This is the final project (IPA) of our apprentice Cédric Girardin at the computer science department of the Berne University of Applied Science.
The VR-Application runs on a Oculus-Quest and uses its Hand tracking feature. The Application can validate up to 23 Hand signs from the German Alphabet.
This Project was made in 2 weeks/ 10 Work days: 30.03.2020-08.05.2020 (Yes, the documentation was also made within does 2 weeks).
The Readme file explains shortly folowing topics:
- Setup
- Documentation
- Plugins
- Handform validation
- Handdata
- Domainmodel
- Clone the unity project
git clone https://github.com/cpvrlab/vrTrainingFingerAlphabet.git- Open Unity Hub->Add, Select the Project where its cloned
- Select Unity Version (2019.3.06f) and open the project
The Documentation is in German and uses the HERMES 5.1 Project method.
└───Assets
│ └───Documents
│ │ IPA_CédricGirardin.pdf
The Unity Project uses folowing plugins:
| Name | Link |
|---|---|
| ui_ux_in_vr_2019_2 | https://github.com/soma1294/ui_ux_in_vr_2019_2 |
| Oculus Integration | https://assetstore.unity.com/packages/tools/integration/oculus-integration-82022 |
| JSON .NET For Unity | https://assetstore.unity.com/packages/tools/input-management/json-net-for-unity-11347#description |
| Monumental Journey | https://www.youtube.com/audiolibrary/music?nv=1 |
| All Sound Effects | https://bigsoundbank.com/ |
To validate a Handform, it loads a saved hand into the scene, and compares it values with the current hand.
The Validation compares folowing values:
| Handsubdivision | Value Type | BoneIds |
|---|---|---|
| Handorientation | Vector.up | Hand_WristRoot |
| Fingerangle | localEulerAngles Z | Intermediate BonesId’s and distal BoneId’s |
| Fingertipdistances | Diffrence of Vector.position | Hand_tipId’s |
The diffrences then gets weighted and it determents if its right or wrong. This process can be observed in realtime in the insight scene (insight the oculus quest).\
In the insight scene is a save button.
When pressed it creates a Handformstruct:
This Handformstruct will then be saved in the handdata.json file.
The Jsonfile will be converted to the values of a handformstruct...
