Why is captured motion not used directly without editing?

Raw capture has noise and gaps, is tied to the performer's exact proportions, and rarely matches the character or shot precisely, so it must be cleaned, retargeted, and often blended or edited before use.

What is motion retargeting?

It is adapting motion captured from one body to a character with different proportions while preserving important constraints such as keeping feet on the ground and hands on objects, so the motion still looks correct.

Character and Motion Capture

Motion capture records the movement of real performers and maps it onto digital characters, providing realistic motion that would be difficult to animate by hand.

Definition

Motion capture is the measurement of the movement of a person or object over time, and character animation from capture is the process of cleaning, retargeting, and reusing that data to drive a digital character.

Scope

This topic covers optical, inertial, and marker-based capture systems, the cleanup and retargeting of captured data to characters of different proportions, the blending and reuse of motion clips, and data-driven structures such as motion graphs that synthesize new sequences from a captured library.

Core questions

How is the movement of a performer measured accurately?
How is captured motion adapted to a character of different proportions?
How are clips of motion blended and concatenated smoothly?
How can new motion be synthesized from a captured library?

Key concepts

Optical and inertial capture
Marker tracking
Motion cleanup
Motion retargeting
Motion blending
Motion graphs

Key theories

Motion retargeting: Captured motion must be mapped to characters whose skeletons differ in size and proportion, requiring constraints such as foot contacts to be preserved so the retargeted motion stays physically and visually plausible.
Motion graphs: A library of captured clips is organized into a graph whose edges are smooth transitions, so that walking the graph synthesizes long, controllable streams of natural motion from a finite dataset.

Clinical relevance

Motion capture is standard in film and game character animation and virtual production, and the same technology supports clinical gait analysis, sports biomechanics, and rehabilitation assessment.

History

Rotoscoping and early electromechanical rigs gave way to optical marker systems in the 1990s; data-driven methods such as motion graphs in 2002 enabled flexible reuse, and markerless and learned capture are now advancing rapidly.

Key figures

Lucas Kovar
Michael Gleicher
Alberto Menache

Seminal works

kovar2002
menache2011

Frequently asked questions

Why is captured motion not used directly without editing?: Raw capture has noise and gaps, is tied to the performer's exact proportions, and rarely matches the character or shot precisely, so it must be cleaned, retargeted, and often blended or edited before use.
What is motion retargeting?: It is adapting motion captured from one body to a character with different proportions while preserving important constraints such as keeping feet on the ground and hands on objects, so the motion still looks correct.