The field of motion recognition systems (e.g. facial recognition) is continuously
developing. Included in this is hand-gesture recognition, through which we aim to
understand the articulated nature of the hand as a set of joint angles and positions.
Current methods involve visual markers such as painted gloves, stickers or LEDs
attached to the hand and observed by cameras, which fail to capture the complexity
of the articulations. This new method synchronizes magnetic sensors with camera
images and records matched image-articulation pairs aligned in both time and
position. The elastic, sensory finger loops used for this have minimal impact on finger
profile and do not impede natural motion, making this an accurate and efficient
method.
1586128594.png)