Motion capture automated customized presets

Abstract

Motion Capture technologies transfer coordinate data from the human body movement and locomotion to a digital structure in order to move the avatar according to the actions performed by a person, creating digital animation function curves, marking as a keyframe each frame captured in a timeline. Those keyframes create excess of short movements as they try to correct the coordinates to a distinct virtual character from the real person who originated them, making the avatar quiver each time it performs any action. For animation purposes, in order to produce visually harmonic movements, it is necessary to remove manually the exceeded keyframes. The present study proposes an automated scripted method to reduce the amount of keyframes, keeping the shapes of the function curves, in order to customize the aesthetic gestural properties of characters animated by MoCap. It is presented a graphic comparison from before and after applying the automated customization proposed.