AUTOMATED BUILDING DETECTION USING RANSAC FROM CLASSIFIED LIDAR POINT CLOUD DATA

For the past 10 years, the Philippines has seen and experienced the growing force of different natural disasters and because of this the Philippine governement started an initiative to use LiDAR technology in the forefront of disaster management to mitigate the effects of these natural phenomenons. The study aims to help the initiative by determining the shape, number and distribution and location of buildings within a given vicinity. The study implements a Python script to automate the detection of the different buildings within a given area using a RANSAC Algorithm to process the Classified LiDAR Dataset. Pre-processing is done by clipping the LiDAR data into a sample area. The program starts by using the a Python module to read. LAS files then implements the RANSAC algorithm to detect roof planes from a given set of parameters. The detected planes are intersected and combined by the program to define the roof of a building. Points lying on the detected building are removed from the initial list and the program runs again. A sample area in Pulilan, Bulacan was used. A total of 8 out of 9 buildings in the test area were detected by the program and the difference in area between the generated shapefile and the digitized shapefile were compared.