Intelligent Recognition Method of Low Dimensional Human Motion Data Boundary

An intelligent recognition method based on radial basis function (RBF) neural network for moving data of low dimensional human body is proposed. The median filter and a posteriori Wiener filter are used to filter the boundary of human moving data in order to improve the edge denoising intensity, and the wavelet transform is used to separate the human motion characteristics to obtain the separation matrix and independent components. The data feature extraction is completed to obtain the most suitable feature subspace for recognition. Through radial basis The feature subspace is recognized by the function neural network, and the intelligent recognition of the moving data boundary of the low-dimensional human body is completed. Experiments show that this method can simplify the extraction process of human motion data boundary and improve the accuracy of human motion data boundary recognition.