Abstract: Currently, multi-touch gestures on the touch system lack of an ideal description and analysis. Research on Multi-touch gesture analysis and recognition algorithm presents a general multi-touch gesture analysis and design framework, rational analysis of high performance algorithms and optimizing multi-touch commands. Contact displacement and time to set the threshold to improve the dual function of identifying the accuracy of contact, to prevent the sudden increase in the noise of false judgments, slow down fast operation fitter; using RBF neural network model to solve the problem of dynamic gesture recognition, and the introduction of Euclidean distance-based clustering statistics as a function of the characteristic parameters of the network, greatly improving the multi-touch gesture recognition efficiency and accuracy.

Key words: Multi touch,Touch gesture,RBF neural networks,Euchdean distance,Dual-threshold