Human body action recognition method based on muscle signals

Abstract

The invention discloses a human body action recognition method based on muscle signals. The method comprises the steps of obtaining depth image data corresponding to the muscle signals on the basis ofdepth signal flow, and distinguishing action information of muscle junction points in a human body in a three-dimensional space; and optimizing muscle junction signals of the human body on the basisof the action information of muscle junctions, filtering the muscle junction signals or redundant muscle junction signals which do not influence or slightly influence human body behavior recognition,building a human body behavior conditional random field model, obtaining a human body behavior recognition model, and predicting subsequent actions of the human body based on the human body behavior recognition model. According to the method, an improved sheep-flock optimization algorithm is used for performing selection. Compared with an unimproved method, the improved sheep-flock optimization algorithm has the advantages that a population is initialized by using a point set method, so that the algorithm is prevented from falling into local optimum; the convergence speed of the algorithm is increased by using an ordered subset method and introducing a sheep-flock wait-and-see method; and meanwhile, the recognition effect of video human body behaviors is improved.

Description

technical field [0001] The invention belongs to the technical field of pattern recognition, and in particular, the invention relates to a method for recognizing human body movements based on muscle signals. Background technique [0002] With the rapid development of the economy, the continuous improvement of people's living standards and the continuous innovation of computer science, control engineering, and rehabilitation medical technology, disabled people gradually realize that it is no longer far away to make up for their "incomplete" living conditions through artificial limbs. How to make prosthetics more practical and ideally respond to the physiological needs of the disabled has become a scientific research cause that benefits mankind. Intelligent bionic prosthetics emerged as the times require. Compared with ordinary prostheses, they can simulate the movement posture of ordinary people better, and have an incomparable positive impact on the physiology and psychology ...

AI Technical Summary

Problems solved by technology

[0005] However, in the existing technology, the ability to analyze the information sent by the muscle signal needs to be improved, and the data processing ability for the muscle signal is still low, and it is impossible to accurately and effectively identify the action represented by the muscle signal.

Images