Journée CNRS GdR IASIS

M. Jouni presented the RGB-Hyperspectral fusion pipeline at the CNRS scientific meeting "Solutions matérielles pour la fusion de données".

The presentation slides are accessible via this link (OneDrive).

- Abstract -

Snapshot hyperspectral imaging is valuable in domains such as agriculture, security, surveillance, and waste sorting, but current cameras face trade-offs between spatial resolution, spectral resolution, and cost. A promising alternative is to pair a low-resolution hyperspectral sensor with a low-cost, high-resolution RGB camera and fuse their data. We present a dual-camera demonstrator for snapshot hyperspectral super resolution guided by a low-cost RGB camera. The system is implemented in ROS2, which orchestrates the full pipeline from data acquisition and preprocessing to registration and spectro-spatial fusion. Its design enables the evaluationof diverse fusion strategies in a reproducible and extensible framework. This work illustrates a pathway toward integrated hardware–software solutions for multimodal real-time fusion, contributing to the democratization of hyperspectral technology for practical applications.