Vegetation Stratum Occupancy Prediction from Airborne LiDAR 3D Point Clouds

TL;DR

This paper introduces a deep learning method for estimating vegetation layer occupancy from airborne LiDAR data, using aggregated supervision to produce accurate, interpretable 3D vegetation maps.

Contribution

It presents a novel deep learning approach that predicts vegetation occupancy from LiDAR point clouds with supervision from aggregated plot data, improving interpretability and accuracy.

Findings

Outperforms baseline methods in precision

Provides visually interpretable occupancy maps

Uses aggregated supervision for training

Abstract

We propose a new deep learning-based method for estimating the occupancy of vegetation strata from 3D point clouds captured from an aerial platform. Our model predicts rasterized occupancy maps for three vegetation strata: lower, medium, and higher strata. Our training scheme allows our network to only being supervized with values aggregated over cylindrical plots, which are easier to produce than pixel-wise or point-wise annotations. Our method outperforms handcrafted and deep learning baselines in terms of precision while simultaneously providing visual and interpretable predictions. We provide an open-source implementation of our method along along a dataset of 199 agricultural plots to train and evaluate occupancy regression algorithms.