Using Octrees to Detect Changes to Buildings and Trees in the Urban Environment from Airborne LiDAR Data

Change detection is a major issue for urban area monitoring. In this paper, a new three-step point-based method for detecting changes to buildings and trees using airborne light detection and ranging (LiDAR) data is proposed. First, the airborne LiDAR data from two dates are accurately registered using the iterative closest point algorithm, and a progressive triangulated irregular network densification filtering algorithm is used to separate ground points from non-ground points. Second, an octree is generated from the non-ground points to store and index the irregularly-distributed LiDAR points. Finally, by comparing the LiDAR points from two dates and using the AutoClust algorithm, those areas of buildings and trees in the urban environment that have changed are determined effectively and efficiently. The key contributions of this approach are the development of a point-based method to effectively solve the problem of objects at different scales, and the establishment of rules to detect changes in buildings and trees to urban areas, enabling the use of the point-based method over large areas. To evaluate the proposed method, a series of experiments using aerial images are conducted. The results demonstrate that satisfactory performance can be obtained using the proposed approach.