Modelling of LiDAR return on surface object reflected as specular and diffuse components / Noraain Mohamed Saraf

The useful of LiDAR data in the industry necessitate user the ability of managing LiDAR data according to their purpose, for example in 3-D mapping. Dealing with 3- D mapping, features information of LiDAR data (e.g building and tree) have been tremendously useful. This research intent to investigate the characteristics of LiDAR return pulses coming of specular (building) and diffuse (tree) objects from LiDAR derived surface models of the urban study site in Ampang, Kuala Lumpur. Contour, slope and aspect are three of those products that could be derived from LiDAR data. The data were initially checked and verified accordingly. Digital models (DEM and DSM) were generated based on this LiDAR data involving classification, filtering and masking. A normalized DSM was extracted to separate the buildings from other spatial features. Slope and aspect analysis were conducted based upon segmentation on the rooftop. Furthermore, height of building and tree were estimated. Hence, an error assessment was done and findings were highlighted and documented. The result of LiDAR verification certifies that the data is reliable and useable where the Root Mean Square (RMS) error obtained is within the tolerance value of vertical accuracy (z), which is 0.096m. Thus, specular and object extraction were conducted with error assessment less than 10 percent. The segmentation applied based on contour, slope and aspect analyses indicate that the approach can derive the reliable and accurate 3-D building. The finding from this study demonstrates the capability and the effectiveness of LiDAR data.