd International Congress

on Human-Computer Interaction, Optimization and

Robotic Applications, June 11-13, 2021, Turkey.

Takhirov, S.M., Gilani, A., and Allen, J. (2021).

Comparative Evaluation of Laser Scanning and Image

Tracking for Deformation Monitoring of Buckling

Restrained Brace. 10th International Conference on

Structural Health Monitoring of Intelligent

Infrastructure. 30 June - 2 July, 2021, Porto, Portugal.

Trimble Inc (2016). The Trimble TX6 laser scanner.

http://trl.trimble.com/docus hare/dsweb /Get/Docum

ent-82496 5/022516-282_Trimb leTX6 DS_US_10

16_LR.pdf. Accessed 3 Jan 2020.

Comparative Study of Two Approaches for Application of Terrestrial Laser Scanner in Structural Health Monitoring and Damage

Assessment

171