Low-level vision is the initial stage of visual information processing in biological and artificial systems, focusing on the automatic extraction of fundamental visual features such as edges, corners, motion, color, and texture directly from raw sensory input. This foundational processing is data-driven and serves as the essential precursor for subsequent higher-level visual tasks like object recognition and scene understanding.