Triangulated, connected sets for building detection from irregularly spaced LiDAR

Abstract

A novel method for building detection from irregularly spaced Light Detection and Ranging (LiDAR) data is presented. Using features from the triangulation of the LiDAR data, the proposed method identifies ground points, and differentiates those points from building and non building points. Furthermore, the method differentiates the buildings from one another, subsequently recognizing each building as a unique entity. The primary objective in the design of the proposed algorithm is to completely automate the building detection process such that no user intervention (window or parameter adjustment, selection of control points, apriori knowledge dependency, etc.) is necessary.