Lidar Filtering Algorithms for Urban Flood Application: Review on Current Algorithms and Filters Test

Digital terrain model (DTM) is one of the important input parameters in urban flood application. This is because it influences the flow direction, flow velocity, flood extend and flood depth. LiDAR offers accurate DTM for large areas within a short period of time. From the overall LiDAR data processes, filtering (classification) poses the greatest challenge. Evaluation and comparison of current filtering algorithm is(are) done to find out which one can best filter the LiDAR data in order to develop an accurate DTM that suits the urban flood application. We have tested eight different algorithms. The results have been analysed in a qualitative assessment (i.e.: visually assessing the performance of the algorithm in several terrain types) and then followed by a quantitative analysis (i.e.: height comparison) using the RMSE formula. The result is then used in flood simulation by using MikeFlood software to see the outcome of filtering process to DTM and finally to the flood model. Accurate results in urban flood application depend on how close DTM can represent the urban surface. Objects like buildings and bridge should be removed while objects like ramps which give impact to the flood flow should be maintained. From the overall results and assessment, the advantages and disadvantages of each filter are analysed to formalize a new assumption for the new filtering algorithm that is suitable for the urban flood application. This paper also explains the next tasks, which are to focus on improving the filtering algorithm to detect bridges using geometric method and implementing procedures to remove the bridges during DTM generation. * A.F. Abdullah a