Multiple-Camera-Based Gesture Recognition by MDA Method
Cheng-Yueh Tsai, Yung-Hui Lee · 2008
In the present day, human-computer interaction is more important. This study proposes a new ideal for hand recognition model by combining VICON and multiple discriminant analysis (MDA). After building and testing, this model showed the recognition rates are 98.60% (by train set) and 82.64% (by test set) respectively. The average recognition rate is 90.62%. Besides, this research discussion the effect of subjects and eigen vectors on recognition rate of model.