A high-resolution electrostatically actuated MEMS Fabry–Pérot tunable filter for LWIR imaging and sensing applications

denotes threefold rotational symmetry with a 120° periodicity), enables high spectral finesse and stable large-stroke actuation. The device achieves a minimum full width at half maximum (FWHM) of 188 nm, a maximum quality factor (Q) of 63.7, and continuous tuning from 10.3 to 12 μm, representing advanced performance among LWIR MEMS-FPTFs. To further demonstrate application potential, we developed a dual-channel spectral-visual proof-of-concept recognition demonstration based on the MEMS-FPTF. Compared with a vision-only neural-network baseline, the hybrid model exhibits a significant improvement in recognition performance under illumination variations and complex background conditions. The proposed MEMS-FPTF offers a compact, low-power, and high-precision solution for LWIR spectral filtering, with strong promise for mineral identification, autonomous perception, and remote spectral sensing.