Building boundaries extaction from point clouds using dual-threshold Alpha Shapes

Abstract

Modern urban buildings have complex model structure, and the building boundary extraction from airborne LiDAR data is still a relatively difficult problem. This paper proposes a dual-threshold Alpha Shapes algorithm and contour simplification algorithm based on least squares to extract the initial building contour. Dual-threshold Alpha Shapes algorithm consists of four steps: (1) dual-threshold a-shape acquisition, (2) based on a simple ring breadth-first search query algorithm, (3) the corresponding path screening, (4) deriving a closed contour (4) a closed contour, (5) Least Squares contour simplification. Experiments are conducted to demonstrate the efficiency of the proposed method on two datasets.