Page 65 - 2024F

58 UEC Int’l Mini-Conference No.53

Leap Motion and Hand Tracking

Technologies: Air­writing
Tiesheng YIN, C.K.CHOO
The University of Electro­Communications, Tokyo, Japan

0. Introduction 3. 3D Air­Writing[3]
This review research studied two papers
on Leap Motion and air writing, providing
insights into the capabilities of Leap
Motion, thereby laying a foundation for
future research.
1. Background

Leap Motion is a hand­tracking controller
that uses infrared emitters and cameras.

Fig 4. Working flow of the system

(1) Data acquisition
(2) Preprocessing
•Alignment
Fig 1. Leap Motion's internal structure(left)
and hand recognition results(right) [1] •Normalization
(3) Feature extraction
F = {P, D, C}
­ 3D Point Feature (P)
­ Writing Direction (D) Fig 5.
­ Curvature features (C) recognize
result
Data processing for word
2. 2D Air­Writing[2] recognition

4. Conclusion
Comparing the two studies, we can see
that the features of both approaches are:
(1) 2D Air­Writing
•Small amount of data and computation
•Simple to development
Fig 2. Leap Motion(left)
and Coordinate systems(right) •Existing tools can be used
(2) 3D Air­Writing
•More degrees of freedom in writing,
however, the data would be more
complecate to process.

Fig 3.writing
result Reference:
Use x, y coordinates for writing, and z [1]Bird, J.J. et al. Sensors 2020, 20, 5151.
coordinate for control [2] A. Morikawa, et al. Proceedings of the Companion of the 2017
The handwriting is presented based on ACM/IEEE International Conference on Human­Robot Interaction
the derived brush point p(UVW ). The (HRI '17). Association for Computing Machinery, New York, NY,
width d of the handwriting trajectory is USA, 215–216 (2017)
calculated from the height of brush point p. [3] Roy, P.P., et al. Multimed Tools Appl. 80, 11671–11689 (2021)