Knowledge-Based Classification Method for Urban Area Objects Feature Extraction Based on LIDAR Points
Laser scanning technology can quickly capture a large area of high-precise 3D spatial data, and get the information of buildings, roads, vegetation and other urban objects from raw data. Based on this information general frame of these objects can be modelling. In this paper, an object-based classification method is proposed for urban objects based on LIDAR points: determine the contents of the objects contained in the scene; extract inherent features of different objects; establish objects feature knowledge database; combine and compare objects’ features and distribution of LIDAR points; derive a set of rule to express the point cloud classification which can be received by computer through fuzzy judgement. The method has been applied to LIDAR points by LYNX. The experiment results show that the proposed classification method is promising and usable.
H. G. Xu et al., "Knowledge-Based Classification Method for Urban Area Objects Feature Extraction Based on LIDAR Points", Applied Mechanics and Materials, Vols. 128-129, pp. 1157-1162, 2012