标题：Model and implementation of body movement recognition using Support Vector Machines and Finite State Machines with Cartesian coordinates input for gesture-based interaction
摘要：The growth in the use of gesture-based interaction in videogames has highlighted the potential for the use of such inter-action method for a wide range of applications. This paperpresents the implementation of an enhanced model for ges-ture recognition as input method for software applications.The model uses Support Vector Machines (SVM) and FiniteState Machines (FSM) and the implementation was basedon a KinectRdevice. The model uses data input based onCartesian coordinates. The use of Cartesian coordinates en-ables more .exibility to generalise the use of the model todifferent applications, when compared to related work en-countered in the literature based on accelerometer devicesfor data input. The results showed that the use of SVM andFSM with Cartesian coordinates as input for gesture-basedinteraction is very promising. The success rate in gesturerecognition was 98%, from a training corpus of 9 sets ob-tained by recording real users' gestures. A proof-of-conceptimplementation of the gesture recognition interaction wasperformed using the application Google EarthR. A prelim-inary acceptance evaluation with users indicated that the in-teraction with the system via the implementation reportedwas satisfactory
