Abstract

Learning a new language is always difficult. This is especially true for character-based languages like Japanese since each character also comes with a unique stroke order. To learn this stroke order teachers often make students draw these characters into the air.

To make memorizing this stroke order easier a framework is proposed in which the users can draw the different characters in front of their smartphones. The framework then gives feedback on the stroke order. Since nowadays the field of augmented reality is incredibly active and new improvements are made constantly, a lot of value is put into the framework being component based so improvements can be easily made alongside new technologies.

Implementation

The framework is implementet in Unity and uses Unity Colliders in combination with an event system to compare the drawn strokes.

ManoMotion is used for the hand and gesture detection. While the tracking isn't as good Leap Motion, ManoMotion works with only a RGB camera and therefore is usable with only a smartphone camera.

The Framework loads the Kanji from a XML file. Because of this it is easily expandable and also works with other languages while only changing the input. 

The following shows the finished framework, how it works behind the scene as well as an example of a saved Kanji.

Media

 Drawing.mp4LineDetection.mp4