Abstract
The Leap Motion is a gesture recognition device. Introduced in 2013, it is the first compact device to recognize hand gestures.To critically analyze Leap Motion, three starting questions are:
- Meet user expectations for accuracy, reliability and robustness?
- What are the most promising applications of Leap Motion?
- What are the best practices in creating an application that integrates the Leap Motion?
To explore the subject, the gesture recognition domain is established: definitions, classifications, categorization of devices and types of recognition. The Leap Motion is presented as a gesture recognition device by the vision and its interest is demonstrated.
In order to answer the first question, recent scientific articles dealing with the qualities of the Leap Motion are analyzed. Based on this analysis, accuracy, robustness, pointing and interaction capabilities, and sensor recognition capabilities are evaluated.
To answer the second question, the observed applications of the Leap Motion are classified in the state of the art by theme, based on scientific articles. This classification then makes it possible to categorize it. Finally, a critical analysis identifies the most promising and suggests improvements.
API of Leap Motion is analyzed: its dynamic library, its classes, its six data types and their parameters.
Tests conducted using Java programs can evaluate its recognition capabilities.
Then, a use case of the application on a map is chosen. He composes this map in all four directions and then zooms in and out.
In order to identify the most intuitive gestures to perform the application operations, the behaviors of different users are collected during an investigation. The map is presented using a computer on which no program is applied. The expectations of gestural interaction are explained then the results are synthesized.
Based on the gestures most used during the investigation, a gesture interaction optimization software prototype is created. This prototype is presented to the users of the survey to test it. These tests check the satisfaction of the users and identify the elements to be improved regarding gestural interaction.
| Date of Award | 22 Jun 2018 |
|---|---|
| Original language | French |
| Awarding Institution |
|
| Supervisor | Bruno Dumas (Supervisor) |