Research On The Function And Model Of Cerebellar Motion Control Based On 3D Gesture Tracking

One of the prerequisites for human-computer interaction is to achieve real-time tracking of 3D gestures,the high-dimensional structure of gestures,and the temporal and spatial coordination between arm motion and grasping actions are the main bottlenecks that affect real-time tracking.Therefore,how to effectively deal with the high-dimensional structure of gesture tracking and arm-extension and grasping the time-space coordination between the two issues is the research difficult and hot spots in the field.The research of these two hot spots can provide a new research topic or research idea for the application of cognitive psychology theory and method in computer theory.Firstly,hand tracking is introduced,including the basic concept,cognitive model,hand gesture tracking and some basic methods of dimensionality reduction and adjusting arm movement grasping movement between the temporal coordination of neural network model.Then,combined with the functional role of the cerebellum in some sports and cognitive tasks,as well as the existing functional model of the cerebellum,this paper discusses the dimensionality reduction effect of the real-time tracking of small and middle brain.On this basis,the particle filter algorithm is extended,and an adaptive control model with cerebellar prediction function is added to the particle filter algorithm.For simplicity of calculation and Realization of the cerebellar control model embedded in the "function",rather than complete on the neuroanatomy of cerebellar model.The simulation results show that the improved algorithm can improve the speed and accuracy of real-time tracking.This paper also studies the properties of cognitive psychology in arm movement by combining Anatomical structure and knowledge of neurophysiology,and aims at the time coordination controlling mechanism of arm extension and gripping in 3D gesture tracking.Based on the neural network model of classical arm extension and gripping time and space coordination,this paper first analyzes the visual channel of arm movement behavior,and constructs self-organizing neural network model and cerebellum controlling model adapting with visual coordination according to the characteristics of these channels,and then trains and simulates.The experimental results show that the proposed model can effectively improve the control mechanism and precision of time coordination between arm extension and gripping.