Abstract: This tutorial will introduce computational imaging as a broad range of techniques in which algorithms play a major role in the final image formation process. The basic recipe for computational imaging involves coming up with a forward model that can simulate/predict what your imaging system measures. This model is computationally inverted to reconstruct the object under investigation. We will also discuss how to incorporate prior knowledge and constructs from machine learning into the computational reconstruction algorithms to improve the image reconstruction fidelity. The tutorial will include examples from computational optical imaging, such as diffusercam and diffraction tomography.
Bio: Kevin C. Zhou is a postdoctoral researcher at UC Berkeley, working with Profs. Laura Waller and Hillel Adesnik. His research is centered around developing high-throughput, data-intensive computational 3D imaging systems for a broad range of applications, with interests spanning optical coherence tomography, diffraction tomography, lidar, ptychography, nonlinear microscopy, and parallelized 3D microscopy. Kevin is a recipient of the Barry M. Goldwater Scholarship, the NSF Graduate Research Fellowship, and the Schmidt Science Fellowship.