Super-field-of-view non-line-of-sight imaging via spatial encoding of a translated point spread function

Non-line-of-sight (NLOS) imaging aims to reconstruct objects beyond line-of-sight view, offering potential applications in various fields. However, conventional transient NLOS methods are constrained by the detection region, preventing the reconstruction of targets outside its normal space and thereby limiting practical applicability. In this paper, a computational imaging method for super-field-of-view (Super-FoV) reconstruction based on spatial encoding of a translated point spread function (PSF) is proposed. By formulating a forward model with PSF translation and incorporating spatial-domain encoding to eliminate circular convolution artifacts, the proposed method achieves efficient and high-fidelity reconstruction of targets beyond the detection region. Numerical simulations demonstrate that our method achieves multi-target Super-FoV reconstruction covering an area equivalent to 9× the detection region. Experimental results further validate that the proposed method, with low computational and memory costs, reconstructs targets located beyond 2× the detection range along a single direction. This is equivalent to an imaging capability covering an area approximately 19.6× that of the original detection region. Moreover, by performing multiple translated reconstructions and stitching based on a single scan, the approach achieves large-FoV imaging, accurately recovering the spatial locations and morphological features of multiple targets. This study overcomes the inherent FoV limitations of conventional transient NLOS imaging and introduces a novel computational paradigm for flexible and efficient Super-FoV reconstruction.