The Dragon Rider Interface was designed as part of the Human Computer Interaction bachelor project of University Hamburg.
The goal of this project was to develop new methods of moving in a virtual world by leaning. These methods are based on body tracking via Microsoft Kinect as well as weight distribution measured with the Wii Balance Board. Additionally, we wanted to design an application where these methods could be applied and evaluated against each other.
We decided on a Dragon Riding simulation in which the user is seated on a dragon and has to steer through a specific route.
The goal of this project was to develop new methods of moving in a virtual world by leaning. These methods are based on body tracking via Microsoft Kinect as well as weight distribution measured with the Wii Balance Board. Additionally, we wanted to design an application where these methods could be applied and evaluated against each other.
We decided on a Dragon Riding simulation in which the user is seated on a dragon and has to steer through a specific route.
To give context to this proposed mode of virtual travel we implemented an environment of tree-covered mountains and a route marked by checkpoints (an egg flying above a tower) that can be experienced by the user through an Oculus Rift HMD.
As for the input methods, we set up a seat with a balance board placed on top and a Kinect positioned in front. This way we were able to receive weight distribution and joint position data from the user.
After calibrating this data by calculating the amount of weight put in any direction as a fraction of the user’s whole eight (for the Wii Balance Board) or comparing the current leaning angle to a previously measured individual maximum leaning angle (for the Kinect) the user’s real movements were translated to the dragon’s movement in the virtual world.
Left and right leaning were mapped to a yaw and roll movement in the corresponding direction, leaning forward or backwards leads to a pitch movement of the dragon - forward leaning resulting in downward and backward leaning resulting in upward movement.
A gesture similar to the spurring of a horse’s reigns was implemented to enable a short speed boost.
As for the input methods, we set up a seat with a balance board placed on top and a Kinect positioned in front. This way we were able to receive weight distribution and joint position data from the user.
After calibrating this data by calculating the amount of weight put in any direction as a fraction of the user’s whole eight (for the Wii Balance Board) or comparing the current leaning angle to a previously measured individual maximum leaning angle (for the Kinect) the user’s real movements were translated to the dragon’s movement in the virtual world.
Left and right leaning were mapped to a yaw and roll movement in the corresponding direction, leaning forward or backwards leads to a pitch movement of the dragon - forward leaning resulting in downward and backward leaning resulting in upward movement.
A gesture similar to the spurring of a horse’s reigns was implemented to enable a short speed boost.
We conducted a study in which participants were asked to collect all eggs on the route by flying through them. All participants were instructed to fly the route twice, once with the Balance Board controls and once with the Kinect, alternating with which control they started.
We measured how long it took to collect all eggs in each trial and had our participants answer questionnaires about usability and motion sickness. The results showed that it took significantly longer to complete the trial with the Balance Board Controls accompanied worse ratings regarding intuitiveness, ease of use, feeling of control and fun in this condition.
Concluding, the Kinect shows far more promising results as an input device for leaning based controls, shown in the speed in which participants were able to complete the given task and also in the preference of the users. In further research it would be interesting to find out if the higher sensibility of the Balance Board could be used to advantageously if the users were given more information about the mechanics of the controls.
We measured how long it took to collect all eggs in each trial and had our participants answer questionnaires about usability and motion sickness. The results showed that it took significantly longer to complete the trial with the Balance Board Controls accompanied worse ratings regarding intuitiveness, ease of use, feeling of control and fun in this condition.
Concluding, the Kinect shows far more promising results as an input device for leaning based controls, shown in the speed in which participants were able to complete the given task and also in the preference of the users. In further research it would be interesting to find out if the higher sensibility of the Balance Board could be used to advantageously if the users were given more information about the mechanics of the controls.
The paper resulting from this project was presented at the GI VR/AR 2016 conference in Bielefeld.
