Deep learning and Canopy Lidar Point Clouds

Estimating dendrometric parameters by processing lidar data into deep neural networks:

This project aims to estimate biophysical parameters of trees, which are - tree height, tree crown diameter and diameter at breast height (DBH), by processing the UAV borne tree lidar point clouds into the the pointnet deep neurals. The purpose of choosing pointnet was due to its special characteristic to process point clouds without any prior voxelisation or 3D blocks of the points. The output will further be modeled using the "Whitbox Tools" to determine biophysical parameters.

The purpose of this project is to create an automated and easy workflow for calculating biophysical trees. Because these parameters play a huge role for maintaining the forest grounds, tracking its changes and even preparing for any unprecedented environmental hazards. The future work involving this project will be to create allometric models that can direct these parameters to estimate the forest biomass and carbon stock.

Please refer my ArcGIS storymap theorizing this project.