Integration of dynamic LiDAR and image sensor data for route corridor mapping

Abstract

Building and maintaining modern transportation infrastructure demands considerable expenditure for any nation. These terrestrial route corridor zones include road, rail and to a lesser extent waterways. Road networks range from the large highways and motorways covering hundreds of kilometres down to smaller street networks that may only be few hundred meters in length. These route networks attract their own unique set of spatial information requirements in terms of overall management. These include transportation planning, engineering and operation. High quality, timely spatial information is required of the entire route corridor which now extends past the narrow confines of the road surface and includes the area adjacent to the road edge as well as areas above and below the road surface. Comprehensive 3D spatial information is required, not only, of the network itself but also objects occurring along these route corridors. This information can be used to address the day to day engineering problems as well as more strategic issues such as road safety, congestions management and noise modelling. LiDAR systems are widely available and now used to record data from both aerial and terrestrial survey platforms. LiDAR outputs X,Y,Z points enabling reliable 3-D measurements as well as 2.5-D geometric surface to be produced. High quality imagery is also collected from similar airborne and terrestrial mobile mapping platforms. This paper examines the integration of road survey imagery and airborne LiDAR data-streams within a GIS in order to satisfy these spatial information requirements.

Full text

Integration of dynamic LiDAR and image sensor data for route corridor mapping
Timothy McCarthya, Jianghua Zheng b and A. Stewart Fotheringhama
aNational Centre for Geocomputation, National University of Ireland, Maynooth, Ireland. tim.mccarthy@nuim.ie
bCollege of Resources & Environment Science, Xinjiang University, Urumqi, 830046, Xinjiang, China.
Key Words: Dynamic, Mapping, Mobile, LiDAR, Imagery, GIS, Fusion
ABSTRACT:
Building and maintaining modern transportation infrastructure demands considerable expenditure for any nation. These terrestrial
route corridor zones include road, rail and to a lesser extent waterways. Road networks range from the large highways and
motorways covering hundreds of kilometres down to smaller street networks that may only be few hundred meters in length. These
route networks attract their own unique set of spatial information requirements in terms of overall management. These include
transportation planning, engineering and operation. High quality, timely spatial information is required of the entire route corridor
which now extends past the narrow confines of the road surface and includes the area adjacent to the road edge as well as areas above
and below the road surface. Comprehensive 3D spatial information is required, not only, of the network itself but also objects
occurring along these route corridors. This information can be used to address the day to day engineering problems as well as more
strategic issues such as road safety, congestions management and noise modelling. LiDAR systems are widely available and now
used to record data from both aerial and terrestrial survey platforms. LiDAR outputs X,Y,Z points enabling reliable 3-D
measurements as well as 2.5-D geometric surface to be produced. High quality imagery is also collected from similar airborne and
terrestrial mobile mapping platforms. This paper examines the integration of road survey imagery and airborne LiDAR data-streams
within a GIS in order to satisfy these spatial information requirements.
Background
The importance of transportation networks is well documented
and information across a broad spectrum is required to manage
various activities that occur along these transportation corridors
(McCarthy et al, 2007). This information is required for a
variety of activities including; road maintenance, pavement
condition, street furniture upgrade, safety analysis, road user
charging and noise modelling. The importance of this
information is borne out by a recent compilation of a
specialised base mapping datasets by national mapping
agencies, the creation of national standards for route corridor
asset registers as well as the increase in specialist road network
asset inventory software providers (McCarthy, 2007).
Geosptial data can be collected by a variety of remote sensing
methods including; spaceborne, airborne and terrestrial sensor
systems. Terrestrial-based systems include stereoscopic cameras
mounted on road survey vehicles and airborne systems
including LiDAR. Stereoscopic camera systems, usually
mounted orthogonal to direction of travel, collect image data
enabling 3-D, in-frame measurements to be extracted. These,
together with any visual data, such as road-sign damage can be
stored in a database. Airborne LiDAR systems acquires XYZ
point data using a vertically pointing sensor along the route
network enabling a high resolution 2.5-D point-cloud structure
of the route corridor to be constructed. These two sensors
produce datasets with intrinsically different spatial properties
with associated strengths and weaknesses in terms of the spatial
data recorded. In this investigation, stereoscopic data collected
from a road survey vehicle was integrated with airborne LiDAR
data within a dynamic GIS environment. This study builds on
initial work carried out at the National Centre for
Geocomputation in 2007 (McCarthy, 2007).
Sensor Systems
Ohio State University’s Centre for Mapping was one of the first
research groups to pioneer the development of dynamic
stereoscopic image mapping systems for route corridor mapping
in the mid-1990s (Blaho and Toth, 1995; Bosler and Toth,
1995; Bosler and Toth, 1996; and Jeyapalan, 2004).
Developments extended beyond stereoscopic image collection
and measurement to include automatic feature extraction (Habib
et al., 1999; Habib 2000; Tao 2000; Tau 2001 and Toth and
Grejner-Brzezinska, 2004). Mobile stereoscopic image mapping
technology is now reasonably well established. This is borne
out by the wide ranging technologies and applications presented
at the recent fifth international mobile mapping symposium, 28th
to 30th May 2007, in Padua, Italy (MMT, 2007). There are a
number of companies throughout the world offering this as a
commercial service (McCarthy, 2007). RouteMapper is a
typical example of one of these systems (McCarthy et al, 2007).
This system is in operation in Europe and some of the datasets
acquired by this system have been used to investigate the
usefulness of integrating stereoscopic imagery with LiDAR
data. In summary, the RouteMapper© system comprises four
progressive scan cameras (1392*1024), a navigation unit,
triggering modules and data logging capability. The system is
calibrated any time the cameras are moved and resulting
calibration transformation enables either monoscopic or
stereoscopic measurements to be performed.
The mapping and analysis software, developed by the main
author, comprises image, 2D map and database displays
together with associated toolbars and drop down menus, Figure
1. This allows the user to navigate through the recorded data
using interactive video controls or via the mapping interface.
The user can click the play button and view all four cameras
whilst position of survey van updates dynamically in a moving
1125

map display. 3-D in-frame measurements can be carried out,
recording both dimensional as well as positional information.
These measurements are usually confined to a 3D wedge shaped
volume measuring 30m X 30m X 7m directly in front of the
survey vehicle. This together with any additional attribute can
be stored in the survey database. Standard GIS functionality is
available including spatial and aspatial query. The browser is
lightweight and designed so that users can learn basic functions
in a very short time.
Figure 1. Road survey image mapping system depicting
imagery and map displays.
Additional modules have been developed to enable specialist
asset register construction. A wider audience can access this
data over a recently developed, easy-to-use server based
component. This can be particularly useful if an organisation
comprises many departments such as road planning,
maintenance and operations.
Airborne LiDAR has been available since the late-1990s and
has been accepted as an accurate, effective method for data
collection (Iavarone, 2005). These high-resolution XYZ point
data can be collected during the day or night onboard survey
aircraft. LiDAR data acquisition has been well documented for
a range of applications (Kidder et al 2004 and Veneziano et al.,
2002). Very fast airborne LiDAR scanning technology at rates
of up to 167kHz (Optech, 2008) enable reasonably large swaths
of ground to be surveyed in a short time for a variety of end-
user applications including flood-plain mapping, utilities,
transportations and municipal surveying (Hill et al., 2000).
Data processing still demands a reasonable amount of manual
input but the resulting information content is also quite high.
Data volume also needs to be taken into account with a typical
4km X 4km survey resulting in 64 million XYZ & intensity
value points at 50cm sample spacing. Airborne LiDAR has also
been used for route corridor design (Uddin 2002 and Veneziano
et al., 2002) and route inventory (Shamayleh et al., 2003). In all
cases, LiDAR has been found to increase mapping efficiency
whether it is for planned routes or mapping out existing
infrastructure.
The test section chosen for this study was along the A421
primary road in UK. This data was collected 6th May 2007.
Corresponding airborne LiDAR was required for integration
and evaluation. Environment Agency (UK) provided processed
1m resolution, LiDAR dataset for this study. The airborne
point-cloud dataset was originally acquired over UK Midlands
23rd November 2006 (EA, 2008). The dataset is part of a vast
archive of existing airborne LiDAR data acquired over various
dates, at varying resolutions. This repository extends over a
large part of England and Wales, These data are available to the
researchers and public alike. For researchers, this is an
invaluable source of high quality, base mapping data that can be
put to a range of uses. Quite a number of countries across the
globe are also building up large archives of airborne LiDAR.
Extending the use of these data can help lower the up-front cost
of collecting and processing LiDAR.
Integration of stereoscopic imagery and airborne LiDAR
A number of researchers have examined the advantages of
integrating LiDAR with other datasets within a GIS. Kidder et
al (2004) carried out an evaluation of methodologies employed
to make LiDAR compatible, consistent and useable within a
GIS. These focused on data handling, error detection and
geodetic transformation. One of the chief conclusions centred
on understanding the errors in LiDAR data and advised further
research before wholly relying on this dataset for certain
applications. Kressler et al. (2006) integrated LiDAR, image
data and spatial databases to produce a higher resolution
building/land classification map. Rottensteiner et al. (2003)
used aerial imagery to aid building outline extraction and
recommended further examination of GIS datasets for assessing
data quality. In all cases, GIS was perceived as useful whether
utilised as a spatial repository, aiding LiDAR processing, or by
providing an environment for checking data quality. This
research project followed on from preliminary evaluation of
imagery and LiDAR integration initiated at NCG in 2007
(McCarthy 2007), concentrating on value of integrating ground
imagery and LiDAR geometric datasets. Various information
regarding road side assets and infrastructure is required for
efficient management. These include information relating to
object location, dimension and condition state. Objects and
features include bridges, walls, street-lamps, traffic-sign, road-
markings, drainage, road-reflectors, pavement-condition, grade
and camber (McCarthy, 2007).
A prototype browser enabling LiDAR, conventional 2D base-
maps and stereoscopic data to be integrated together within a
GIS was designed. This entailed modifying the software to
allow a point cloud module to be integrated within existing
browser, Figure 2.
Figure 2. Browser depicting stereoscopic, mapping and LiDAR
display, approach a roundabout on the A421 in UK.
A range of functions such as loading external XYZ points,
setting viewing geometry, measurement extraction and
graphical drawing were used to integrate the LiDAR point cloud
environment with image space. The test data covered a 10km
section of road that was surveyed by both road image mapping
data as well as airborne LiDAR. The road survey vehicle
acquired imagery using four forward facing cameras; one
camera pointing to the left, the next two, in stereoscopic
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008
1126

configuration, pointing forward with the final camera pointing
to the right. All three data displays were linked with the default
master position set to the image module since the camera
position field of view were fixed at the time of survey. The user
can control navigation through the integrated image and LiDAR
datasets in any of the three display environments using mouse
controls or toolbar buttons along the tool bar. This allows the
user zoom, pan, query and measure. Measurements in one
window were mirrored in the other two display environments.
For example, measurements can be carried out in the image
window with the resulting point or line plotted in all three
displays. Measurement of objects in 2D map display was
confined to the XY plane.
Useful measurement range within the road image survey system
was confined to a 3D wedge shaped volume extending
outwards, in front of the vehicle, 30m along a centre line, +/-
15m to the left and right of this centre line and 10m upwards in
height. The user was alerted anytime measurements fell outside
this calibration volume since results could be erroneous. Objects
in any display could be identified, measured and associated
attributes stored in the main browser database. Later, these same
stored objects could be picked from a table and their associated
location, dimensions and attributes could be instantly displayed.
This action caused all three mapping environments to display
retrieved object’s position and/or dimensions.
Results and Discussion
Imagery was collected at rates of 4Hz whilst travelling at speeds
of between 75km/hr and 100km/hr. LiDAR, on the other hand,
was collected by a downward pointing scanning sensor. The
LiDAR data was processed into geometric blocks covering large
tracts of topography, infrastructure and buildings. Loading route
data using this conventional data structure can be inefficient
since the user is typically only interested in a 3D corridor
represented by the route centre line and extending typically +/-
100m on either side. Some pre-processing is required to subset
out this route corridor from a much larger dataset. Bringing road
survey imagery and LiDAR datasets together within a GIS
results in a more dynamic mapping environment. The static
LiDAR display needs to update user position and mimic the
smooth movement along the route corridor, updating the view-
shed each time the survey position and camera images refresh.
LiDAR and image data acquire different scene properties in
terms of thematic attribute, resolution and field of view so,
choosing various viewing geometries to handle both displays is
important. The default user interaction mode naturally follows
the path taken by the road survey imaging system since this is
fixed at the time of survey. The default viewing geometry needs
to take into account the various data acquisition parameters
including; camera models, sensing geometry, resolution of both
survey cameras and LiDAR instruments In this test, user
position in the point-cloud display had to be elevated and field
of view (fov) extended in order to orientate the user and carry
out measurements. The 1m resolution of the point-cloud made it
difficult to instantly spot correspondence in terms of visual cues
between LiDAR and more visually rich image data.
The obvious differences between road survey imagery and
airborne LiDAR deal with each sensor’s ability to record
various scene content properties, Figure 3. Road survey cameras
collect data orthogonally as the vehicle moves along the road.
Imaging range is controlled by horizontal and vertical fov as
well as vehicle position on the road network.
Figure 3. Browser depicting scene content differences between
image and LiDAR sensors datasets
This tends to limit the view of the came to the immediate road
corridor, in front of the vehicle, if pointing forward. There is
very little chance of recording data behind walls, over bridges
or on top of embankments. High quality imagery with
resolution at the centimetre level can be acquired even at speeds
of 100km/hr. This enables the 1cm cracks on the road pavement
surface to be viewed and measured. Information on road signs
can be read, as can their condition, Figure 4.
Figure 4. Road-sign at full resolution, taken from main image
display, Figure3, illustrating detail recorded by survey vehicle.
The advantages and disadvantages of the two sensors in terms
of collecting useful road network information are summarised in
Table 1.
Sensor Advantages Disadvantages
Road
Survey
Imagery
Very high resolution
detail View-shed limited to
number of cameras
available
Multiple view sheds Unable to view
behind walls, tops of
bridges,
embankments
Record data under
tree canopy, tunnels
and bridges
Affected by
environmental
conditions such as
sunlight, rain, low
illumination
conditions
Comprehensive
measurement and
attribute information
available
Poor measurement
accuracy sometimes
results from poor
navigation eg long
tunnels, heavy traffic
in city centre.
Offers a very realistic
view of immediate
route corridor
Airborne
LiDAR Uniform, synoptic
dataset Relatively coarse 1m
resolution
Ability to record
sight-lines, camber,
grade, curvature over
long lengths along
road lines
A lot of standard
roadside features
missing such as street
lamps, barriers, road
markings
Table 1. Advantages and disadvantages of sensor datasets
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008
1127

The position, dimensions and state of road-side safety barriers,
communication cabinets, street lamps, traffic signs and drainage
infrastructure can usually be recorded using one of the multiple
camera views. Safety barriers, communication, drainage and
ventilation systems under bridges and in tunnels can be imaged
and mapped. In contrast, point-cloud data acquired by airborne
LiDAR, is a poor second at first glance. Compared to the
visually rich spatially encoded imagery, the 1m resolution
point-cloud seems quite sparse, with very little detail, devoid of
obvious road-side objects like road markings, road signs, street-
lamps, barriers and kerb-stone. This LiDAR dataset was
collected from an aerial platform a few thousand feet above the
ground. Current sensor technology onboard airborne platforms
at these sorts of altitudes will be limited when compared to road
survey systems. A lot of routes in the Northern Hemisphere are
surrounded by trees and hedges. Downtown city areas can have
very tall buildings and relatively narrow thoroughfares. These
natural and man-made features often hinder comprehensive
surveying from the air.
However, LiDAR enables large feature classification and
measurement including those outside road survey cameras’ field
of view. Road lengths, widths, camber, curvature, grade can be
easily measured. Blind-spots along linear route sections can be
computed. Embankments, bridges, roundabouts can be
measured. Offsets from road centre-line to buildings behind
walls can readily be recorded. All of these mapping tasks would
be impossible to carry out using the road image mapping system
on its own. It is certainly more efficient to measure, classify and
record details of road side features using this integrated
approach. Measurement accuracy was in close agreement where
objects were clearly identifiable and within range in both image
and LiDAR datasets. In one example a bridge was identified
and measured as 6.8m above the ground, Figure 5.
Figure 5. Browser depicting integrated measurement.
The main overall shortcoming of road survey image data relate
to the relatively short effective measurement range and the lack
of an overall, synoptic view. The chief shortcoming of this
LiDAR dataset is the associated relatively coarse 1m resolution.
However in most cases, the shortcomings of one dataset are
handled by the other. A summary of advantages of integration
are listed in Table 2. Having the capability to analyse both
stereo imagery and LiDAR datasets within a GIS meant that
objects and features could be classified with additional
descriptor information and with greater certainty. Work-flow, in
terms of understanding, classifying and measuring the network
becomes faster. The two datasets intrinsically integrated extends
the usefulness of the browser enabling other engineering
disciplines make use of the same data.
Browser Feature Comment
Realistic
representation of
route corridor
Combined image and point-cloud
display result in more comprehensive
representation of route
Object measurement More comprehensive record of road-
and classification side asset since user can see detail at
road side as well as behind walls,
embankments. Accuracy of
measurements of large items can be
checked
Work-flow Much faster work flow since large
objects (bridges, embankments) can
be measured using LiDAR and
smaller objects located and measured
using road survey imagery. Less time
spent checking accuracy of
measurement.
Extended usefulness Ability to classify and measure
relevant features within a 100m
distance of road centre line extends
the usefulness of this route mapping
tool eg Noise Mapping
Table 2. Listing of advantages of integrating road survey
imagery and airborne LiDAR within a GIS
Integrating road survey imagery and aerial point-clouds poses
some questions in how these data should be handled. Is XYZ
point-cloud data the most efficient data structure to store and
display this data? Should some attempt be made to use
information from both image and point-cloud data to help
automate or semi-automate infrastructure mapping along route
corridors? How are differences in sensor acquisition geometries
dealt with? These questions are more easily handled within a
more advanced data fusion investigation. A number of
researchers have examined data fusion of image and LiDAR
datasets. These research projects dealt with data collected from
similar platforms and highlight some of the problems such as
modelling scene objects, handling occlusions and dealing with
multi-sensor registration (Habib et al., 2005; Habib et al., 2004;
Iavarone, 2005; Lee et al, 2008; Zhang et al. 2006 and Youn et
al., 2008). Therefore, additional hurdles need to be overcome in
order to develop future data fusion techniques for multi-sensor
datasets collected from multi-platforms.
In parallel, dynamic road based terrestrial LiDAR survey
systems are being developed commercially to collect route
corridor data from moving survey vehicles (Geospatial, 2007;
3DLaserMapping, 2008). One recent system is based on the
200kHz pulse rate, RIEGL LMS-Q560 sensor. This system has
an effective measurement rate of 100kHz with a range of
between 30m 1800m and associated accuracy of 20mm. These
can be combined with 39MP, calibrated digital cameras for
route corridor surveying. User requirements are increasingly
seeking high resolution, 3D models of transportation networks
together with timely information pertaining to road-side asset
condition for a variety of engineering, management and
strategic planning tasks. These involve developing new
algorithms to increase automation in road asset inventory to
help reconstruct the network and associated infrastructure in 3D
as well as the ability to detect and record change. All of these
developments indicate the growing interest in data fusion and
development of automated asset-focused change detection
algorithms using multiple, cross platform imagery and LiDAR
datasets.
Conclusions
Road networks are expensive to build and maintain but occupy
a strategic position in any modern nation’s infrastructure
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008
1128

inventory. Road survey imaging systems, although a relatively
recent technology, now play a significant role in mapping route
networks throughout the world. Airborne LiDAR data are
becoming increasingly available and provide an additional
mapping layer. Bringing these two datasets together in a GIS
enables a more comprehensive 3D representation of the route
corridor as well as increased efficiencies in terms of mapping
the underlying infrastructure. A prototype browser was designed
and assessed in terms of user interaction, object measurement
and classification, work flow and overall usefulness.
Shortcomings, in terms of road inventory information, arising
from one sensor are handled by the other sensor’s dataset. The
resulting integrated approach extends the usefulness of the
datasets enabling additional engineering tasks to be
accomplished such as road-safety assessment and noise
modelling to be carried out.
Data fusion techniques are topics under investigation by a
number of researchers. These are focused around multi-sensor
datasets acquired from similar platforms. These techniques will
need to be extended to handle multi-sensor, multi-platform
datasets. Integrating these two multi-sensor, multi-platform
datasets provides an environment to understand the various
issues. New, rapid scanning, terrestrial LiDAR data acquisition
systems integrated with 39 MP digital cameras already exist.
These will continue to produce vast quantities of data. Data
fusion and automated change detection algorithms will play a
pivotal role in ensuring that data processing and geo-
information generation keeps pace with developments in the
data acquisition field. All of these factors have to match the
rising expectations of users in terms of more comprehensive 3D
representations and automation in not only mapping route
infrastructure but also identifying and recording change in a
timely, cost-effective fashion. Integrating these two datasets
within a GIS, albeit a small step, is nevertheless a step closer
towards attaining these goals.
References
3DLaserMapping 2008. Dynamic LiDAR mapping.
http://www.3dlasermapping.com/uk/mobile/streetmapper/overvi
ew.htm
Blaho, G., and Toth, C. 1995 Field Experiences with a Fully
Digital Mobile Stereo Image Acquisition System, Proc. ASPRS-
ACSM Annual Convention, Vol. 2, pp. 97-104, 1995.
Bossler, J. D., and Toth, 1995 C. Accuracies Obtained by the
GPSVan™, Proceedings, IS/LIS’95, Vol. 1., pp. 70-77, 1995.
Bossler, J., and Toth, C. 1996. Feature Collection by the
GPSVan™, Analog Video or Stereo Digital Video?,
Proceedings, ASPRS-ACSM Annual Convention, Vol. 2, pp.
108-115, 1996.
EA 2008. Environment Agency (UK). LiDAR Data
Acquisition. www.environment-agency.gov.uk.
Geospatial 2007. Survey of 566km of road using LiDAR
mounted on a survey vehicle. http://eirs.geospatial-
solutions.com/gsseirs/article/articleDetail.jsp?id=338981
Habib, A., Uebbing, R., Novak, K. 1999. Automatic Extraction
of Road Signs from Color Terrestrial Imagery, Journal of
Photogrammetric Engineering and Remote Sensing, Vol. 65, pp.
597-601, 1999.
Habib A. 2000. Matching road edges in stereo-image sequences
using data association techniques. PERS. January 2000. pp 91-
98
Habib A, Ghanma M and Tait M. 2004. Integration of LiDAR
and photogrammetry for close range applications. ISPRS.
Comm IV. Integration and fusion of data models. Istanbul 2004.
Habib A, Ghanma M, Mitishita E, Kim E and Kim C. 2005.
Image georeferencing using LiDAR data. Geoscience and
Remote Sensing Symposium, 2005. IGARSS '05. Proceedings.
2005 IEEE International
Hill J, Graham L and Henry R. 2000 Wide area topographic
mapping and applications using airborne LiDAR technology.
PERS. August 2000. pp 908-914.
Hudson, W, Haas R and Uddin W 1991 Infrastructure
Management: Design, Construction, Maintenance,
Rehabilitation, Renovation. McGraw-Hill, 1997
Iavarone A 2005. Merging airborne and ground based LiDAR
data. PERS. September 2005. pp 1001-1003
Jeyapalan K 2004. Mobile digital cameras for as-built surveys
of roadside features. PERS. March 2004. pp 301-312.
Kider D, Thomas M, Leight C, Oliver R and Morgan C. 2004.
Coastal monitoring with LiDAR: Challenges, problems and
pitfalls. Remote Sensing for Environmental Monitoring, GIS
Applications and Geology IV, edited by Manfred ehlers,
Francesco posa, Hermann j Kaufmann, Ulrich Michel, Glacomo
De Carrolis, Proceedings of SPIE Vol 5574. SPIE Bellingham
WA 2004. pp 80 – 89.
Kressler F and Steinnocher K. 2006. Image classification with
LiDAR and GIS data – moving from land cover to land use.
Remote Sensing for environmental monitoring, GIS Appls. and
Geology IV, edited by Mafred Ehlers, Ulrich Michel, Proc. Of
SPIE Vol 6366. 2006.
Lee H, Lee K, Lee S. 2008. Fusion of LiDAR and Imagery for
relaibale building extraction. PERS. Vol.74, No. 2, Feb 2008.
pp 215-225.
McCarthy T., Fotheringham A.S. Charlton M., Winstanley A.
and O’Malley V. Integration of LiDAR and Stereoscopic
Imagery for Route Corridor Surveying. MMT 2007. The 5th
International Symposium on Mobile Mapping Technologies.
CD-ROM. 28th - 30th May 2007, Padua, Italy
MMT 2007. The 5th International Symposium on Mobile
Mapping Technologies. CD-ROM.
http://www.cirgeo.unipd.it/cirgeo/convegni/mmt2007. 28th -
30th May 2007, Padua, Italy
Optech 2008. Gemini 167kHz ALTM.
http://www.optech.ca/prodlatm.htm
Rottensteiner F and Briese C. 2003. Automatic generation of
building models from LiDAR data and the integration of aerial
images. ISPRS Vol XXXIV. Dresden 2003.
RouteMapper 2008 Stereoscopic image mapping system.
http://www.routemapper.net/
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008
1129

Shamayleh H and Khattak A. 2003. Utilization of Lidar
technology for highway inventory. Proceedings of the 2003
Mid-Continent Transportation Research Symposium, Ames,
Iowa, August 2003.
Tao V 2001. Database-guided automatic inspection of vertically
structured transportation objects from mobile mapping image
sequences. PERS. December 2001. pp1401-1409
Tao V 2000. Semi-automated object measurement using
multiple-image matching from mobile mapping image
sequences. PERS. December 2000. pp 1477-1485.
Toth C and Grejner-Brzezinska D 2004. Redefining the
paradigm of modern mobile mapping: An automated high-
precision road centreline mapping system. PERS. June 2004. pp
685-694
Uddin W. Evaluation of Airborne LiDAR digital terrain
mapping for highway corridor planning and design. . Integrated
Remote Sensing at the Global, Regional and Local Scale.
ISPRS Comission I. Mid-Term Symposium in conjunction with
Pecora 15/Land Satellite Information IV Conference. 10-15
November 2002, Denver, CO USA
Veneziano D, Souleyrette R and Hallmark S. 2002. Evaluation
of LiDAR for highway planning, location and design. Integrated
Remote Sensing at the Global, Regional and Local Scale.
ISPRS Comission I. Mid-Term Symposium in conjunction with
Pecora 15/Land Satellite Information IV Conference. 10-15
November 2002, Denver, CO USA
Youn J., Bethel J, Mikhail E. and Lee C. 2008. Extracting urban
road networks from high resolution true orthoimage and
LiDAR. PERS. Vol 74, No. 2. February 2008. pp 227 – 237.
Zhang W and Li Q. 2006. A preliminary simulation to study the
potential integration of LiDAR and imagery. Remote Sensing
for environmental monitoring, GIS Appls. and Geology VI,
edited by Manfred Ehlers, Ulrich Michel, Proc. Of SPIE Vol
6366. 2006
Acknowledgements
Environment Agency (UK), Geomatics Group, Phoenix
House, Lower Bristol Road, Bath. BA2 9ES for
providing the test airborne LiDAR dataset.
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008
1130