IEEE ICSPCC 2019 (SPG 10—6)

Semantic Enrichment for Rooftop Modeling
using Aerial LiDAR Reflectance
22 September 2019
Tan, T., Chen, K., Lu, W., & Xue, F.*
Assistant Professor
Dept. of REC / iLab
FoA, HKU, HKSAR, PRC

Outline
1

Background

2

Semantic Enrichment using LiDAR

3

Discussion

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

2

Section 1

BACKGROUND

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

3

1.1 Background
 Global urbanization
 By 2050, 65% world’s population will live in cities (WHO,
2015)

 Irreversible; Even faster in China

 Leads to urban vulnerability (a.k.a. ‘city diseases’)
 ‘Dead’ space/landscape, low familiarity with surroundings,
 Poor waste treatment, environment (air, water) pollution,
 Heritage destruction, aging town blocks, inefficient traffic,

China’s and global urbanization rates
source: gov.cn

 Disasters (earthquake, climate change), resource crisis, …

 Demands smarter and more resilient development
 (a) Smarter decision supports in multiple disciplines
 (b) On basis of accurate, timely urban semantics
F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

Global urban vulnerability level (Birkmann
et al, 2016) source: nature.com
4

1.2 Urban semantics
 Why semantics from signals? (Rowley & Hartley, 2017)
 Answering interrogative questions (what, who, where, when)
 Enabling automated reasoning / checking
 Abstracted, processed from data and signals

 Types of urban semantics

Data: Digital pixels
(0~255 R, G, B)

 Geometric: Dimension, location, rotation, color, …
 Non-geometric facts: Function, materials, history, owner, …
 Instructions (how-to): Manufacturing, installation, access, …

 Common databases / interfaces
 BIM: building information model
 GIS: geographic information system
F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

Semantics: Car,
building, tree, …
5

1.3 Motivation and aims
 LiDAR data
 Light Detection and Ranging
o Different devices: total station, vehicle-borne, drone

 Aerial LiDAR from drones / fixed-wing aircraft
o Large-scale

o Uniform point density (4~1,000 pts/m2)

o Laser reflectance (received photons from object surface)
o Rooftop details

Illustration of aerial LiDAR

 Semantic enrichment using LiDAR ?
 Geometry

 Non-geometric, e.g., green roof
 topology
F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

2.5D “block” map

Infrared laser reflectance
(warmer color = less received)

6

Section 2

SEMANTIC ENRICHMENT USING
LIDAR

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

7

2.1 Semantic enrichment: Geometry
 LiDAR  RANSAC  rectification  LoD2 model (Chen et al. 2018)

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

8

2.2 Semantic enrichment: Green roofs (1/3)
 Inputs of a pilot area: (a) LiDAR
 Intermediate input: (b) Rooftop elements from geometric modeling (previous page)

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

9

2.2 Semantic enrichment: Green roofs (2/3)
 A supervised learning method
 Decision tree (ctree on R)
o Human readable result

 Label: Potted, turf, non-green
Label
Non-green
Non-green
Non-green
Non-green
Non-green
Non-green
Non-green
Non-green
Non-green
potted
turf
turf

…

turf

Avg.
reflectance
(%)
54.5
53.6
36.7
34.6
50.8
29.5
30.5
33.5
28.1
35.1
54.9
53.7

….

50.4

Top area
(m²)

123.6
66.2
400.5
58.6
12.5
5.0
9.5
29.1
5.3
74.0
61.9
529.3

…

74. 4

height
(m)

2.47
2.39
3.53
3.52
2.84
0.80
0.72
0.63
0.72
0.35
–0.35
–0.34

…

–0.39

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

10

2.2 Semantic enrichment: Green roofs (3/3)
 Output: (a) green roof prediction
 Validation: (b) screenshot of Google Earth

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

11

2.3 Semantic enrichment: Symmetry
 3D point cloud  symmetry hierarchy (Xue et al., 2019)

 Time = 98.6s

 A knowledge discovery tool for further 3D modeling

 PCR = 93.7%

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

12

Section 3

DISCUSSION

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

13

3.1 Discussion
 A pilot study of predicting rooftop materials
 From LiDAR
o Using geometric features (from LiDAR)
o Using laser reflectance (from LiDAR)

 For smart city

 Pros

 Automated
 Data readiness

 Cons
 A small-scale test
 No benchmarking against other methods
o More supervised, unsupervised, reinforcement learning methods

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

14

References
 Birkmann, J., Welle, T., Solecki, W., Lwasa, S., & Garschagen, M. (2016). Boost resilience of
small and mid-sized cities. Nature News, 537(7622), 605.
 Chen, K., Lu, W., Xue, F., Tang, P., & Li, L. H. (2018a). Automatic building information
model reconstruction in high-density urban areas: Augmenting multi-source data with
architectural knowledge. Automation in Construction, 93, 22-34.
 Rowley, J., & Hartley, R. (2017). Organizing knowledge: an introduction to managing
access to information. Routledge.
 World Health Organization. (2015). Global Report on Urban Health—Executive Summary.
http://www.who.int/kobe_centre/measuring/urban-global-report/en/
 Xue, F., Chen, K., & Lu, W. (2019). Architectural symmetry detection from 3D urban point
clouds: A derivative-free optimization (DFO) approach. In Advances in Informatics and
Computing in Civil and Construction Engineering (pp. 513-519). Springer, Cham.

F. Xue: Semantic Enrichment using LiDAR, IEEE ICSPCC 2019, Dalian, China

15

Thank you !
Q&A time

16