Motion capture, commonly abbreviated as mocap, is a technology and technique that records the movement of a person or object in physical space and translates that movement data into digital form for use in animation, visual effects, or other applications. By tracking the precise position and orientation of markers, sensors, or body landmarks over time, motion capture preserves the nuance, weight, and character of real human performance as digital motion data that can be applied to animated characters or virtual environments.

There are several approaches to motion capture. Optical systems use reflective markers placed on a performer's body that are tracked by multiple cameras, with software calculating 3D positions from the marker positions across frames. Inertial systems use sensors worn on the body that measure acceleration and rotation, producing motion data without requiring a camera setup. Markerless systems use computer vision to track body movement directly from video footage without physical markers, making capture more accessible and less intrusive. Facial capture records the nuanced movement of facial muscles, either through marker-based or markerless systems, to drive digital facial animation. The captured data is then applied to a digital character's skeleton, translating the performer's movement into the character's motion with all the organic timing and physical specificity that keyframe animation alone is difficult to achieve.

AI tools are increasingly relevant to motion capture workflows, both in enhancing markerless capture accuracy and in generating motion data synthetically from text or video descriptions. For AI video generation, understanding motion capture contextualizes why generated human movement can feel more or less natural - models trained on motion capture data have strong foundations for generating believable human motion.