Refocusing performance of multi-spectral light field camera through spectrally and depth-resolved flame imaging

Flames are highly dynamic, and their radiation, temperature, and radical emissions vary with combustion conditions. Different spectral bands reveal distinct radical emissions. Flames are also inherently three-dimensional (3D), and imaging at multiple depths is needed to resolve spatial variations. Multi-spectral imaging enables the capture of 3D flame characteristics through refocused imaging across spectra and depths, simultaneously capturing spatial, angular, and spectral data. This study presents a systematic investigation of how various filter array designs and multi-spectral light field camera (MS-LFC) configurations influence the quality of refocused flame images. Simulations were conducted to assess the effects of filter array designs and MS-LFC configurations on the sharpness and structural similarity of refocused images. The results show that centrally distributed spectral filters and a sufficiently dense set of sub-aperture images significantly enhance the sharpness and structural similarity of refocused flame images, enabling accurate depth-resolved flame reconstruction across different spectra. Conversely, reduced angular sampling leads to sparser data acquisition, degrading image similarity and overall quality, except at the focal plane, where refocusing is less affected. A detailed analysis of these findings offers valuable guidance for the design and optimization of MS-LFC systems, contributing to improved accuracy and efficiency in spectrally and depth-resolved flame imaging for combustion diagnostics.