IEEE Consumer Electronics Magazine - May 2018 - 75

AR TECHNOLOGY FOR CITY DATA VISUALIZATION
Currently, many cities are developing rapidly toward becoming smart cities, which differ from traditional cities in the way
that their information for multiple infrastructures is organized
in a smart way, adopting ICT and IoT technologies [9]. Most
of these cities, such as Barcelona, Spain, and Amsterdam, The
Netherlands, are facing the requirements of increasing capabilities and achieving sustainable ways of development, which
cannot be fulfilled by traditional methods [10], [11]. Building
smart cities can greatly improve the quality of public services
for residents, obtain a rational use of the resources, reduce the
work in analyzing the information from the city data for management, and help cities become more environmentally
friendly through saving energy and causing less damage to the
environment [1].
Along with the rapid development of smart cities, AR technology and 3-D reconstruction technology have also been widely used to benefit this progress [12]. Schall et al., e.g., applied
AR technology on a mobile device to interactively visualize
underground infrastructure for on-site inspection, data capture,
and as-built surveying [13]. Porzi et al. used smartphones to augment mountainous landscapes with georeferenced data, such as
the names of peaks and the positions of hiking trails, through
estimating the device location and orientation with a global positioning system and inertial sensors [14]. In addition, to realize
efficient city planning, Santana et al. use both AR and VR technologies to analyze and visualize the patterns of city energy production and demand with location-based services [15].
Because city data play an important part in the smart city
and HoloLens adopts cutting-edge AR technology, we came
up with the idea of visualizing Toronto city data with HoloLens to take advantage of AR technology.

VISUALIZING TORONTO CITY DATA WITH HOLOLENS
In this section, we discuss our method of visualizing Toronto
city data with HoloLens. We first obtain a Toronto 3-D city
model and related city data sets from the city of Toronto
open data, transform these files to desired formats, import
them into our HoloLens application, and then integrate interaction designs.

ally contains detailed descriptions, and we needed to process
this information to fulfill the purpose of mapping related data
with the 3-D city model geographically. Table 1 is an example of the data table description we extracted from a redlight-camera file. Because this file contains position data with
x and y coordinates in a modified traverse Mercator (MTM)
coordinate system, we can directly map them with the reference frame of the city model accurately.
However, some files do not contain straightforward MTM
coordinate information and require further transformation.
Table 2 stores the location data of washroom facilities in a
geographic information system (GIS) with latitude and longitude. The solution we adopt is to apply a transverse Mercator
map projection on the GIS coordinate to transform it into an
MTM coordinate with the related parameters listed in Table 3
[16]. Through this process, we successfully obtained the
buildings' locations on the city model's coordinate system.
In addition to location information, many other types of
data are also used to visualize Toronto city data. When a data
file about taxi stand locations is loaded, e.g., the maximum
number of taxis permitted at each data point could be used in
comparison with others to make them visualized differently.
Moreover, the data set of bicycle shops in Toronto can be

Table 1. The data table description
from a red-light-camera file.
Name

Description

Road 1

First road of intersection

Road 2

Second road of intersection

Easting in MTM NAD27 3° projection

Northing in MTM NAD27 3° projection

Longitude

Longitude in WGS84 coordinate system

Latitude

Latitude in WGS84 coordinate system

Object ID

Unique system identifier

NAD27: North American Datum of 1927; WGS84: World Geodetic
System.

TORONTO CITY MODEL AND DATA SETS PROCESS
We obtained the Toronto 3-D city model and approximately
220 open city data sets from the city of Toronto. The original
city model is composed of several 3-D massing AutoCAD
files in .dwg format, which is not suitable for HoloLens, so we
converted them into .fbx 3-D files. The open data sets cover a
variety of city areas, including transportation, city government, business, and health, among others. Since they are
stored in several file formats, such as extensible markup language (XML), comma-separated values, and excel spreadsheet, we first converted them all into XML files by extracting
meaningful information. Then we selected the data sets with
complete information and processed them for visualization.
Based on the contents of each data set, the processing
steps may be different. The read-me file of each data set usu-

Table 2. The data table description
from a washroom facilities file.
Column Name

Value

Washroom building asset ID

ID number

Park name

Name of a park

Building name

Name of a building

GIS coordinate

(Longitude, latitude)

Access

Public, private, or none

MAy 2018

IEEE Consumer Electronics Magazine

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - May 2018