Hyper-Drive
Infrared Hyperspectral Imaging Data Sets for Robots in Unstructured Environments
Affiliation:
Northeastern University: RIVeR Lab
Role:
Time Range:
Mechanical Design Engineer
2023
Skills:
Mechanical Design, CAD (Solidworks), Optical Design, Waterproof Design, Precision Location, Optical Calibration, Rapid Prototyping, Technical Communication
Background:
This academic project was a mobile robotic data collection platform that used hyperspectral imaging to make inferences and construct a hyperspectral outdoor navigation dataset. This required a number of mechanical systems to protect the expensive and sensative cameras, and allow for automatic and continuous calibration to adapt to changing lighting conditions.
Images:
My Contributions:
• Designed and built water-resistant cases for multiple different cameras with optically-clear windows
• Designed and built fixture for in-situ optical calibration, including a hyperspectral point sensor and spectralon reference surface with adjustable viewing angle.
• Assisted in development of academic deliverables
Outcomes:
This project waspublished in IEEE WHISPERS 2023 and the camera cases have been used for other hyperspectral research projects.
A follow up paper is expected to be published in RA-L this year.