New simple virtual walking method – walking on the spot

In CAVE-like environments human locomotion is significantly restricted due to physical space and configural constraints. Interaction techniques based upon stepping in place have been suggested as a way to simulate long range locomotion. We describe a new method for step detection and estimation of forward walking speed and direction in an immersive virtual environment. To calibrate our system and to help in the modeling of the stepping behaviour, we collected motion capture data during real locomotion down a hallway while walking at different freely selected speeds, from very slow to very fast. From this data, the empirical relation between the forward speed of real walking and the trajectory of the leg motion during stepping was established. A simple model of stepping motion was fit for individual subjects. The model was used to estimate forward walking speed and direction from step characteristics during walking in place in a six-walled virtual environment. The system provides natural and effective simulated gait for interaction and travel within the virtual environment and provides the ability to study human locomotion and navigation in a CAVE-like environment.