3D City Model from LiDAR of Akoka, Lagos Using Geographical Information Technique

Abstract

3D city models are now widely adapted globally and since Lagos state is fast developing into a megacity, there is the need to have proper 3D city visuals. This study therefore created and visualized a 3D building model from LiDAR and orthophoto data of Akoka in Lagos State using geographical information techniques. A recently digitized Google Earth satellite image of the study area was compared with the Orthophoto to identify major changes in the area. The LiDAR point cloud data was classified using the automatic classification of LiDAR data method and then subsequently classified into buildings, vegetation, water body, grounds and roads. Elevation models such as Digital Terrain Model (DTM), Triangular Irregular Network (TIN) surface model and Digital Surface Model (DSM) were generated from the processed LiDAR data. Furthermore, maximum likelihood algorithm was used to classify the LiDAR image. And then accuracy assessment technique performed on the classified image. The result of the accuracy assessment indicated that the overall classification was 66.00% while the overall kappa statistics was 0.5677. This study concluded that the application of geographical information technique to LiDAR images can be used to assess, visualize and model 3D cities which can be very effective and useful in advancing the geospatial industry and enhancing urban planning for smart cities.