Abstract

Thermal imaging has shown promise in a variety of clinical settings, including diagnosis and monitoring of diseases such as skin infection and vascular lesions, as well as non-contact monitoring of vital signs like breathing rate and body temperature.
While thermal cameras can provide a spatial temperature map, even high-resolution thermal cameras lack visible landmarks and thus are typically combined with visible cameras to supply the needed visual context. However, in existing devices the visible and thermal sensors are mounted side-by-side and not optically coaligned, producing parallax and mis-registration between thermal and visible images.
To overcome this limitation, we developed a low-cost, coaligned visible/near-infrared (VIS/NIR) and long-wave infrared (LWIR) imaging system in which a Raspberry Pi NoIR camera shares the optical axis with a Lepton 3.5 thermal camera via a 45° beamsplitter fabricated from Indium Tin Oxide (ITO)-coated glass.