Computational diffraction-limited imaging with a non-imaging spatial information transfer lens

The demand for low-cost, high-performance miniaturized optical imaging systems requires creating a new imaging paradigm. In this paper, we propose an imaging paradigm that achieves diffraction-limited imaging with a non-imaging spatial information transfer lens. The spatial information transfer lens realizes a perfect match between the space–bandwidth product (SBP) of the lens and that of the image sensor so that the collected spatial information from the object can be totally recorded and fully resolved by the image sensor. A backward wave propagation model is developed to reconstruct the object by propagating the light wave modulated by the information transfer lens back from the image space to object space. The proposed imaging paradigm breaks the point-to-point imaging structure and removes the focusing-distance constraint, allowing a flexible arrangement of the object and the image sensor along the optical axis with a compact form factor of the optical system. We experimentally demonstrate the versatility and effectiveness of the proposed imaging paradigm. The proposed imaging paradigm is low-cost, simple in configuration, flexible in arrangement, and diffraction limited with great potential applications in biomedical imaging.