What is it about?
We design a new unsupervised approach to delineate building footprints on large-scale LiDAR point clouds. By computing an α-shape on low-height points, we delineate the building bottoms on the ground. We then measure the roughness of the entire points to find flat surface areas. Finally, valid building footprints are located by checking flat surfaces in the detected bottom areas. Compared to the Artificial Intelligence (AI)-assisted mapping results from Microsoft Building Footprints, the accuracy of the proposed method is 17% higher in the test areas. The simple and effective pipeline makes the proposed method easy to use and suitable for a wider range of applications.
Featured Image
Photo by Peter Hammer on Unsplash
Why is it important?
(1) Our approach is effective and easy to use without a time-consuming and data-hungry training process. (2) The heuristic parameter settings in our method are physically meaningful and easy to adjust for better performance. (3) The simplicity of our approach makes it suitable, for example, for interactive applications where the user wants to quickly visualize the result of building delineation on large-scale ALS point clouds.
Perspectives
This work is presented as part of the 11th SIGSPATIAL Cup competition, which considers the problem of delineating building footprints from point clouds generated by airborne laser scanning (ALS). It is challenging to solve a real and specific question by providing an executable program in a limited amount of time. At the same time, it was also meaningful, and I had a good time.
Xin Xu
University of Maryland at College Park
Read the Original
This page is a summary of: An unsupervised building footprints delineation approach for large-scale LiDAR point clouds, November 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3557915.3565986.
You can read the full text:
Contributors
The following have contributed to this page
