Today, tree distribution maps can only be generated manually in a very time-consuming process, and real-time microclimate mapping of a large 3D volume under tree canopy is not possible. A prototype small quadrotor unmanned aerial vehicle (UAV) system was developed that is able to maneuver in cluttered environments like forests, and under tree canopy to map tree distributions from 3D point clouds gathered from an onboard stereo vision system. The UAV uses a small onboard sensor board to record micro-climate parameters during flight.
Based on an existing micro air vehicle system that executes an autonomous navigation framework to maneuver in GPS-denied environments, close to the ground and in highly cluttered environments, a data processing framework was developed to analyze imagery from an onboard, forward-looking stereovision system to calculate tree trunk distribution maps. This includes estimating the forest floor profile, detecting tree trunks in the images, and calculating the diameter of tree trunks at a certain height above the ground. While data processing is intended to run onboard the vehicle, it is currently performed offline from collected images, which are manually labeled to compare tree diameter and position estimations with ground truth. The framework was tested on data from an example forest of live oak.
Applications for
this work include rapid assessment of biomass for commercial forestry and carbon offsets, and science data collection such as structure and microclimate that are currently not feasible at this scale. Additionally, mapping below-canopy microclimate within forest canopies is an essential aspect of carbon cycle model validation, because it has the potential to falsify parameterizations that are currently poorly quantified. Microclimate sensing of radiative fluxes also has potential for improved remote sensing retrievals under complex canopies. NASA’s terrestrial modeling efforts in carbon and water cycles can be significantly improved by the proposed spatial data collection of microclimate. This approach also expands carbon inventorying and microclimate mapping into environments that are traditionally very difficult to map (e.g. mangrove mapping).
This work was done by Roland Brockers, Stephan M. Weiss, Kyle Ashley, and Lorenzo Alligo of Caltech; and Adam Wolf, Adam Yabroudi, Mike Lee, and Kelly Caylor of Princeton University for NASA’s Jet Propulsion Laboratory. This software is available for commercial licensing. Please contact Dan Broderick at
This Brief includes a Technical Support Package (TSP).
Rapid Forest Triage by Sub-Canopy Micro Air Vehicles
(reference NPO-49673) is currently available for download from the TSP library.
Don't have an account?
Overview
The document outlines the "Rapid Forest Triage by Sub-Canopy Micro Air Vehicles" project conducted by NASA's Jet Propulsion Laboratory (JPL). The primary objective of this initiative is to utilize small quadrotor unmanned aerial vehicles (UAVs) that can navigate autonomously beneath tree canopies to perform in-situ sensing for forest monitoring applications. This innovative approach aims to revolutionize forest inventorying and ecosystem monitoring, which traditionally requires extensive manual labor and time.
Key achievements from the fiscal year 2014 include the first successful demonstration of stereo-vision-based forest inventorying using sub-canopy UAV flights, with a human-in-the-loop component for validation. The project developed and integrated a lightweight sensor payload capable of monitoring microclimate parameters during flight, which is crucial for understanding forest ecosystems. The UAVs are equipped with advanced imaging and processing capabilities, allowing for tree trunk detection, diameter estimation, and position estimation through stereo point clouds.
The significance of these results is profound. The technology enables real-time, fully autonomous forest inventorying, drastically reducing the time required to map a one-hectare forest plot from weeks to potentially minutes or hours. This efficiency is particularly beneficial for applications such as biomass assessment in commercial forestry, carbon offset initiatives (e.g., REDD+), and scientific data collection related to forest structure and microclimate, which are currently challenging to achieve at such scales.
Moreover, the project addresses the need for mapping below-canopy microclimates, which is essential for validating carbon cycle models. The ability to accurately assess microclimate conditions in forest environments that are typically difficult to access, such as mangroves, expands the scope of ecological research and monitoring.
The document also emphasizes the collaborative nature of the project, involving researchers from Princeton University and highlighting the importance of integrating human oversight in the initial phases of data collection and analysis. Overall, the "Rapid Forest Triage" project represents a significant advancement in the use of UAV technology for environmental monitoring, promising to enhance our understanding of forest ecosystems and contribute to more effective management practices.
