High-resolution Imaging Through a Multimode Fiber: From Raster-scanning to Compressive Sensing

Optical multimode fibers (MMFs), known for their compactness, flexibility, and high mode density within a small footprint, are ideal tools for imaging across various applications from neuroscience to semiconductor metrology. The continuously increasing computing power makes computational imaging through MMFs particularly promising. In this thesis, we investigated two computational imaging methods, wavefront shaping (WFS) based raster-scan (RS) imaging and compressive imaging (CI). We studied the performance and the limitations of both approaches, and developed a new compressive imaging technique for high-speed super-resolution 3D imaging through a flexible ultimately-thin probe.