Gesture Estimation Based On Point Cloud Network And PSO Optimization Algorithm

In recent years,the pace of national development is stable,the material and spiritual life is extremely rich,and the achievements in the field of science and technology are frequent.With the popularization and development of deep learning and machine learning,gesture prediction research has gradually become a new star in the field of scientific research.Gesture estimation research has many similar fields,gesture recognition,fingertip diagnosis,gesture evaluation as the core technology of human-computer interaction,production and life in various fields widely used with outstanding application of virtual reality technology(vr)and augmented reality(ar),as a class input and output module,play an irreplaceable role.Nowadays,with the sudden emergence of gesture prediction technology,there are still some problems,such as declining accuracy,counting buoy and overfit.In order to remedy the above problems,this dissertation proposes a gesture estimation research scheme based on point cloud network and PSO optimization algorithm.Firstly,the modified YOLO frame’s hand area is used to test the model in the data collection module to decide whether to receive the important area of each data frame.Next,binocular camera and open source library PCL are used to convert the collected hand information into point cloud information.Then,the point cloud information is input into the PSO and multi-layer point cloud network feature fusion network model,and the feature fusion of information is carried out first,and then the particle swarm optimization algorithm(PSO)is used to select the best,so as to improve the accuracy of specific information.Finally,the obtained specific information of the hand is input into the simulation estimation model of MANO hand frame to simulate the specific state and position of the hand part.Through the above experimental process,the accuracy of its identification and simulation can be greatly improved,which will play a certain role in the development of subsequent hand research.