Spherical images and video are widespread due to the popularization of capture and visualization devices. Unlike traditional pinhole-based, spherical media are defined on the surface of a sphere, have a full field of view (FoV), and store light intensities from an entire scene. Besides classical applications, those exploring immersive augmented, mixed, and virtual reality enormously benefit from full-FoV vision. Although spherical images lie on the sphere, they are often mapped to one or many planes. Those sphere-to-plane mappings lead to distortions, which may degrade the quality of traditional visual computing algorithms. This talk revises the spherical imaging model, standard capturing device types, and common representation formats. It also discusses the prominent challenges in spherical image processing and showcases the advances in selected applications.