2.2 Movement Recognition
Several types of motion capture systems are presented in the literature and some of them were instrument of interest in our work. The types of motion capture that we were interested are: Marker-based motion-capture, color markers and bare-hand tracking
Marker-based motion-capture: these systems require obtrusive retro-reflective markers and many camera setups Our pretension is to make use of one single camera to make less expensive for users, so we discarded this option.
Color markers: the usage of color markers are used in and presented good results. The work introduced a special glove design consisted in large color patches accounts for camera limitations to identify the movement. Our approach has no intended to make use of gloves because our proposition is to identify movements of the body, not only color marked regions.
Bare-hand tracking: edge detection and silhouettes are the most common features used to identify the pose of the hand. Our approach is intended to make use of this type of motion capture technique but we are going to make use of skeletonization applied in skin segmentation. the work shows a model-based hand tracking using a hierarchical Bayesian filter. It presents set of contributions in this type of tracking like: a hierarchical filtering algorithm, which combines robust detection with temporal filtering and the formulation of likelihood function that fuses shape and color information significantly improving the robustness of the tracker
However the work exposes a good model, it is restricted only for hand movements. Our objective is to identify not only hand movement but either other parts of the body like head and neck. That’s why were are going to use a skin segmentation technique and it is in development to present results in real-time, implemented in GPU, making use of a filter and thinning to handle body movements.
The next section presents our JECRIPE game which has the pretension to stimulate children with trisomy21