THE TEAM

The DASL Drone-Based IR Leak Detection Team | 2024-2025

ABOUT

At the Sanghi College of Engineering on the NAU campus, the Drone-Based IR Leak Detection project is conducted under the Dynamic and Active Systems Lab, led by Dr. Michael Shafer. The project addresses pipe network degradation in the Greater Phoenix Area, where leaks often are visible at the surface from older pipes. Using thermal imagery taken from a UAV, the team processes these images to identify and visualize potential leaks, providing these insights to the Salt River Project (SRP).

DRONE USAGE

To perform surveys of these pipe networks, the team needed a drone capable of taking thermal imagery pictures which will provide the bulk of data needed for analysis. The drone the team owns and operates for the image collection is the Mavic 3T Drone from DJI, which produces reliable outputs for many of the flights the team performs. Flying at a range of 150 to 200 ft in the air, the drone takes multiple images at a slow pace until the end of the flight path is reached. Then, the images taken are extracted from the drone and then moved into an external processing software.

INSTALLATION

Follow the link to our GitHub or scanning the QR code to view the README file for step-by-step instructions on how to install all necessary processing programs and setup the parameters for the drone flights.

PREPROCESSING

Once the flights have been completed and the images have been extracted from the drone, they need to be exported into an external preprocessing program. For DJI drones, the collected images are processed using the Fiji program which converts the proprietary JPG images to non-proprietary TIFF images. These TIFFs can then be used for the later processing stages. To download the Fiji program, follow the GitHub link and go through the installation instructions for the IR imagery here. Then, follow the instructions in this document for DJI specific parameters.

PROCESSING

Once the TIFFs have been generated from Fiji, they are then processed in Agisoft Metashape to produce orthomosaic products. These products consist of the temperature profile, the greyscale imagery, and the digital elevation model (DEM). In the post-processing stage, all three will be utilized in the analysis of each site. In the image provided, this is an example of what should be provided as the result of the processing.