Data Glove Gesture Recognition With The RBF Neural Network Modified By LM And Genetic Algorithm

With the introduction of intelligent manufacturing 2025,industrial development is in the Fourth Era of industrial revolution dominated by intelligent technology,The so-called intelligent technology is human-computer interaction technology.At present,Scholars both at home and abroad are focusing on the information interaction technology of human and computer,and it becomes one of the most important research topics in the fourth wave of industrial revolution.The gesture recognition system has a very broad application prospect in the fields of industrial production,family entertainment,distance education and other fields.In this paper,the recognition accuracy and recognition rate of the gesture recognition system of 5DT data glove are studied.In order to improve the accuracy and real time of the gesture recognition system.LM algorithm and genetic algorithm are used to optimize RBF neural network,and use minimum error method to match gesture to prevent errors in gesture handover,the minimum error method is used to match the gestures so as to prevent the effect of the error during the handover.Firstly,the composition of the joints in the human hand is analyzed,and the distribution of the 5DT data glove sensor is combined to realize the mapping relationship between the joints of the human hand and the sensors,according to the characteristics of the hand joint distribution,the constraint conditions of the corresponding joints are put forward.Secondly,all kinds of gesture data are collected by Using 5DT data glove,the training sample library of gesture recognition and test samples are set up.In order to improve the existing defects in the application of RBF neural network,genetic algorithm and LM algorithm are proposed to alternately optimize RBF neural network.The simulation of RBF,GA-RBF and GL-RBF algorithm is realized by MATLAB function,at the same time,the performance of GL-RBF algorithm in gesture recognition is also verified.Finally,Open GL,3DSMax and MFC are used to draw the robot model.Based on the MFC technology,the 5DT data glove gesture recognition experiment platform is built.The remote robot manipulation of data glove is realized through the experimental platform,and the performance of humancomputer interaction is further improved.