Reconstructing Ground-Penetrating Radargrams into 3D
Point Cloud Format for Building Inspection: A Web AR
Approach Evaluated by Technology Acceptance Model
Meng Sun1*, Longyong Wu1, Ying Katherine Deng1, Xiao Li2, and Fan Xue1
1

Department of Real Estate and Construction, The University of Hong Kong, Pokfulam, Hong Kong, China
2
Department of Civil Engineering, The University of Hong Kong, Pokfulam, Hong Kong, China
sunmhku01@connect.hku.hk, wulongyong@connect.hku.hk, ydeng@hku.hk, shell.x.li@hku.hk, xuef@hku.hk
Abstract –
Non-destructive Testing (NDT) is important for
building structure inspection. Ground Penetrating
Radar (GPR) is an effective non-destructive testing
technique for building inspections, yet the complexity
of the radargram data has limited its widespread
adoption in engineering practice. This paper presents
an approach for reconstructing raw GPR radargrams
into 3D point clouds for intuitive visualization and
interactive analysis. To evaluate the usability and
industry acceptance, this paper conducts a
questionnaire survey based on the Technology
Acceptance Model (TAM) among 44 industry
professionals in Hong Kong. The results indicated
that the 3D reconstruction approach received
perceived usefulness of 3.26 (Belief that technology
improves performance) and perceived ease of use of
3.55 (Belief that technology is easy to use), with the
“see-through” function being the most highly
regarded. Additional attitude and behavior intention
research further confirmed the approach’s
advantages in terms of intuitiveness and crossplatform compatibility. This paper verified the
feasibility and acceptance of reconstructing GPR
radargrams into 3D point clouds, which can benefit
digital transformation in building inspection.
Keywords –
Ground Penetrating Radar (GPR); Technology
Acceptance Model (TAM); 3D reconstruction; Point
cloud; Building inspection

1

Introduction

People spend “an average of 87% of their time in
enclosed buildings,” according to a U.S. national survey
[1]. The aging and deterioration of buildings will
inevitably lead to a degradation of structural performance,
posing a potential threat to residents, public safety, and
even cultural heritage buildings. Building inspection is

1
2345

crucial for identifying potential risks and related facility
management issues [2, 3].
The verification of internal structures in the past
relied on destructive tests, which could leave scars on
buildings [4]. Such destructive tests are infeasible for
certain types of buildings, such as protected heritage
buildings [5]. As a Non-destructive Testing (NDT)
method, Ground Penetrating Radar (GPR) has been
introduced for subsurface building inspection using
electromagnetic signals [6, 7]. GPR can detect subsurface
features in as-built buildings and cultural heritage
practices [8, 9, 10]. However, a persistent knowledge gap
exists between inspectors and the interdisciplinary
heritage conservation team, as complex raw radargrams
(B-scan/C-scan) are difficult for inspectors to interpret
and use in decision-making [4, 11]. Advanced 3D
reconstruction is the key to filling the gap in processing
the complexity of GPR results, and to enable complex
shapes analysis using stereopsis vision [12, 13].
This paper presents a Web AR approach to
reconstructing raw GPR radargrams into 3D point clouds
(GPR-AR) to overcome subjectivity in conventional
interpretations by intuitive, conclusive visual evidence.
However, whether this approach can be accepted by
industry professionals requires further evaluation and
analysis.
The Technology Acceptance Model (TAM) is an
information system theory in various fields to interpret
and evaluate how users will accept and utilize new
technology [2]. TAM clarifies the underlying mechanism
of technology adoption by revealing the causal
relationships between each external variable and actual
system use. The external variables (EVs) encompass
external environmental factors that affect potential users;
EVs include respondents’ backgrounds, technology
characteristics, and usage context [14, 15]. The central
conceptual component of TAM includes two key
constructs, i.e., perceived usefulness (PU) and perceived
ease of use (PEOU) [16, 17]. PU refers to the extent to
which the respondents believe a particular technology

can enhance their work performance, while PEOU
represents the degree to which the respondents believe
using this technology is effortless [14].
This paper further presents a TAM-based analysis to
assess the acceptance of the proposed GPR-AR. The
GPR-AR employed advanced array GPR devices to
acquire raw electromagnetic signals and reconstructed
them into a 3D point cloud. A TAM-based questionnaire
survey targeting professionals was conducted to assess
the perceived potential for usability and ease of use. The
contribution of this paper is twofold: (i) an intuitive and
precise approach to assist the stakeholders in inspecting
the building materials; (ii) a high degree of acceptance of
reconstructing GPR radargrams into 3D point clouds in
the Architecture, Engineering, and Construction (AEC)
industry.

2

The 3D Reconstructed GPR in the Web
AR (GPR-AR) Approach

2.1

Overview

Fig. 1 shows an overview of the 3D reconstructed
GPR presented in the web AR (GPR-AR) approach. It
comprises four layers: acquisition, processing,
visualization, and interaction. The approach begins with
the acquisition layer to collect on-site data. Then, the raw
GPR radargram is converted to a 3D point cloud. Next, a
3D reconstruction of GPR scanning results is performed
in the processing layer, followed by the visualization
layer. The interaction layer provides valuable functions
to construction stakeholders and facilitates building
inspection.

Fig. 1 The technical framework of GPR-AR

2
2346

2.2

Layer-wise technical implementation

The acquisition layer bridges real-world conditions
into data. Regarding hardware control, the GPR device’s
parameters should result in sharp and clear
electromagnetic signals. Then, a standard GPR grid scan
is carried out to acquire a GPR project data file. Note that
GPR data acquisition precedes any processing.
The processing layer generates C-scans and then
transforms raw GPR radargrams into 3D point clouds.
The Inverse Distance Weighted gridding method is used
to convert B-scan data to C-scan data. Besides gridding,
cell size, search radius, and moving average are
necessary parameters. The spacing between C-scans is
set as 0.01m. After the above processing, several C-scan
slices can be generated and then passed into the point
cloud reconstruction. The outcome point clouds serve for
3D visualization and enable 3D object detection. Each Cscan slice will be converted into a 2D point cloud slice,
where the pixel coordinates (x, y) and slice depth (z) are
mapped to spatial positions. Rather than RGB colors, the
reflection intensity should be labeled at the discretized
points. In visualization equalization of GPR-AR,
logarithmic and normalization operations lead to define a
new index:
= (ln − ln
)/(ln
− ln
),
where is the reflection intensity,
and
are
the maximum and minimum reflection intensity in all
slices. The labeled point cloud slice is finally stacked
based on the spatial depth and merged into a structured
point cloud.
The third layer visualizes the incoming point cloud in
two modalities: isometric 3D and augmented reality (AR).
The isometric 3D is convenient for offline visualization
and analysis. AR mode aims to map an isometric 3D
model for online rendering and visualization.

User functions in the interaction layer facilitate user
operations, as shown in Fig. 2. A sliding filter controls
point-level visibility based on the RI range. Adequately
adjusting the filter hides low-value noises and highlights
building elements. The 3D interface adopts adjustable
XYZ-axis cutting planes for focused zones. Length and
area measurements enable users to efficiently measure
spatial data using mouse clicks.

3

Evaluation Using a
Survey Based on TAM

3.1

Survey design

Questionnaire

A questionnaire survey was designed based on the
TAM, as shown in Fig. 4. The respondents were
professionals in the AEC industry to reflect the
acceptance levels of experts [16, 18, 17]. EVs use
respondents’ experience and their technology usage
context. PU in TAM is measured by evaluating the core
functions of the platform, while the overall usability
rating measures PEOU. The outcomes of PU and PEOU
surveys reflect the respondents’ attitudes and behavioral
intentions towards use (ABI). This paper employs TAM
and designs a structured questionnaire to evaluate the
acceptance of the GPR-AR platform among professionals.

Fig. 2 Interface of the GPR-AR platform

2.3

A demonstration case of reconstruction

Fig. 3 shows a demonstration case of GPR-AR. In
this case, a GPR scan was conducted on an aging stair of
the MuiWo watchtower in Hong Kong, and the results
were processed using GPR-AR. The raw B-scans are
isolated and difficult for non-specialists to interpret. In
contrast, GPR-AR provides a simple, easy-to-use 3D
visualization of results. Using the length measurement
function, the spacing between vertical and horizontal
rebars was easily measured at 0.17m and 0.16m,
respectively. The result was confirmed by the exposed
rebar layout in Fig. 3. Two horizontal rebars are detected
at a depth of 16 cm, spaced approximately 23 cm apart in
the 3D visualization. The user-friendly GPR-AR
interface and functions enabled stakeholders to measure
and interpret the internal details.

Fig. 4 GPR-AR platform survey based on the
TAM [16, 14]
The questionnaire comprises three sections, as listed
in Table 1: background information about the
respondents, the GPR-AR platform evaluation and openended feedback on the GPR-AR platform. Each question
corresponds to one specific TAM construct. The
questionnaire was electronically distributed to industry
professionals after two professional workshops on GPR.
Table 1 The structure of the GPR-AR evaluation
questionnaire
Section
Backgro
und

Evaluati
on

Fig. 3 A demonstration case of GPR-AR

3
2347

Questions
Q1 How long have you been working in
the AEC industry?
Q2 How much do you know about
GPR?
Q3 How useful do you find GPR
technology for building inspection?
Q4 How intuitive is the “see-through”
visualization feature of the GPR-AR
platform?
Q5 How useful do you find the length
measurement tool in the GPR-AR
platform to be for the building
inspection?
Q6 How useful do you find the area
measurement tool in the GPR-AR

Constructs
EV
EV
PU
PEOU
PU

PU

platform to be for the building
inspection?
Q7 How would you rate the ease of use PEOU
of the GPR platform?
Q8 To what extent can the GPR-AR PU
platform
improve
the
building
inspection?
Q9 Which aspect of the GPR-AR ABI
platform needs the most improvement?
Feedbac
Q10 What other suggestions do you ABI
k
have for developing and applying the
GPR-AR platform?
(Note: EV refers to external variables; PU refers to perceived
usefulness; PEOU refers to perceived ease of use; ABI refers to
attitudes and behavioral intentions towards use)

A five-point Likert scale was adopted in the
evaluation [3]. Four questions correspond to PU (Q3, 5,
6, 8) to collect respondents’ responses, including Not at
all helpful = 1, Slightly helpful = 2, Moderately helpful =
3, Very helpful = 4, and Indispensable = 5. For the
question corresponding to PEOU (Q4, 7), a five-point
Likert scale was also used, where Very difficult = 1,
Somewhat difficult = 2, Neutral = 3, Somewhat easy = 4,
and Very easy = 5. Notably, although the “see-through”
visualization is the prerequisite for platform use, it
primarily serves to reduce the burden of data
interpretation. Therefore, Q4 was classified as PEOU,
which in turn has an indirect impact on PU.

3.2

Data collection

To ensure the accuracy and relevance of the
investigation, the target population was limited to active
professionals in the fields of construction, safety
inspection, and heritage protection in Hong Kong. These
professionals, selected from the Hong Kong Institute of
Engineers (HKIE, 90.9%) and the Society of Registered
Safety Officers (SRSO, 9.1%), were affiliated with
contractors, developers, government agencies, and

universities. All participants hold a high degree of
professional standing and technical expertise. The survey
was conducted during industry seminars and workshops
from June to July 2025, ensuring that the feedback
received was from practitioners familiar with on-site
inspection challenges and emerging digital tools.
The collected data was processed into three
categories. The data collected in the background section
were categorical. The data in Section Evaluation are
ordinal (e.g., Indispensable = 5 in the PU-related
questions). The data in Section Feedback is open-ended
qualitative data. All the ordinal data were analyzed using
descriptive statistics to evaluate the GPR-AR platform.
Means, variances, and standard deviations were
calculated for each question to summarize central
tendency and response dispersion.

4

Findings

4.1

Questionnaire survey results

A thorough survey was carried out to evaluate the
functions and applications of the GPR-AR platform.
Among the 44 respondents, more than 75% have at least
1 year of professional experience in building construction
or building safety, with 21 having over 10 years.
For those professional respondents, 45% had no
prior knowledge of GPR, while others had only basic
awareness (41%) or were familiar with general principles
(14%), indicating that GPR is well-known but not widely
applied within the construction and safety industry.
The evaluation results of three core functions are
listed in Table 2. “See-through visualization” was the
most recognized function, with more than half of the
respondents finding it very helpful (score ≥ 4). Length
and area measurement are also considered essential

Table 2 Evaluation of core functions on the GPR-AR platform
Functions
Construct
Mean
Variance
Standard
deviation

“See-through”
visualization
PEOU
3.55
0.38

Length measurement

Area measurement

PU
3.30
0.39

PU
3.34
0.54

0.62

0.74

0.62

Distribution

1: Not at all helpful; 2: Slightly helpful; 3: Moderately helpful; 4: Very helpful; 5: Indispensable

4
2348

functions in engineering practice. Moreover, these
industry stakeholders believed that the GPR-AR platform
should be improved in several key functions of concern
in the industry, including interface design (mentioned by
52%), functional design (50%), and processing speed
(26%). Some professionals also suggested that
conversion from a 3D scan to a 2D drawing/BIM is
needed [19].

4.2

TAM model assessment

Four-item PU and two-item PEOU constructs were
created to calculate composite scores (Table 2). The
reliability of the PU was tested using Cronbach’s alpha
value (α), while the internal consistency of PEOU was
evaluated through the correlation coefficient between the
only two items [20]. For the ordered Likert-scale data, a
nonparametric Spearman rank correlation analysis was
employed to examine the influence of respondent EVs on
platform evaluation [3].
As shown in Table 3, PU demonstrated excellent
reliability (α = 0.803) based on its four-item scale. The
PU composite was 3.26 (mean score of four items), with
a distribution indicating a tendency towards higher scores.
PEOU only included two items that significantly reduced
α (0.434); therefore, an alternative 2-item correlation
analysis was applied [21]. The 2-item construct
demonstrated acceptable reliability for short scales, as
indicated by the Spearman coefficient (ρ > 0.30, p < 0.05).
Consequently, the 2-item composite score was calculated
to provide preliminary evidence of internal consistency.

loading and composite reliability (CR). In Table 3, the
items demonstrated significance and had factor loadings
greater than 0.4 (Q4’s was close to 0.4), which can be
considered a sufficient variance-explained rate for their
construct. CR of PU was greater than 0.7, indicating good
aggregation validity; CR of PEOU was within (0.4, 0.7),
indicating a moderate but acceptable aggregation effect.
The distributions of PU and PEOU by EVs
(professional experience and GPR knowledge level) were
elaborated in Fig. 5. PU of GPR-AR by professionals was
significantly influenced by their professional background.
PU had a “U-shaped relationship” with their professional
experience, while PU was positively correlated with the
GPR knowledge level. This indicates that GPR-AR is of
great value to new practitioners and experienced
professionals alike. On the other hand, PEOU was not
significantly affected by their GPR knowledge level. This
is because ease of use is an objective attribute of a
technology, while usefulness is an evaluation that
depends on the cognition and needs.
A Spearman correlation matrix was established to
analyze the correlations among EV, PEOU, and PU. The
correlation between EV and other constructs indicated
the extent to which technology has become popularized.
The work experience and GPR knowledge in EV were
mutually independent. It was found that the correlation
between the two items and other assessment variables
was insignificant (Table 4). Organizations and
stakeholders continue to maintain a conservative attitude,
which aligns with the TAM. There was a significant
correlation between PU and PEOU (r = 0.654), consistent

Table 3 Questionnaire analysis of evaluation items
Construct

Perceived
usefulness (PU)

Item

Factor Loading

Q3

0.357

Q5

0.894

Q6

0.850

Q8

0.784

Q4

0.487

Perceived ease of
use (PEOU)
Q7

α

CR

0.803

0.821

0.434

0.437

Distribution

0.569

Confirmatory factor analysis (CFA) assesses
whether the measurement model accurately reflects the
construct or variable being measured. Content validity
was confirmed through classification using TAM.
Convergent validity was evaluated by the extracted factor

5
2349

with a prior TAM study [22]. This positive correlation
reflects the interrelationship between PU and PEOU.
Especially in the context of emerging technologies,
PEOU has a significant impact on PU.

The live poll indicated that students recognized that
GPR-AR can improve: (i) 3D presentation of subsurface
over 2D sections (93%), (ii) uses for non-destructive
testing (81%), and accurate rebar locations/details (76%).
We received 15 pieces of feedback comparing GPR-AR
with the mainstream GPR processing software Geolitix.
Generally, it was agreed that GPR-AR has advantages in
terms of intuitive 3D visualization, a lightweight user
experience, an open-source and free nature, and crossplatform compatibility. However, the feedback also
revealed deficiencies in functional depth and data
analysis capabilities. It supplemented the ABI of the
TAM-based survey and provided a clear direction for the
functional optimization of the web AR approach.

Fig. 5 Distribution of PU and PEOU by EVs
Table 4 Discriminant validity: Spearman correlation and
AVE root value
EV
PU
PEOU

EV
0.927
-0.077
(p=0.618)*
-0.165
(p=0.284)*

PU

PEOU

0.744
0.654
(p=0.001)*

0.53

*Values in parentheses represent the significance level, and the
diagonal values are √AVE

4.3

Attitude and behavior intention from
students’ feedback

Given the limited practical experience with GPR
among professionals in the above survey, it is valuable to
gather additional comparative feedback from users with
technical awareness. An exploratory study was
conducted targeting 70 postgraduates with primary
knowledge of GPR. Comparative evaluations and
development suggestions for the GPR-AR platform and
the mainstream commercial software Geolitix were
collected. The thematic analysis method was employed
to code and categorize the textual data, with a focus on
respondents’ attitudes and behavior intentions towards
the platform. This provides an additional perspective for
the TAM study, jointly forming a triangular validation
framework for evaluating GPR-AR.

6
2350

5

Discussion and Limitations

5.1

Survey analysis

The questionnaire design in this paper was based on
the TAM, aiming to structurally assess the potential
application of the GPR-AR platform for building
inspection. The survey results indicated that respondents
had a high PU, especially in the “see-through”
visualization aspect. The PEOU also reached a positive
level, indicating that the platform has a strong foundation
of user acceptance. Few correlations were identified
between EV and the evaluation variables (Table 4),
which demonstrates the ease of use of GPR-AR.
Although the Q3 description directly focuses on
usability, limited familiarity with GPR and the wording
of Q3 hinder respondents from clearly distinguishing
between PU and PEOU for an emerging technology,
resulting in a relatively low factor loading. Besides,
correlated PU and PEOU (r = 0.654, p < 0.001) in the
TAM results in PEOU items inadvertently capturing
some of the PU construct [16, 22]. The boundary
between “usability” and “ease of use” among these
professionals, lacking sufficient experience, may be
blurred [23]. Additionally, reliability analysis for PEOU
involving a small number of items may limit the accuracy.
Future research should adopt a more comprehensive PU
and PEOU scale to clarify reliability and validity.
Based on the ABI feedback from students and the
results of industry questionnaires, the platform is
expected to be systematically enhanced. Expected
improvements include optimizing user interface
interactions, improving the intuitiveness of operations,
and developing advanced analysis functions, such as
machine-learning-based rebar recognition and intelligent
defect diagnosis [24]. More case studies will also be used
to verify engineering usability and ease of use.
In summary, the web AR approach (GPR-AR) has
addressed the limitations of interpreting 1D and 2D GPR
radargrams in building inspection. The investigation

results regarding the GPR-AR platform indicate that its
visualization function effectively leverages augmented
reality to make complex GPR data more comprehensible
and accessible. This TAM analysis demonstrates that
reconstructing raw GPR radargrams into 3D point clouds
effectively reduces professional barriers to GPR, making
it more accessible to diverse stakeholders.

5.2

References

Limitations and prospects of GPR-AR

While the TAM-based survey indicates that GPRAR could contribute to AEC practice, it is limited to
directly measuring improvements in inspection. Future
research will focus on objective performance metrics
using GPR-AR, such as accuracy and time efficiency, to
verify perceptions.
Beyond the perceptual scope of this study, several
technical challenges should be addressed for the wide
adoption of GPR-AR. Firstly, the next phase of research
should focus on the precise co-registration of volumetric
point clouds between the surface and the subsurface.
Secondly, the operation-led trajectory deviations will
affect the accuracy of the results. Reducing the influence
of operational errors during the analysis process is a
research goal for future development. Moreover, the 3D
model is currently an intuitive visualization without any
annotations or segmentation. Future implementations
aim to integrate semantic models and geometric
algorithms to achieve a detailed digital twin model [25,
26, 27].
Despite these challenges, GPR-AR holds significant
promise for a wide range of applications in the AEC
industry. During building inspections, intuitive 3D
visualization enables non-professional stakeholders to
locate internal rebars and detect defects without
destructive tests. In the field of heritage protection, the
approach is helpful for conducting subsurface heritage
documentation. In addition, the approach's intuitive,
easy-to-use nature makes it more convenient for
collaborative
infrastructure
monitoring.
These
application scenarios demonstrate the great potential of
GPR-AR.

6

integrating machine learning algorithms to enable
automatic object recognition. This research provides
empirical evidence for the digital applications of GPR in
building inspections, which align with Hong Kong’s
smart city development goals and the global digital
transformation of the construction industry.

Conclusion

This paper integrates the reconstruction of Ground
Penetrating Radar (GPR) radargrams into 3D point
clouds within a web AR platform for building inspection.
A questionnaire survey based on the Technology
Acceptance Model (TAM) revealed that this integration
was well accepted. The scores for perceived usefulness
(PU) and perceived ease of use (PEOU) performed well,
indicating potential application in engineering practice.
Future work can focus on the following aspects:
achieving precise registration of surface and subsurface
point clouds, optimizing operation-leading errors, and

7
2351

[1] N. E. Klepeis et al., "The National Human Activity
Pattern Survey (NHAPS): a resource for assessing
exposure to environmental pollutants," Journal of
Exposure Science & Environmental Epidemiology,
vol. 11, no. 3, pp. 231-252, 2001.
https://doi.org/10.1038/sj.jea.7500165

[2] Tsukasa Mizutani, Jingzi Chen, and Shuto
Yotsumoto, "The 3D localization of subsurface
pipes from Ground Penetrating Radar images using
edge detection and point cloud segmentation,"
Measurement, vol. 236, p. 115102, 2024. [Online].
https://doi.org/10.1016/j.measurement.2024.115102

[3] Songbo Hu, Yihai Fang, Eric Windholz, and
Kourosh Khoshelham, "Factors influencing the
adoption of remote visual inspection in building
regulatory activities," Journal of Management in
Engineering, vol. 41, p. 05025002, 2025. [Online].
https://doi.org/10.1061/JMENEA.MEENG-6559

[4] Wenting Zhan, Wei Pan, Winnie Law, and Haipeng
Shen, "Stakeholders’ knowledge, attitudes, and
intentions of adopting Modular Integrated
Construction for sustainable development in Hong
Kong," Journal of Management in Engineering, vol.
40,
p.
04023070,
2024.
[Online].
https://doi.org/10.1061/JMENEA.MEENG-5642

[5] D. M. McCann and M. C. Forde, "Review of NDT
methods in the assessment of concrete and masonry
structures," NDT & E International, vol. 34, pp. 71–
84, 2001. [Online]. https://doi.org/10.1016/S09638695(00)00032-3

[6] Wallace Wai-Lok Lai, Xavier Dérobert, and Peter
Annan, "A review of Ground Penetrating Radar
application in civil engineering: A 30-year journey
from Locating and Testing to Imaging and
Diagnosis," NDT & E International, vol. 96, pp. 58–
78,
2018.
[Online].
https://doi.org/10.1016/j.ndteint.2017.04.002

[7] Kevin Lee and Chen Feng, "GPR-Former: Contextaware
moisture
detection,"
International
Symposium on Automation and Robotics in
Construction (ISARC) Proceedings, vol. 2025
Proceedings of the 42nd ISARC, Montreal, Canada,
pp.
1–8,
2025.
[Online].
https://doi.org/10.22260/ISARC2025/0002

[8] Eduardo Diz-Mellado et al., "Non-destructive
testing and Finite Element Method integrated
procedure for heritage diagnosis: The Seville

Cathedral case study," Journal of Building
Engineering, vol. 37, p. 102134, 2021. [Online].
https://doi.org/10.1016/j.jobe.2020.102134

[9] Anne Lalagüe, Matthew A. Lebens, Inge Hoff, and
Eivind Grøv, "Detection of rockfall on a tunnel
concrete lining with Ground-Penetrating Radar
(GPR)," Rock Mechanics and Rock Engineering,
vol. 49, pp. 2811–2823, 2016. [Online].
https://doi.org/10.1007/s00603-016-0943-y

[10] Mercedes Solla et al., "A Building Information
Modeling approach to integrate geomatic data for
the documentation and preservation of cultural
heritage," Remote Sensing, vol. 12, p. 4028, 2020.
[Online]. https://doi.org/10.3390/rs12244028
[11] F. Hassani, "Documentation of cultural heritage;
techniques, potentials, and constraints," , vol. XL5-W7,
2015,
pp.
207–214.
[Online].
https://doi.org/10.5194/isprsarchives-XL-5-W7-2072015

[12] Alexander Bornik and Wolfgang Neubauer, "3D
Visualization techniques for analysis and
archaeological interpretation of GPR data," Remote
Sensing, vol. 14, p. 1709, 2022. [Online].
https://doi.org/10.3390/rs14071709

[13] Fan Xue, Weisheng Lu, Christopher J. Webster, and
Ke Chen, "A derivative-free optimization-based
approach for detecting architectural symmetries
from 3D point clouds," ISPRS Journal of
Photogrammetry and Remote Sensing, vol. 148, pp.
32–40,
2019.
[Online].
https://doi.org/10.1016/j.isprsjprs.2018.12.005

[14] Stanisław Fel, Jarosław Kozak, and Piotr Horodyski,
"Responsibility and AI: Exploring technology
acceptance models," Journal of Innovation &
Knowledge, vol. 10, p. 100852, 2025. [Online].
https://doi.org/10.1016/j.jik.2025.100852

[15] Jiwoong Shin, Sungwoo Moon, Bong-ho Cho,
Sungjoo Hwang, and Byungjoo Choi, "Extended
technology acceptance model to explain the
mechanism of modular construction adoption,"
Journal of Cleaner Production, vol. 342, p. 130963,
2022.
[Online].
https://doi.org/10.1016/j.jclepro.2022.130963

[16] Fred D. Davis, "Perceived Usefulness, Perceived
Ease of Use, and User Acceptance of Information
Technology," MIS Quarterly, vol. 13, pp. 319–340,
1989. [Online]. https://doi.org/10.2307/249008
[17] Viswanath Venkatesh, "Determinants of perceived
ease of use: Integrating control, intrinsic motivation,
and emotion into the Technology Acceptance
Model," Information Systems Research, vol. 11, pp.
342–365,
2000.
[Online].
https://doi.org/10.1287/isre.11.4.342.11872

[18] Nikola Marangunić and Andrina Granić,
"Technology acceptance model: a literature review
from 1986 to 2013," Universal Access in the

8
2352

Information Society, vol. 14, pp. 81–95, 2015.
[Online]. https://doi.org/10.1007/s10209-014-0348-1
[19] Alexis Toledo Bórquez, Felipe Muñoz La Rivera,
Rodrigo F. Herrera, and Javier Mora Serrano,
"Augmented reality and ultrasonic sensor-based
tool for rebar inspection on construction sites,"
Automation in Construction, vol. 176, p. 106288,
2025.
[Online].
https://doi.org/10.1016/j.autcon.2025.106288

[20] Lee J. Cronbach, "Coefficient alpha and the internal
structure of tests," Psychometrika, vol. 16, pp. 297–
334,
1951.
[Online].
https://doi.org/10.1007/BF02310555

[21] Jose M. Cortina, "What is coefficient alpha? An
examination of theory and applications," Journal of
Applied Psychology, vol. 78, pp. 98–104, 1993.
[Online]. https://doi.org/10.1037/0021-9010.78.1.98
[22] William R. King and Jun He, "A meta-analysis of
the technology acceptance model," Information &
Management, vol. 43, pp. 740–755, 2006. [Online].
https://doi.org/10.1016/j.im.2006.05.003

[23] Mark Turner, Barbara Kitchenham, Pearl Brereton,
Stuart Charters, and David Budgen, "Does the
technology acceptance model predict actual use? A
systematic literature review," Information and
Software Technology, vol. 52, pp. 463–479, 2010.
[Online]. https://doi.org/10.1016/j.infsof.2009.11.005
[24] Meng Sun, Dong Liang, Jiajia Wang, Bate Bate,
and Fan Xue, "Clarity in DEM cementation
predictions: Integrating automated machine
learning and interpretability analysis of shear wave
velocity," Advanced Engineering Informatics, vol.
69,
p.
104060,
2026.
[Online].
https://doi.org/10.1016/j.aei.2025.104060

[25] Fan Xue, Wenjin Zhang, Guangji Xu, Qqianyun
Zhou, and Yijie Wu, "Surface or skeleton?
Automatic hierarchical clustering of 3D point
clouds of bronze frog drums for heritage digital
twins," ISPRS Annals of the Photogrammetry,
Remote Sensing and Spatial Information Sciences,
vol. X-M-1-2023, pp. 293–299, 2023. [Online].
https://doi.org/10.5194/isprs-annals-X-M-1-2023-2932023

[26] Meng Sun, Rui Zhao, and Fan Xue, "The
Opportunities and Challenges of Multimodal
GenAI in the Construction Industry: A Brief
Review," The 29th International Symposium on
Advancement of Construction Management and
Real Estate (CRIOCM2024), 2024
[27] Dong Liang et al, "Transfer learning from building
information model-based synthetic data for threedimensional module detection in point clouds of
modular-integrated
construction
hoisting,"
Engineering Applications of Artificial Intelligence,
vol. 164, p. 113243, 2026. [Online].
https://doi.org/10.1016/j.engappai.2025.113243