Extraction of Different Buildings from Lidar Data

Airborne laser scanning has become an arch technology and which is now widely accepted for 3D data collection. These 3D points represent the terrain surface as well as objects on top of the terrain surface. Airborne LIDAR systems are able to record data through multiple pulses which make the data very accurate. LIDAR data is a relatively new technology for obtaining Digital Surface Models (DSM) of the earth’s surface which can be very useful for detection of different physical features like trees, buildings, roads, etc. in rural and urban areas. In urban areas problem arises due to high density of points and therefore the demarcation between the different physical objects becomes difficult. In the paper LiDAR data of an urban area is used for buildings extraction. Firstly the data is smoothened to remove random error in it and then a new parameter openness is used for the extraction of building from it. Openness is a novel digital image processing technique which expresses degree of dominance or enclosure of a location on a surface. Openness helps in differentiating different types of physical features present in the data. All the results obtained are presented with discussion. The test results show considerable success in extracting buildings from an urban area where buildings and trees are of equal heights. Generation of the results in this project were done with the help of MATLAB programming.