NetBurner In The Field: UAV-Hyperspectral Imager for Ecological Monitoring

UAV Sorasak

At NetBurner, we love to see what people are building with our products. What’s even better is when we get permission to showcase those inspiring innovations, products and systems. If you get fired up about airborne drones, embedded computing, remote sensing, environmental research, and geo-rectified aerial 3D mapping all rolled into one, then this article, and the referenced paper, are well worth the read.

The National Research Council of Canada (NRC) has partnered with the Applied Remote Sensing Lab at McGill University in Montreal to develop a UAV-based hyperspectral imaging system to allow for remote ecological monitoring. As if we didn’t think this project was awesome enough already, this system was made even better after we found that it incorporates the NetBurner MOD54415 System on Module. We thought it would be worth sharing how one of our customers is using our products in the field.

NetBurner McGill UAV Hyperspectral Imaging
Figure 1: The system comprises of a DJI M600Pro Hexacopter, the micro compact airborne spectrographic imager (μCASI), the IMU/GPS unit (which incorporates a MOD54415), among other subsystems. The μCASI and IMU/GPS are mounted on the hexacopter’s gimbal to maximize system stability and minimize image distortion.

Hyperspectral imaging is the process of capturing and processing information from across a continuous broad swath of frequencies in the electromagnetic (EM) spectrum – the sensors used are capable of detecting EM radiation at frequencies beyond that of visible light, and into the infrared (IR) and ultraviolet (UV) ranges. The μCASI (short for micro Compact Airborne Spectrographic Imager) used in this research is capable of detecting 288 discrete spectral bands from 401-996 nanometers. That spans from the deep blue edge between the visible light band and the near ultraviolet band, through visible light and into the near-IR (NIR) band, also known as shortwave infrared (SWIR), regions of the electromagnetic spectrum.

UAV-Hyperspectral remote imaging is becoming more common in vegetation studies and can be used to provide detailed information on plant chemical and structural characteristics. This information is useful for determining plant species richness, studying invasive species, plant health, and can even be used to identify plant species.

geocorrected hyperspectral images
Figure 2. Ultra-high spatial resolution geocorrected hyperspectral images of the three study sites from the UAV-μCASI system.

The objective of this specific research study is to complement airborne hyperspectral research for environmental applications as well as to advance the understanding of UAV-hyperspectral systems as an alternative to satellite and in situ measurements (field spectroscopy).

UAV mounted hyperspectral imaging systems are in their early stages. These systems are still expensive and face many challenges, including battery performance (or flight duration) and the geocorrection of the imagery. With the growing popularity of turn-key UAV-hyperspectral systems on the market, this research demonstrates the basic requirements and technical challenges for these systems to be fully operational.

System diagram of the NetBurner powered UAV with Hyperspectral imaging
Figure 3: Schematic of the IMU/GPS unit (also referenced as the IGDR, or IMU-GNSS Data Recorder, in the research paper). The “SBC”, or Single Board Computer, is a NetBurner MOD54415.

The system developed uses a NetBurner MOD54415 (Single Board Computer or “SBC” as labeled in the diagram) to communicate with the GPS and IMU. Several metrics from the ADIS16490 IMU and Novatel OEM719 GPS (like pitch, roll, yaw, position, and velocity) are sampled, timestamped, and stored onto the microSD card of the MOD54415 in real-time. 

The MOD54415 is also tasked with sending GPS information via Ethernet to the μCASI, which is then used in real-time to determine its location relative to a user-defined geofence. After a flight is complete, the IMU and GPS data are retrieved from the microSD card via the Ethernet port for further analysis and compared against IMU/GPS data retrieved from the DJI M600Pro.

There are a lot more cool tidbits! Using the data produced by the system with further video imaging, 3D surfaces could be reconstructed for additional visualizations.  Using SfM-MVS workflow (Structure from Motion and Multi-View Stereo, respectively) the paper describes how 3D point clouds and a digital surface model (DSM) of the study areas were rendered as 3D geo-rectified orthomosaic maps. The figure below shows one of the locations from their ecological studies captured using a DJI Inspire 2 UAV with an X5S camera system, as described in the full research paper. It’s well worth the read.

3D point cloud from a flight path
Figure 4. A subset of the 3D point cloud from a flight path through Cowichan Garry Oak Preserve. The interactive version of the point cloud can be viewed here: http://bit.ly/CGOP-MD.”

Here at NetBurner, we love to see the amazing things our customers and community of innovators are up to. We hope you enjoyed this showcase. Please comment below and contact us if you want us to feature your project or if you need help getting it off the ground!

Original research article and figure attributions: https://www.mdpi.com/2504-446X/3/1/12

Under the CC BY4 license, the material can be freely used (no journal copyright) (https://creativecommons.org/licenses/by/4.0/deed.ast ).

Cover photo credit: instagram.com/boontohhgraphy

Share this post

Subscribe to our Newsletter

Get monthly updates from our Learn Blog with the latest in IoT and Embedded technology news, trends, tutorial and best practices. Or just opt in for product change notifications.

Leave a Reply
Click to access the login or register cheese