True Leaf Area Index Retrieval Using Terrestrial LiDAR for Broadleaf Trees via Novel Multiinclined-Planes Method

True leaf area index (LAIt) is a more crucial and efficient structural parameter to characterize the photosynthetic capacity of broadleaf forests than the concept of effective leaf area index, which is the predominant form retrieved by remote sensing. Conventional LAIt retrieval methods rely on the accurate terrestrial laser scanning (TLS), regarded as an essential means for capturing the size and distribution of leaves within the tree crown. However, results of current TLS-based LAIt retrieval algorithms are highly sensitive to physical parameters such as voxel configuration and gap size. In this study, we proposed a point cloud multi-inclined-planes (MIP) method to address the above issue. The MIP begins by extracting leaf point clouds from individual tree point clouds and clustering these leaf points. Independent leaf clusters are identified through clustering, and each cluster is treated as a complex surface composed of multiple inclined planes. The area of this complex surface is then calculated to derive the LAIt. Through validating on both simulated and measured data with comparison to other methods, MIP achieved a root mean square error of 1.197 m²/m² and demonstrated robustness. By eliminating the factors that may affect retrieval of LAIt while ensuring insensitivity to parameter variations, it holds great promise in forest monitoring and application.