Binocular low-distortion metalens for depth sensing

Metasurfaces provide a promising solution for compact and lightweight binocular depth sensing. However, most existing methods either neglect the impact of imaging quality and geometric distortion on accuracy, or rely on complex post-processing for correction and refinement. Here, we propose a low-distortion binocular metalens system for depth sensing without a backend correction algorithm. A pair of polarization-independent silicon-based metalenses, designed with a 30° field of view and <0.20% maximum distortion, are monolithically integrated on a single substrate alongside an imaging sensor for left and right view acquisition. The binocular system is then experimentally characterized for 3D object imaging under incoherent near-infrared light. This work provides a feasible route toward miniaturized and low-complexity three-dimensional depth-sensing systems, with potential applications in compact security monitoring and intelligent vision terminals.