Section: New Results

Serious Games Interfaces Using an RGB-D Camera : Results and Perspectives

Participants : Baptiste Fosty, François Brémond.

Keywords: RGB-D camera analysis, walking speed, serious games, startup project

Within the context of the development of serious games for people suffering from Alzheimer disease (Az@Game project), we have developed algorithms to interact with the virtual environment through simple gesture recognition and walking speed computation. We have shown in previous work that the walking speed measured by our system is accurate enough within this context and reproducible. A paper has been submitted in Gait and Posture journal (now in reviewing process).

Concerning the gesture recognition algorithm, it consists in recognizing three basic gestures (right arm left on the side, left arm left on the side, right or left arm left on top). We performed a small experimentation to test the robustness of the system in detecting these gestures where participants (10 in total) had to perform 10 times each gesture while walking at 2.5km/h on the treadmill (see Figure 23 ). The results are shown in Table 13 .

Figure 23. Display of gesture recognition and speed computation. Distance is red when right arm is left, speed is red when left arm is left, both are red when one arm is on top. Yellow dots correspond to feet and arm detection.