This project explores the application of persistent homology to classify Airborne Laser Scanning (ALS) data of forest landscapes into landscape-scale units, enhancing the effectiveness of Terrestrial Laser Scanning (TLS) data collection and facilitating the scaling of TLS to ALS data. By employing topological data analysis, specifically persistent homology, we analyze the complex structure of ALS-derived point clouds to identify distinctive patterns and structures within the forest. These patterns are categorized into defined landscape units, which represent various forest types and structures. The classification achieved through this method allows for targeted TLS data collection, focusing efforts on representative sample areas that encapsulate the diversity of landscape features identified by ALS. Furthermore, the project develops methodologies for aligning and scaling TLS data with the broader, more comprehensive ALS datasets. This alignment is crucial for improving the accuracy and utility of forest models at multiple scales. Through this approach we aim to bridge the gap between high-resolution, localized TLS data and extensive, landscape-level ALS data, providing a framework for enhancing forest management and research initiatives.
Topological Data Analysis to Identify Landscape Scale Units in Aerial LiDAR Data