Conference PaperPDF Available

Mobile mapping system performance – an initial investigation into the effect of vehicle speed on laser scan lines

Authors:

Abstract and Figures

Mobile mapping systems (MMS) are becoming an increasingly popular method for collecting high quality near-3D information in terrestrial environments. One of the primary reasons for this is technological advances in laser scanning. When a modern laser scanner is mounted on a moving platform and combined with a GPS and navigation system, mobile mapping systems can produce millions of geo-referenced points per minute which can then be used to create accurate near-3D models. The development of processing algorithms for these point clouds has been the focus of the research community to date. However, given an arbitrary known static object positioned at a specific distance away from a moving mobile mapping system the resolution and accuracy of the resulting point cloud which will describe the object is unknown. It is this resolution and accuracy which is the underlying limit of these point cloud processing algorithms. We are in the process of developing a method for determining the quantitative resolution and accuracy of point clouds collected by a mobile mapping system with respect to known objects. In this paper, we will demonstrate our initial investigation into the effect that vehicle speed has on laser scan lines. Speed has an impact on the physical distance between sequential laser scan lines and will also influence the angle of individual scan lines. We have developed a system to calculate this information about laser scan lines based on the position and orientation of the laser scanner on the vehicle and the vehicles speed. We will verify our equations and analysis by comparing our simulated data to the point cloud data collected by our XP-1 mobile mapping system.
Content may be subject to copyright.
Mobile mapping system performance an initial investigation
into the effect of vehicle speed on laser scan lines
C.Cahalane, T.McCarthy and C.McElhinney
National Centre for Geocomputation, NUI Maynooth, Kildare, Ireland
Email: conor.cahalane.2010@nuim.ie
Abstract
Mobile mapping systems (MMS) are becoming an increasingly popular method for collecting
high quality near-3D information in terrestrial environments. One of the primary reasons for
this is technological advances in laser scanning. When a modern laser scanner is mounted on
a moving platform and combined with a GPS and navigation system, mobile mapping
systems can produce millions of geo-referenced points per minute which can then be used to
create accurate near-3D models. The development of processing algorithms for these point
clouds has been the focus of the research community to date. However, given an arbitrary
known static object positioned at a specific distance away from a moving mobile mapping
system the resolution and accuracy of the resulting point cloud which will describe the object
is unknown. It is this resolution and accuracy which is the underlying limit of these point
cloud processing algorithms. We are in the process of developing a method for determining
the quantitative resolution and accuracy of point clouds collected by a mobile mapping
system with respect to known objects. In this paper, we will demonstrate our initial
investigation into the effect that vehicle speed has on laser scan lines. Speed has an impact on
the physical distance between sequential laser scan lines and will also influence the angle of
individual scan lines. We have developed a system to calculate this information about laser
scan lines based on the position and orientation of the laser scanner on the vehicle and the
vehicles speed. We will verify our equations and analysis by comparing our simulated data to
the point cloud data collected by our XP-1 mobile mapping system.
1 Introduction
The focus of the research community to date has largely been on developing automated or
semi automated algorithms for processing the large point clouds captured by modern
terrestrial or mobile mapping systems (Becker and Haala, 2009, Hammoudi et al., 2009, Pu
and Vosselman, 2007). However, other than accuracy tests on specific systems (Barber et al.,
2008, Haala et al., 2008) little research exists assessing the performance of generic mobile
mapping systems. Further research in this area is important as one of the underlying problems
facing research groups working with extraction algorithms is how many scan lines (or
profiles) will hit an object at a certain range, and how many points can they expect to return
from each profile. For example, work by Kukko et al., (2009) and Lehtomaki et al., (2010)
require a minimum number of profiles on post objects for them to be detected. Circular
objects need a minimum number of points on each profile to recognise a circular shape. Each
algorithm performs differently, and from Kaartinen et al., (2005) we can see that the point
density directly impacts on the accuracy of the resulting extracted model. Mobile mapping
systems (MMS) are new to the market, and to date there has been no concerted effort to
assess their combined capabilities. This paper will focus entirely on laser based systems.
One of the fundamentals for a laser based mobile mapping system is the location and
orientation of the scanner on the vehicle. Although there have been tests investigating the
best scanner configuration to minimise occlusions (Yoo et al., 2009), there does not appear to
have been research carried out to find the optimal location for a single scanner (i.e. rear, side,
front) that will provide the highest point density. The solution to date has been to increase the
number of scanners. Our system is a single scanner system, so we hope to provide a definitive
view of the capabilities of such a system which we anticipate will then be of use to systems
with more hardware. Scanner orientation is also of importance. Scan lines cannot be
perpendicular to the direction of travel or they will miss objects whose sides are also
perpendicular to it. A horizontal rotation of the scanner solves this problem, and a vertical
rotation deals with structures above the vehicle which would otherwise be missed, such as
overhead road signs, bridge faces etc. We hope to be able to define what the optimum
orientation is.
When safe to do so, mobile mapping systems are capable of operating at highway speeds.
However, point density decreases as vehicle velocity increases and this necessitates multiple
passes to ensure a dense point cloud (multiple passes are also employed to ensure all sides of
an object are captured) that will meet project specifications. To ensure a high point density,
projects have been carried out at low speed (Goulette et al., 2006, Graefe, 2007), which in a
commercial situation would impact on the productivity of a MMS. It is our hope that when
completed our work will allow us define the maximum speed for specific scanner
configurations that will provide a required point density, and also define the minimum
number of passes required. This should help to minimise survey time and also the file size of
each survey.
To date there has been some interesting work in this area. Kukko et al., (2007) and Hesse and
Kutterer (2007) have detailed profile spacing at various mirror speeds and vehicle velocities.
We hope to improve on this by providing a generic formula which will work for any mirror
speed, vehicle velocity and importantly, will incorporate scanner orientation into the system.
Hoffman and Brenner (2009) have included in their work on theoretic point density some
interesting results on the effect change in vehicle direction and velocity has on scan lines.
In section 2 we will look at mobile mapping systems in general and the platform we have
developed at StratAG, followed in section 3 by the theory and processes behind our work. In
section 4 we will present the results of our test data, and finally in section 5, our conclusions.
2 Mobile Mapping & Experimental Platform
MMS enable high density spatial data to be collected along route networks and in urban
environments. These data can then be utilised in a number of ways, such as route safety
audits, road authorities GIS, infrastructure surveys and change detection for national mapping
agencies. Combining high accuracy GPS/INS, LiDAR and imaging sensors onboard a
moving platform enable surveys to be carried out rapidly with significant cost savings. Land
based MMS compliment existing ground based survey and aerial surveying activities in a
number of ways. Large scale detail such as road sign detail or detailed infrastructure
condition can be recorded. Additionally, extensive ground control is not required and these
systems can capture features that are sometimes obscured from aerial platforms.
The multi-disciplinary research group StratAG, established to research advanced
geotechnologies at NUI Maynooth have recently completed design and development of a
multipurpose, state of the art, land based experimental platform (XP-1) Mobile Mapping
System as shown in Figure 1. The primary components of the XP-1 are an IXEA LandINS
GPS/INS, a Riegl VQ-250 300KHz laser scanner and an imaging system consisting of 6
progressive-scan cameras. Additional imaging sensors include a FLIR thermal (un-cooled)
SC-660 camera and an innovative 5-CCD multispectral camera capable of sensing across
blue, green, red and two infrared bandwidths.
Figure 1. The XP1 Mobile Mapping System.
3 Laser scan line theory
This paper details our initial investigation into the effects of vehicle speed on scan lines for
certain scanner mirror speeds, positions and orientations. For these initial steps, a number of
necessary assumptions were made. The first assumption was that the vehicle was operating
on a plane surface, so road cross fall or surface deformities could be ignored. The second was
that the vehicle maintains a constant speed between each laser profile. The final assumption
was that we would not take into account occlusions, traffic or obstructions. Once these
assumptions had been made, the factors impacting upon scan lines were identified.
3.1 Profile spacing
Before exploring the effect of scanner rotations on profile spacing, it is important to define
what we mean by a vertical and horizontal rotation of the laser scanner. On the XP1, the
scanner is located at the rear of the vehicle. Figure 2, displays the our laser
scanner/GNSS/INS mount, and the 45° horizontal and 45° vertical rotations of the scanner. It
can be seen in Figures 3 and 4 the effect that horizontal and vertical rotations of the scanner
have on profile spacing. The distance travelled, d, is no longer the profile spacing. The profile
spacing is now a function of the scanner angle, mirror speed and vehicle velocity.
Figure 2. XP1 Laser scanner rotations: (a) top view (b) side view.
b
a
Figure 3. Profile effect due to vertical scanner rotation: (a) overview (b) detail view
Figure 4. Profile effect due to horizontal scanner rotation: (a) overview (b) detail view
Once the variables affecting the profile spacing were identified (see Table 1), it was possible
to create a formula to calculate profile spacing for a generic system.
Table 1. Profile spacing variables.
Profile Spacing
v
Speed in m/s
f
Mirror rotation frequency
d
Distance travelled between mirror rotations
θh
Laser scanner horizontal-rotation angle
θv
Laser scanner vertical-rotation angle
Ph
Profile distance on ground surface
Pv
Profile distance on vertical structures
To calculate profile spacing, the distance covered in one mirror rotation for any speed is
required (Equation 1).
d = v/f (1)
From this, the effect of the scanner rotation is calculated by trigonometry (Equation 2) for a
right angled triangle, as shown in Figures 3 and 4.
Ph = cosθh.d or Pv = cosθv.d (2)
b
a
b
a
3.2 Motion effect
In this paper, the effect of motion on profiles refers to their angular change due to the speed
of a vehicle. A stationary vehicle exhibits profiles whose angle is entirely dependent on the
scanner angle, however, when the MMS is in motion this will change. In our system, it takes
0.01s for each mirror rotation. During this time the vehicle will have moved, altering the
angle of the scan line, as shown in Figure 5. We have designed a system to calculate this
effect for a generic system, based on scanner height, vertical scanner rotation, vehicle speed
and mirror speed. The variables involved in this calculation can be found in Table 2.
Figure 5. Motion effect: (a) overview (b) detail view
Table 2. Motion effect variables.
Motion effect
θv
Laser scanner vertical-rotation angle
h
Scanner height above ground
d
Distance travelled between mirror rotations
θm
Motion amended angle
For this initial investigation, the effect of motion on profiles can be viewed as part of a group
of parallelograms and triangles (Figure 5). Using trigonometric ratios, the hypotenuse ‘Y’ of
the triangle formed with the vehicle is found (Equation 3). Accurate measurement of the
vehicle height (h) is important.
Y = h / (sin (90° θv) (3)
In this calculation, because using the distance travelled during one full mirror rotation would
mean that we would now be on a new scan line we only deal with the bottom half of the
circular scan, hence the use of d/2. This is followed by use of the cosine rule to calculate the
short diagonal length ‘X’ in the parallelogram (Equation 4). This value is one of the variables
for the last triangle, and by using the sine rule θm is found (Equation 5).
X = √ ((d/2) 2 + (Y)2 2(d/2)(Y)cos (90°- θv)) (4)
θm = sin-1 ((sin (90°- θv) * d/2)/x) (5)
b
a
4 Results
We will demonstrate the capabilities of this method using two datasets. One, a theoretical test
dataset verified with CAD models and the second an actual dataset from a MMS. As our test
dataset, we applied our formula to a selection of velocities, varying mirror rotation speeds
and varying vertical and horizontal rotations of the laser scanner. This theoretical test dataset
was compared against CAD models and in each case it agreed with our results. To improve
on this and provide a real world test of our system, we chose to verify our equations and
hypotheses by comparing our simulated data to a point cloud captured by our XP1 mobile
mapping system. Since relative accuracy and not absolute accuracy is what is important to
this study, we ignored the GPS conditions and quality of the combined navigation solution.
By analysing the navigation files, varying areas velocity from 5m/s to 16m/s were identified.
By plotting this navigation data on the survey data and selecting straight sections of road
only, errors due to course deviation could be kept to a minimum. The reference data exhibited
a point spacing that one would expect from a laser scanner according to its specifications.
This made accurate measurements difficult, so we interpolated a 2d linear fit (shown in
Figure 5) as it was the most suitable approximation of the scan lines that we were
investigating and our matches our assumption that the road is a flat plane.
Figure 6. Scan line measurements: (a) linear interpolation (b) subsequent measurements
After analysing the data, and plotting the predicted versus measured values for ground
profiles(Figure 7), it becomes clear that our system is performing satisfactorily. Measured
profile spacings ranged from 0.031m at 4.1m/s to 0.12m at 16.34 m/s. Overall the predicted
figures matched the measured values, with a minimum percentage error of 0%, an average
percentage error of 2.68% and a maximum of 6%. All measurements were within 5mm of
their predicted value. Outliers can be attributed to errors in measurement of the reference data
or possibly from the linear interpolation process of the scan lines. This is something that we
hope to improve on in the near future.
Initial inspection of the results from the motion effect tests identified an issue (see Figure 8).
However, it soon became apparent that there was a problem with the measured rather than
our predicted values. Minus values for angular changes had been recorded and these should
not have been present unless the vehicle velocity was negative i.e. reversing. This suggests
the error was made when selecting the navigation data for comparison. We believe that a
more rigorous approach to ensuring there was no course deviation between scan lines will
lead to better results in the future. It also became apparent that a θm value of over 90° will
cause our system to fail due to the change of trigonometric quadrant. However, this is
unlikely to happen in practice as the mirror speed would have to be extremely low and the
vehicle speed extremely high, but is worth noting.
b
a
Figure 7. Ground profile spacing due to vehicle speed and 45° horizontal scanner rotation.
Figure 8. Angular change of scan lines due to 45° vertical scanner rotation and vehicle
velocity.
5. Conclusion
This study has proven by validating with real world data that our system is capable of
predicting profile spacing for differing vehicle speeds, mirror frequencies and scanner
rotations. Although initial tests were promising, the results from real world data of our
motion effect tests demonstrates that a more rigorous approach to selecting areas of 0° course
deviation needs to be taken we believe that we have identified the cause of error and will be
able to rectify this in the near future. Further to this, future work will involve eliminating the
error sources in the profile spacing process and attempting to incorporate point spacing into
the profile spacing system.
Acknowledgements
The authors would like to acknowledge the support received from the Irish Research Council
for Science, Engineering and Technology(IRCSET) and the Enterprise Partner, Pavement
Management Systems Ltd.
6 8 10 12 14 16
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
Speed (m/s)
Profile Spacing (m)
Predicted Values
Measured Values
6 8 10 12 14 16
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Speed (m/s)
Angular Change (dec. deg.)
Predicted Values
Measured Values
References
Barber, D. Mills, J. and Smith-Voysey, S., 2008. Geometric validation of a ground-based mobile laser
scanning system, ISPRS Journal of Photogrammetry and Remote Sensing, 63, (1), 128-141.
Becker, S. & Haala, N., 2009. Grammar Supported Facade Reconstruction from Mobile LIDAR
Mapping. In International Archives of Photogrammetry, Remote Sensing and Spatial Information
Sciences, Vol.XXXVIII, Part3/W4 (pp. 65-70). Paris, France.
Goulette, F., Nashashibi, F., Abuhadrous, I., Ammoun, S. and Laurgeau, C., 2006. An integrated on-
board laser range sensing system for on-the-way city and road modelling. In: ISPRS Commission I
Symposium, “From Sensors to Imagery”. Paris, 3-5.7.2006. Internationa Archives of
Photogrammetry, Remote Sensing and Spatial Information Sciences 36 (1B).
Gräfe, G., 2007. Kinematic surveying with static accuracy. In: Grün, A., and H. Kahmen, (Eds.) 8th
Conference on Optical 3-D Measurement Techniques, July 9- 12, 2007 - ETH Zurich, Switzerland.
Haala, N., Peter, M., Kremer, J., Hunter, G., Stuttgart, U. and Cloud, P., 2008. Mobile lidar mapping
for 3d point cloud collection in urban areas a performance test. In: ISPRS Congress, ISPRS, Beijing,
China, pp. 11191124.
Hammoudi, K., Dornaika, F. and Paparoditis, N., 2009. Extracting building footprints from 3D point
clouds using terrestrial laser scanning at street level. In International Archives of Photogrammetry,
Remote Sensing and Spatial Information Sciences, Vol.XXXVIII, Part3/W4 (pp. 65-70). Paris,
France.
Hesse, C. and Kutterer, H., 2007. A mobile mapping system using kinematic terrestrial laser scanning
(KTLS) for image acquisition. In: Grün, A., and H. Kahmen, (Eds.) 8th Conference on Optical 3-D
Measurement Techniques, July 9- 12, 2007 - ETH Zurich, Switzerland.
Kaartinen, H., Hyyppä, J., Gülch, E., Vosselman, G., Hyyppä, H., Matikainen, L., Hofmann, A.D.,
Mäder, U., Persson, Å., Söderman, U., Elmqvist, M., Ruiz, A., Dragoja, M., Flamanc, D., Maillet, G.,
Kersten, T., Carl, J., Hau, R.,Wild, E., Frederiksen, L., Holmgaard, J. and Vester, K., 2005. Accuracy
of 3D city models: EuroSDR com- parison. International Archives of Photogrammetry, Remote Sens-
ing and Spatial Information Sciences 36 (Part 3/W19), 227232.
Kukko, A., Jaakola, A., Lehtomäki, M. and Kaartinen, H., 2009. Mobile mapping system and
computing methods for modelling of road environment. 2009 Joint Urban Remote Sensing Event, 20-
22, May Shanhgai, China.
Kukko, A., Andrei, C.-O., Salminen, V.-M., Kaartinen, H., Chen, Y., Rönnholm, P., Hyyppä, H.,
Hyyppä, J., Chen, R., Haggrén, H., Kosonen, I. and Čapek, K., 2007. Road Environment Mapping
System of the Finnish Geodetic Institute FGI Roamer. In Proceedings of the ISPRS Workshop
„Laser Scanning 2007 and SilviLaser 2007‟, Espoo, Finland, 12-14 Sept. 2007; Rönnholm, P.;
Hyyppä, H.; Hyyppä, J. Eds.; pp. 241-247.
Lehtomäki, M., Jaakola, A., Hyyppä, J., Kukko, A. and Kaartinen, H., 2010. Detection of Vertical
Pole-Like Objects in a Road Environment Using Vehicle-Based Laser Scanning Data. Remote
Sensing, 2(3), 641-664.
Pu, S., and Vosselman, G., 2007. Extracting windows from terrestrial laser scanning. In International
Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. XXXVI, Part
3/W52 (pp. 320-325). Espoo, Finland.
Yoo, H.J., Goulette, F., Senpauroca, J., Lepère and G., 2009. Comparative analysis based on
simulation for the design of laser terrestrial mobile mapping systems, 6th International Symposium on
Mobile Mapping Technology, 2009, Sao Paolo.
... (i) point accuracy and density can be significantly higher than those of ALSs, (ii) point clouds often follow a certain pattern (i.e. produced by the movement of the sensors and the fixed relative perspective from the MLS vehicle), or the fact that (iii) some occlusions are avoided by using more than one sensor [14][15][16][17]. Nevertheless, MLS systems have some disadvantages and limitations. ...
... In the same way, using the MLS system along an ideal rectangular-section tunnel or a very simply-shaped street would create a group of connected straight lines (see Fig. 1B-C). The spatial configuration of those lines depends on the setting parameters of the system, i.e. speed of the vehicle, sensor trajectory, scanner orientation, and sensor measurement and rotation rates [17,23]. ...
... An example of a continuous polyline splitting can be found in Fig. 3. (i) Gaps are identified between points: (10-11), (11)(12), (16)(17) and (19)(20). (ii) Polylines are created by the node strings: (1-10), (11), (12)(13)(14)(15)(16), (17)(18)(19) and (20)(21). ...
Article
Abstract An algorithm for automatic detection of planar and quasi-planar surfaces from Mobile Laser Scanning (MLS) data is proposed. The method uses line clouds for efficient data reduction of point clouds from MLS. The singular geometry of the MLS data on planar surfaces is used to transform the original point cloud into a more structured line cloud, which allows the simplification of the initial data and identification of surfaces by grouping lines. From each profile in the original dataset, strings of aligned points are identified, and a line cloud is defined by the endpoints of these strings. Lines are subsequently grouped following a set of parallelism, proximity and merging rules. The algorithm was tested using an urban dataset, and validated on 27 surfaces, by assessing the correctness and completeness of the point and line grouping. Correctness was, in all the surfaces, higher than 99%, and completeness was 90% on average.
... MIMIC calculates the theoretical position of each laser pulse on the target surface, and therefore random events such as occlusions, multiple returns, different surface types and GNSS/IMU errors are not modelled. MIMIC and the multiple variables that need to be considered when designing a system to calculate point density were first introduced in [28], and the formulae and processes used for its design have been incrementally validated using a selection of real-world MMS point clouds [29][30][31][32][33][34]. MIMIC is compatible with different scanner hardware, scanner configurations, vehicle variables and target variables and provides detailed information in tabular or graphical format (using interpolation techniques designed in [35]) to help visualise point and profile distribution. ...
Article
A Mobile Mapping System (MMS) equipped with laser scanners can collect large volumes of LiDAR data in a short time frame and generate complex 3D models of infrastructure. The performance of the automated algorithms that are developed to extract the infrastructure elements from the point clouds and create these models are largely dependent on the number of pulses striking infrastructure in these clouds. Mobile Mapping Systems have evolved accordingly, adding more and higher specification scanners to achieve the required high point density, however an unanswered question is whether optimising system configuration can achieve similar improvements at no extra cost. This paper presents an approach for improving MMS performance for infrastructure surveys through consideration of scanner orientation, scanner position and scanner operating parameters in a methodology referred to as Geometry Aided Design. A series of tests were designed to measure point cloud characteristics such as point density, point spacing and profile spacing. Three hypothetical MMSs were benchmarked to demonstrate the benefit of Geometry Aided Design for infrastructure surveys. These tests demonstrate that, with the recommended scanner configuration, a MMS, operating one high specification scanner and one low specification scanner, is capable of comparable performance with two high-end systems when benchmarked against a selection of planar, multi-faced and cylindrical targets, resulting in point density improvements in some cases of up to 400%.
... The driver also should maintain a constant speed to the extent possible within the clear constraints of safety and traffic regulations to ensure consistent spacing between scanlines. Cahalane et al. (2010) evaluate the influence of speed on the point density of MTLS point clouds. ...
Book
Manual of Practice 152 is a comprehensive yet general overview to help support education and inform practicing engineers on the important role of the surveying engineer. It provides a much-needed update on the modern practice of surveying and geomatics engineering. Topics include geodesy, coordinate systems and transformations, least squares adjustments and error propagation, modern surveying and remote sensing technology, analysis and establishment of control, geographic building information systems, construction surveying, and best practices. Manual of Practice 152 can be used as a summary and a reference for practicing engineers, surveyors and otherwise, to help provide a solid understanding of the surveying and geomatics field.
... This is very beneficial for shape based recognition of objects in the point clouds. A crucial factor for object extraction from point clouds and classification accuracy is the point density and point distribution, which depends on the one hand on the pulse repetition frequency and on the other hand on the distance between single scan lines, which is a function of the vehicle speed (Cahalane et al., 2010). Work on the extraction of road edges has been reported by McJaakkola et al. (2008). ...
Article
Full-text available
Road safety inspection is currently carried out by time-consuming visual inspection. The latest mobile mapping systems provide an efficient technique for acquiring very dense point clouds along road corridors, so that automated procedures for recognizing and extracting structures can be developed. This paper presents a framework for structure recognition from mobile laser scanned point clouds. It starts with an initial rough classification into three larger categories: ground surface, objects on ground, and objects off ground. Based on a collection of characteristics of point cloud segments like size, shape, orientation and topological relationships, the objects on ground are assigned to more detailed classes such as traffic signs, trees, building walls and barriers. Two mobile laser scanning data sets acquired by different systems are tested with the recognition methods. Performance analyses of the test results are provided to demonstrate the applicability and limits of the methods. While poles are recognized for up to 86%, classification into further categories requires further work and integration with imagery.
... La problématique de l'évaluation de données laser n'étant pas triviale, nous pouvons définir plusieurs critères fournissant des informations sur la qualité des données. Pour leur part [Cahalane et al., 2010], affirment que la performance du système de cartographie mobile doit être étudiée en termes de résolution, d'exactitude et de répétabilité des données acquises. Quant à la résolution, elle se calcule en prenant en compte la répartition des points, à savoir l'espacement entre les profils voisins, mais aussi celui entre les points appartenant au même profil. ...
Article
Full-text available
Over the past few decades, the development of land-based Mobile Mapping Systems (MMS), supported by significant technological progress, has become more prominent. It has been motivated by the growing need for precise geographic information about the environment. In this thesis, we consider approaches for the acquisition of mobile point clouds with topographic quality (centimeter-level accuracy). The aim is to develop techniques for data quality assessment and improvement. In particular, we eliminate the systematic errors inherent to an MMS data using appropriate calibration and registration methods.We describe a novel approach to assess the accuracy and/or the precision of mobile laser point clouds. It is based on the extraction and comparison of line features detected within the urban scene. The computed average distance between corresponding pairs of line segments, taking advantage of a modified Hausdorff distance, is used to evaluate the MMS data with respect to a reference data set. For edge extraction, we propose a method which relies on the intersections between planes modelled via the RANSAC algorithm refined by an analysis of connected components. We also consider an approach to correct point clouds using a line-based rigid registration method. Finally, we study the use of line segments for estimating the boresight angles of a land-based mobile mapping system. We apply our methods to synthetic data and to real data acquired as part of the TerraMobilita project.
Article
Plane detection and segmentation is one of the most crucial tasks in point cloud processing. The output from this process can be used as input for further processing steps, such as modelling, registration and calibration. However, the sparseness and heterogeneity of Mobile Laser Scanning (MLS) point clouds may lead to problems for existing planar surfaces detection and segmentation methods. This paper proposes a new method that can be applicable to detect and segment planar features in sparse and heterogeneous MLS point clouds. This method utilises the scan profile patterns and the planarity values between different neighbouring scan profiles to detect and segment planar surfaces from MLS point clouds. The proposed method is compared to the three most state-of-the-art segmentation methods (e.g. RANSAC, a robust segmentation method based on robust statistics and diagnostic principal component analysis – RDCPA as well as the plane detection method based on line arrangement). Three datasets are used for the validation of the results. The results show that our proposed method outperforms the existing methods in detecting and segmenting planar surfaces in sparse and heterogeneous MLS point clouds. In some instances, the state-of-the-art methods produce incorrect segmentation results for façade details which have a similar orientation, such as for windows and doors within a façade. While RDCPA produces up to 50% of outliers depending on the neighbourhood threshold, another method could not detect such features at all. When dealing with small features such as a target, some algorithms (including RANSAC) were unable to perform segmentation. However, the propose algorithm was demonstrated to detect all planes in the test data sets correctly. The paper shows that these mis-segmentations in other algorithms may lead to significant errors in the registration process of between 1.047 and 1.614 degrees in the angular parameters, whereas the propose method had only resulted in 0.462 degree angular bias. Furthermore, it is not sensitive to the required method parameters as well as the point density of the point clouds.
Conference Paper
Management of the earth structures is important to ensure proper functioning of roads. Given that Japan is home to many earth structures, the basic premise of earth structure management is to prevent traffic hazards by conducting regular inspections. New inspection methods are required to reduce the use of human resources involved in earth structure management. This article discusses the applicability of mobile photogrammetry to the inspection of reinforced soil walls. A photogrammetry tool was fixed to a vehicle to measure the surface geometry of the target. The test wall was an 8.0-m high geogrid reinforced soil wall. Mobile photogrammetry data were validated by comparing the measurements obtained using the proposed method and a classical total station. The results of the mobile photogrammetry field test showed that when the speed of the mobile photogrammetry vehicle was 0 or 30 km/h, measurement errors decreased with increase in the number of pictures used in the analysis. Mobile photogrammetry results using six pictures were less affected by vehicle speed. These results support the use of the proposed system in future inspections of earth structures.
Thesis
Full-text available
Mass data collection systems, both based on images and LiDAR (Light Detection and Ranging), gather field information that can often be visually interpreted. However, this visual inspection is very time and resource intensive, to a point that frequently it is not worth using this technology. For this reason, there is an increasing trend towards the automatization of feature interpretation and identification processes, which are materialized through the creation of algorithms that perform these tasks from images and point clouds. The design of these algorithms consists of: (i) the formulation of accurate abstractions of the goal features (i.e. attributes and descriptors that allow their identification and separation from other features), and (ii) the computation or implementation, involving the transformation of the aforesaid abstractions into a set of computer instructions. In the last few years, the use of terrestrial laser scanners have increased substantially, especially with respect to mobile terrestrial laser scanners. These systems gather hundreds of thousands of points per second on the objects and surfaces around them while the mobile platform is moving, and they overcome some of the disadvantages of the static terrestrial laser scanners and aerial LiDAR. Usually, mobile laser scanners provide point clouds with centimetric accuracy, with a highly variable distribution and point density (according to distances and incidence angles to the surfaces and objects). The point clouds are affected by the occlusions caused by the perspective obtained from a specific mobile point, which is usually close to the ground. In this PhD thesis, two data structures for reducing the effects of the large amount of data and the heterogeneity of the point distribution are developed: (i) voxelization, and (ii) line clouds. These two structures allow the creation of reduced and homogenized versions of the point cloud, and have the advantage of being reversible. That is, the transformation of a point cloud into any of these structures produces a simplified version of the data on which the feature detection algorithms can be applied. Afterwards, the simplified structures (voxels and/or lines) are labelled, and the transformation is reversed in order to recover the original points. These points inherit the labels from the simplified structures. Three algorithms for automatic feature detection are developed based on the two simplification and homogenization structures: (i) an algorithm for automatic detection of pole-like street furniture objects (e.g. traffic signs, traffic lights or lampposts), (ii) an algorithm for automatic surface detection (e.g. façades, walls or panels); and (iii) and algorithm for automatic delineation of road and street edges. For each one of the algorithms, the abstractions, implementation and validation tests are described and specified. In the validation tests, the algorithm for automatic detection of pole-like objects is able to identify more than 92% of the goals in the test point clouds, with actual poles accounting for 84% of the detected features. For the second algorithm, the results show that more than 90% of the points belonging to each one of the 27 test surfaces are assigned to them, and 99.9% of the points assigned to each surface are correct. The third algorithm is able to delineate more than 97% of the test road surface, and 98% of the surface delineated by the algorithm belongs to the actual road. The three algorithms overcome many of the drawbacks of previous algorithms, and they surpass their performance in terms of detection and success rates.
Conference Paper
Full-text available
En este trabajo se explora el potencial del escáner láser 3D y la comparación de nubes de puntos 3D para auscultar fenómenos de subsidencia por disolución kárstica en dolinas activas. El estudio se desarrolla en el Valle del Ebro, donde la composición del sustrato es altamente soluble, lo que propicia el desarrollo de hundimientos del terreno a una alta velocidad, causando la deformación en numerosos tipos de infraestructuras. Las mediciones comenzaron en octubre de 2014, con una periodicidad de unos 6 meses, sobre cuatro dolinas que afectan a diques de contención en el río Ebro, edificios, carreteras y líneas de ferrocarril; además de una dolina en condiciones naturales, sin ninguna infraestructura. El flujo de trabajo desarrollado para realizar el análisis espacial del proceso de subsidencia fue: 1) Adquisición de nubes de puntos mediante escáneres láser (Lecia C10 y Faro Focus X330), 2) Registro de las nubes de puntos utilizando métodos cloud to cloud y dianas de referencia, 3) Georreferenciación de la nube de puntos en sistemas de coordenadas comunes utilizando las dianas de referencia, 4) Limpieza y filtrado de la nube de puntos, y 5) Comparación de las nubes de puntos generadas en cada época para la acotación de las zonas de cambio. La comparación fue realizada estimando la distancia de separación entre nubes de puntos sucesivas, mediante la aplicación de varios métodos. La aplicación de esta metodología permitió evaluar las distintas técnicas, estimar las tasas de hundimiento de las dolinas activas, calcular el grado de afección a las infraestructuras y definir el patrón 3D de la deformación. PALABRAS CLAVE: riesgos naturales, auscultación, monitorización 3D, tasa de subsidencia, escáner láser terrestre 3D
Article
Full-text available
With the high density point clouds obtained with terrestrial laser scanning, detailed building facade structures, such as windows, can be recovered. There are usually only a few laser points available for windows, because window frames are small parts on walls, and window glass reflects no laser beam. Insufficient raw laser information makes it very difficult to recover reliable geometry of a window without human interaction. In this paper, we describe an approach to automatically extract windows from terrestrial point clouds. First, a segmentation process will group laser points in planar segments. Walls, doors and extrusions will be detected by applying feature constraints. Then, two detection strategies for windows are presented, depending on whether a window is covered with curtains or not. Windows which are not covered with curtains reflect no laser beam during the scanning process, and therefore cause holes on the wall segments. Laser points are available for windows which are covered with curtains, but they usually are not on the same plane as its wall and will therefore be grouped into other segments than wall segments. This again results in holes on wall segments. Holes on the wall are recognized by searching long edges from the TIN, generated for wall segments. After filtering out the holes caused by doors and extrusions, the remaining holes are believed to be caused by windows, and hence fitted to rectangles. The result from our approach is evaluated and discussed with examples of reconstructed building facades.
Conference Paper
Full-text available
In this paper, we address the problem of generating building footprints using terrestrial laser scanning from a Mobile Mapping System (MMS). The MMS constitutes a fast and adapted tool to extract precise data for 3D city modeling. Urban environments evolve over time due to human activities and other factors. Buildings are constructed or destroyed and the urban areas are extended. Therefore, the structures of the cities are constantly modified. Currently, building footprints can be generated using aerial data. However, aerial based footprints lack precision due to the nature of the data and to the associated extraction methods. The use of MMS is proposed as an alternative to perform this complex task. In this work, we propose an operational approach for automatic extraction of accurate building footprints. We describe the challenges associated with the terrestrial laser raw data acquired in realistic and dense urban environments. After a filtering stage on the 3D laser cloud point, we extract and reconstruct the dominant facade planes by combining the Hough transform, the k-means clustering algorithm and the RANSAC method. The building footprint is then estimated from these dominant planes. Preliminary experimental results are presented and discussed. The assessments show that this approach is very promising for the automation of building footprints extraction.
Article
Full-text available
Precise realistic models of outdoor environments such as cities and roads are useful for various applications. However, for a high level of detail, and a large size of environment to be digitized, one has to face the issues of quantity of information to be acquired, processed and stored, and of overall processing time. We present in this paper an integrated on-board laser range sensing system addressing this need: it is designed to perform city and road geometric modelling as it moves. It is based on a laser range sensor mounted on a vehicle whose position is known trough GPS-INS localization; it produces raw 3D range data and performs specific modelling for cities and features extraction for roads.
Article
Full-text available
The paper focuses on comparing accuracies obtained with photogrammetry and laser scanning in building extraction and partly com-pares results obtained with various methods. The objective of the EuroSDR Building Extraction comparison was to evaluate the quality, accuracy, feasibility and economical aspects of semi-automatic building extraction based on photogrammetric techniques with the emphasis on commercial and/or operative systems, semi-automatic and automatic building extraction techniques based on high density laser scanner data and semi-automatic and automatic building extraction techniques based on integration of laser scan-ner data and aerial images (hybrid technique). The project consists of three test sites by Finnish Geodetic Institute (FGI), namely Senaatti, Hermanni and Espoonlahti and one test site by Institut Geographique National (IGN), namely Amiens. For each test site fol-lowing data was provided to the partners: aerial images, camera calibration and image orientation information, ground control point coordinates and jpg images of point locations (not for Amiens), laser scanner data and cadastral map vectors of selected buildings. Participants were requested to create the vectors of 3D city models. 3D-models were obtained from 11 participants. Paper confirms with experiments that laser scanning is more suitable in deriving building heights, extracting planar roof faces and ridges of the roof whereas the photogrammetry and aerial images are more suitable in building outline and length determination. CyberCity, Stuttgart and TerraScan (performed by ICC) solutions provided the highest accuracy. There seemed to be a higher variation in the quality of other models depending on test site or remotely sensed information.
Article
Full-text available
The Finnish Geodetic Institute has been developing a mobile road environment mapping system, named as ROAMER, since 2003. The vehicle-borne ROAMER system consists of a carrying platform, positioning and navigation systems, and a 3D data acquisition system. The system employed a 3D laser scanner operated in a profiling mode. The laser scanner can be mounted in several different positions to meet the special needs of some planned applications of the mobile mapping system. In order to be able to accomplish the high automation in 3D modeling, a laser scanner capable to provide dense point clouds was set as the requirement for the system. Additionally, the system is aimed to be a flexible moving laboratory for various road environment applications. The wide field of view and a high point measurement frequency of 120 kHz provided by the laser scanner in use, makes the ROAMER unique. This paper describes the hardware and the navigation solution of the FGI ROAMER. We also discuss the applicability of mobile mapping system in the field of traffic engineering as data source. There is a wide range of laser sensors applicable, as the MMS presented in this paper, but also static laser profilers could be used for real-time traffic flow measurements. This data could be used as input to the MMS based 3D virtual models of the different traffic places to help traffic planners to increase the traffic safety. The detailed 3D models of the transport systems can be used for traffic modeling and traffic simulation systems. In 3D models the interactions between vehicles, pedestrians and bicycles can be examined in high detail. Also the interaction with traffic environment can be studied.
Conference Paper
Full-text available
Mobile mapping is a new way of efficiently collecting three-dimensional data from the road environment. Mobile mapping systems are cost efficient and robust technique to acquire information about even highly dynamic environments like highways and urban streets, where the data collection has previously been laborious and even dangerous for the staff performing the surveying. The dynamic mobile mapping systems could access the site with less risk to the personnel and with less need for road closures. The need for high resolution and details captured in to the data for street and road inventories, or city modelling, are the main reasons for the rapid adoption of the mobile mapping techniques in these fields. Lidar based mobile mapping system produces three-dimensional points from the surrounding objects. Typically, two-dimensional profiling scanner is mounted on the system and the third dimension is achieved by the movement of the vehicle. The characteristics of the obtained point cloud depend largely on the sensor arrangement and the sensor properties. The ROAMER, a single-scanner system for road environment mapping presented in this paper, is able to use various tilted scanning planes for the point acquisition with 120 kHz point measurement frequency and up to 48 Hz profile measurement rate. The relative point precision for the system is estimated to be a few millimetres, but is eventually defined absolutely by the accuracy of the navigation solution that could be provided in real-time, or more reliably through post-processing. We believe that in the future, lidar based mobile mapping will be used considerably for urban and road environment modelling, as well as in many other applications in the fields of construction, forestry, railways, and even in environmental modelling and monitoring e.g. hydrology and glaciology. In urban context, the main applications of these models could include urban and environmental planning, road safety assessment, road construction plan- ning and navigation.
Article
Full-text available
Over the past decade, laser terrestrial Mobile Mapping Systems (MMS) have been developed for the digital mapping of outdoor environments. While the applications of MMS are various (urban security, road control, virtual world, entertainment, etc.), one may imagine that for each application the system designs could be different. Hence, a comparative analysis of different designs may be useful to find the best solution adapted to each application. The objective of this paper is to propose a methodology to compare point-cloud data quality from different MMS designs by modifying spatial configuration of laser imaging system. For this methodology, we define several quality criteria such as precision, resolution, completeness. We illustrate this in the case of urban architecture digital mapping based on the use of a simulator.
Article
Full-text available
Accurate road environment information is needed in applications such as road maintenance and virtual 3D city modelling. Vehicle-based laser scanning (VLS) can produce dense point clouds from large areas efficiently from which the road and its environment can be modelled in detail. Pole-like objects such as traffic signs, lamp posts and tree trunks are an important part of road environments. An automatic method was developed for the extraction of pole-like objects from VLS data. The method was able to find 77.7% of the poles which were found by a manual investigation of the data. Correctness of the detection was 81.0%.
Article
Full-text available
The paper describes an approach for the quality dependent reconstruction of building facades using 3D point clouds from mobile terrestrial laser scanning and coarse building models. Due to changing viewing conditions such measurements frequently suffer from different point densities at the respective building facades. In order to support the automatic generation of facade structure in regions where no or only limited LiDAR measurements are available, a quality dependent processing is implemented. For this purpose, facades are reconstructed at areas of sufficient LiDAR point densities in a first processing step. Based on this reconstruction, rules are derived automatically, which together with the respective facade elements constitute a so-called facade grammar. This grammar holds all the information that is necessary to reconstruct facades in the style of the given building. Thus, it can be used as knowledge base in order to improve and complete facade reconstructions at areas of limited sensor data. Even for parts where no LiDAR measurements are available at all synthetic facade structures can be hypothesized providing detailed building geometry.
Article
This paper outlines a study, carried out on behalf of a national mapping agency, to validate laser scanned point cloud data collected by a ground-based mobile mapping system. As the need for detailed three-dimensional data about our environment continues to grow, ground-based mobile systems are likely to find an increasingly important niche in national mapping agency applications. For example, such systems potentially provide the most efficient data capture for numerical modelling and/or visualisation in support of decision making, filling a void between static terrestrial and mobile airborne laser scanning. This study sought to assess the precision and accuracy of data collected using the StreetMapper system across two test sites: a peri-urban residential housing estate with low density housing and wide streets, and a former industrial area consisting of narrow streets and tall warehouses. An estimate of system precision in both test sites was made using repeated data collection passes, indicating a measurement precision (95%) of between 0.029 m and 0.031 m had been achieved in elevation. Elevation measurement accuracy was assessed against check points collected using conventional surveying techniques at the same time as the laser scanning survey, finding RMS errors in elevation in the order of 0.03 m. Planimetric accuracy was also assessed, with results indicating an accuracy of approximately 0.10 m, although difficulties in reliably assessing planimetric accuracy were encountered. The results of this validation were compared against a theoretical error pre-analysis which was also used to show the relative components of error within the system. Finally, recommendations for future validation methodologies are outlined and possible applications of the system are briefly discussed.