Super-resolution imaging of limited-size objects

Abstract Label-free, far-field super-resolution imaging can be achieved by exploiting prior knowledge about an object, such as sparsity, or by using information accumulated from similar object classes. Here we show that simply knowing that an object is confined within a limited spatial extent is sufficient to surpass the Abbe–Rayleigh diffraction limit: for a fixed photon budget, smaller objects can be resolved with higher spatial resolution. To demonstrate this, we develop a limited-size object microscopy (LSOM) technique. The method relies on representing the coherently scattered field from the object within a limited field of view with Slepian–Pollak functions, a family of prolate spheroidal wavefunctions. The coefficients of such functions can then be recovered from diffraction-limited measurements. We experimentally demonstrate down to λ /8 resolution (where λ is the wavelength) for subwavelength structures and analyse the performance limits of the method using information theory. The technique requires no assumptions about the object’s shape or complexity and does not rely on labels, making it broadly applicable to the study of isolated nano-objects.