ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-5/W1, 35-41, 2017
https://doi.org/10.5194/isprs-annals-IV-5-W1-35-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.
 
13 Dec 2017
PEDESTRIAN PATHFINDING IN URBAN ENVIRONMENTS: PRELIMINARY RESULTS
G. López-Pazos1, J. Balado2, L. Díaz-Vilariño2,3, P. Arias2, and M. Scaioni4 1School of Industrial Engineering, Campus Lagoas-Marcosende, Vigo, 36310 Spain
2Applied Geotechnologies Group, Dept. Natural Resources and Environmental Engineering, University of Vigo, Campus Lagoas-Marcosende, Vigo, 36310 Spain
3Faculty of Engineering of University of Porto, Research Centre for Territory, Transport and Environment (CITTA), Rua Dr. Roberto Frias, s/N, Porto, Portugal
4Dept. of Architecture, Built Environment and Construction Engineering, Politecnico di Milano, via Ponzio 31, 20133 Italy
Keywords: Accessibility, As-built 3D, Dijkstra Pathfinding Algorithm, Graphs, Obstacle detection, Point Cloud Classification, Pedestrian, Smart Cities, Topology Abstract. With the rise of urban population, many initiatives are focused upon the smart city concept, in which mobility of citizens arises as one of the main components. Updated and detailed spatial information of outdoor environments is needed to accurate path planning for pedestrians, especially for people with reduced mobility, in which physical barriers should be considered. This work presents a methodology to use point clouds to direct path planning. The starting point is a classified point cloud in which ground elements have been previously classified as roads, sidewalks, crosswalks, curbs and stairs. The remaining points compose the obstacle class. The methodology starts by individualizing ground elements and simplifying them into representative points, which are used as nodes in the graph creation. The region of influence of obstacles is used to refine the graph. Edges of the graph are weighted according to distance between nodes and according to their accessibility for wheelchairs. As a result, we obtain a very accurate graph representing the as-built environment. The methodology has been tested in a couple of real case studies and Dijkstra algorithm was used to pathfinding. The resulting paths represent the optimal according to motor skills and safety.
Conference paper (PDF, 950 KB)


Citation: López-Pazos, G., Balado, J., Díaz-Vilariño, L., Arias, P., and Scaioni, M.: PEDESTRIAN PATHFINDING IN URBAN ENVIRONMENTS: PRELIMINARY RESULTS, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-5/W1, 35-41, https://doi.org/10.5194/isprs-annals-IV-5-W1-35-2017, 2017.

BibTeX EndNote Reference Manager XML