Hyperspectral Imaging with Active Illumination using Incandescent Lamp and Variable Filament Temperature

Presented novel hyperspectral imaging technique use variable filament temperature incandescent lamp and multi-channel image acquisition. This novel hyperspectral imaging approach is characterized with the simulation.

We simulated an illumination of ColorChecker chart with a black body spectra at varying temperatures and multi-channel image capture in spectral range 400 to 700 nm. We present an algorithm for spectrum reconstruction, addressing the ill-posedness of inverse problem. We assessed the impact of various acquisition parameters on the accuracy of reconstructed spectra, including noise levels, temperature steps, filament temperature range, illumination spectral uncertainties, spectral step sizes in reconstructed spectra, and the number of detected spectral channels.

We successfully reconstructed all spectra, with Root Mean Squared Errors below 5%, reaching as low as 0.1% for specific cases such as Black color. Flat spectra exhibit higher accuracy than complex ones. The illumination spectrum accuracy emerges as a critical factor influencing spectrum reconstruction quality.

Ultimately, our study establishes the feasibility of this innovative hyperspectral imaging approach and identifies optimal acquisition parameters, setting the stage for future practical implementations.