Abstract
Long-range imaging in the Mid-wave infrared (MWIR) is crucial for defense, industrial and environmental applications, requiring very large aperture (~100-1000 mm) lenses for high-resolution imaging. Glass-based refractive optics make such systems bulky and expensive. Metalenses are lightweight alternatives but face fabrication limits at large apertures. We proposed a computational imaging strategy, the Golay metalens, combining (a) a small aperture array in designed spatial configuration and (b) a reconstruction algorithm that recover high-resolution, high-contrast images as though obtained by a single large aperture lens. The design is inherently scalable to larger apertures. A prototype in 89 mm diameter, 356 mm focal length was demonstrated, achieving near diffraction-limited performance enhanced by an image reconstruction algorithm with neural network-based denoising. A large recursive Golay metalens was further designed, demonstrating improved spatial resolution in simulations. Our development represents a significant step toward practical, high-performance MWIR imaging systems that are both scalable and accessible.