Update urban basemap by using the lidar mobile mapping system. The case of abu dhabi municipal system
- Autores: Omar Mubarak Abdulla Yaslam Alshaiba
- Directores de la Tesis: Nieves Lantada Zarzosa (dir. tes.), María Amparo Núñez Andrés (dir. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2020
- Idioma: español
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
Basemaps are the main resource used in urban planning and in building and infrastructure asset management. These maps are used by citizens and by private and public stakeholders. Therefore, accurate, up-to-date geoinformation of reference are needed to provide a good service.
In general, basemaps have been updated by aerial photogrammetry or field surveying, but these methods are not always possible and alternatives need to be sought. Current limitations and challenges that face traditional field surveys include areas with extreme weather, deserts or artic environments, and flight restrictions due to proximity with other countries if there is not an agreement. In such cases, alternatives for large-scale are required.
This thesis proposes the use of a mobile mapping system (MMS) to update urban basemaps. Most urban features can be extracted from point cloud using commercial software or open libraries. However, there are some exceptions: manhole covers, or hidden elements even with captures from defferent perspective, the most common building corners. Therefore, the main objective of this study was to establish a methodology for extracting manholes automatically and for completing hidden corners of buildings, so that urban basemaps can be updated. The algorithm developed to extract manholes is based on time, intensity and shape detection parameters, whereas additional information from satellite images is used to complete buildings.
Each municipality knows the materials and dimensions of its manholes. Taking advantage of this knowledge, the point cloud is filtered to classify points according to the set of intensity values associated with the manhole material. From the classified points, the minimum bounding rectangles (MBR) are obtained and finally the shape is adjusted and drawn.
We use satellite imagery to automatically digitize the layout of building footprints with automated software tools. Then, the visible corners of buildings from the LiDAR point cloud are imported and a fitting process is performed by comparing them with the corners of the building from the satellite image. Two methods are evaluated to establish which is the most suitable for adjustment in these conditions. In the first method, the differences in X and Y directions are measured in the corners, where LiDAR and satellite data are available, and is often computed as the average of the offsets. In the second method, a Helmert 2D transformation is applied.
MMS involves Global Navigation Satellite Systems (GNSS) and Inertial Measurement Units (IMU) to georeference point clouds. Their accuracy depends on the acquisition environment. In this study, the influence of the urban pattern is analysed in three zones with varied urban characteristics: different height buildings, open areas, and areas with a low and high level of urbanization.
To evaluate the efficiency of the proposed algorithms, three areas were chosen with varying urban patterns in Abu Dhabi. In these areas, 3D urban elements (light poles, street signs, etc) were automatically extracted using commercial software. The proposed algorithms were applied to the manholes and buildings.
The completeness and correctness ratio, and geometric accuracy were calculated for all urban elements in the three areas. The best success rates (>70%) were for light poles, street signs and road curbs, regardless of the height of the buildings. The worst rate was obtained for the same features in peri-urban areas, due to high vegetation. In contrast, the best results for trees were found in theses areas.
Our methodology demonstrates the great potential and efficiency of mobile LiDAR technology in updating basemaps; a process that is required to achieve standard accuracy in large scale maps.
The cost of the entire process and the time required for the proposed methodology was calculated and compared with the traditional method. It was found that mobile LiDAR could be a standard cost-efficient procedure for updating maps.
