Road geometry extraction using terrestrial based mobile

Report
Road geometry extraction using terrestrial based mobile
laser scanning system
Pankaj Kumar
Why Road Geometry?
 Road transportation has a central role to play in the progress and socio-economic growth of
society by providing efficient mobility.
 In the member states of the European Union, road traffic accidents claim around 43,000
lives and leave more than 1.8 million people injured annually, thus representing costs of 160
billion euro (ERSO, 2008).
 According to the World Health Organization, road traffic accidents will be the third leading
cause of injury and the sixth leading cause of death in the world by 2020.
 Information about road and its geometry is vital due to its significant correlation with road
safety and road maintenance (RIPCORD, 2007).
 We intend to use mobile mapping technology and active contours to extract the road from
LiDAR data facilitating the automated calculation of road geometry variables.
Figure 1: Road Accident Statistics in Europe
(ERSO, 2008)
Mobile Mapping Technology
 Mobile
(a)
(b)
Figure 2: (a) StratAG’s XP-1 System (b) Mapping and
Navigation Sensors
Mapping Systems (MMSs) enable the collection of high density geospatial data
using mapping and navigation sensors that are rigidly mounted together on a mobile
platform.
 Our MMS (XP-1) comprises of an IXSEA LandINS GPS/INS, Riegl VQ-250 laser scanner
and imaging camera system. The LiDAR system can capture up to 1million points every
3.5 seconds using a 300kHz sensor which leads to approximately 20 GB data per hour.
 Due to the large volume of data produced by MMSs, manual processing of the data is
very time consuming which prompts need for the development of automated methods. Our
research objective is to develop an algorithm for automated route geometry extraction
using terrestrial based laser scanning system.
Active Contour Model
 The
concept of active contours or snakes was first introduced by (Kass et al., 1988) and since then, it has been
widely used in many applications including edge detection, object boundary localisation, medical image analysis,
motion tracking and 3D reconstruction.
 A snake is an energy minimizing spline that moves towards the desired feature or object boundary under the
influence of internal forces within the curve itself and external forces derived typically from 2D imaging data by
minimizing certain energy such as edges or high frequency information.
1
E
*
snake
  ( Eint (v( s))  Eext (v( s)))ds
0
 In
our novel method, we initialise the snake contours over point cloud data based on the trajectory information
produced by the MMS navigation sub-system. The internal energy term provided to the snake contour is based on the
intrinsic properties of the curve, such as elasticity and bending, whilst the GVF (Gradient Vector Flow) energy and
constraint energy terms are derived from a number of the LiDAR point cloud attributes.
Figure 3: Active Contour
over Medical Image.
Extracting Road Geometry
(a)
(b)
(c)
(d)
Figure 3: (a) 3D LiDAR point cloud data (b) 2D road surface (c) Snake Contour without intensity data (d)
Snake Contour with intensity data.
GVF energy for the snake
contour has been extracted
from the localized surface
norm of LiDAR data while
constraint energy is based on
intensity data.
After finding the road edges,
the next step would be to
validate the results and to
extract the road geometry and
other road features.
References:
 ERSO (2008). Annual Statistics Report 2008: ERSO.Transport. Retrieved from http://www.erso.eu.
 Kass, M., Witkin. A. & Terzopoulos, D. (1988). Snakes: Active Contour Models. International Journal of Computer Vision, 1, 321-331.
 RIPCORD (2007). Safety Handbook for Secondary Roads. Retrieved from http://ec.europa.eu.
Research presented in this poster was funded by IRCSET, Pavement Management Services and a Strategic Research Cluster Grant (07/SRC/I1168) by Science Foundation Ireland under the National Development Plan.
The author gratefully acknowledges this support.

similar documents