Parametric study of blind-label acoustic subwavelength imaging

There is a long-existing tradeoff between the imaging resolution and the penetration depth in imaging systems caused by the diffraction limit. We developed a “blind label” approach to tackle this problem, which significantly improves the practicality of acoustic subwavelength imaging in biomedical ultrasound imaging, non-destructive testing, and other acoustic sensing and communication applications. The “blind labels” in our system refer to randomly distributed acoustic scatterers with deep-subwavelength sizes whose exact locations and trajectories are not necessary information in image reconstruction. Our imaging framework is composed of two parts: (1) spatial mixing: a physical process that converts the originally evanescent components in the scattered waves from the object to propagating components that can reach the far-field detector and (2) computational reconstruction. In this talk, we will mainly report our quantitative investigation of the system parameters’ impact on the performance of the blind-label subwavelength imaging system, providing guidance to future system setups in various applications.