Human body lower limb motion detection and recognition method based on multi-source signals

Abstract

The invention discloses a human body lower limb motion detection and identification method based on multi-source signals. The human body lower limb motion detection and identification method comprises the following steps of obtaining human body lower limb motion information through a multi-source signal sensing technology detection method, and obtaining a human body lower limb motion state through the human body lower limb motion information; collecting the multi-source signals and carrying out data preprocessing to obtain input signals for identification; dividing the input signals into training data and test data, carrying out training and detection classification identification through an SVM classifier and a BP neural network method based on the SVM, and obtaining an identification result; and performing accuracy rate output optimization on the recognition result to obtain a final recognition result. According to the method, simultaneous acquisition of human lower limb attitude angle signals, plantar pressure signals and surface electromyographic signals of multiple muscles of legs is effectively realized, signal interference is removed, visual data is output to reflect a human lower limb motion state, and more accurate identification and prediction of human lower limb motion are realized.

Description

technical field [0001] The invention relates to the research field of multi-source signal collection, detection and recognition, in particular to a method for detection and recognition of human lower limb motion based on multi-source signals. Background technique [0002] In recent years, the problem of human-machine lower limb motion coordination has become one of the most important research issues in the field of robotics, especially in the application research of lower limb exoskeleton. It can not only provide theoretical guidance and Technical support, and it is the basis for the lower extremity exoskeleton to realize functions such as assisted walking, weight bearing, and rehabilitation training. [0003] In order to achieve the consistency and coordination of human-machine lower limb movement coordination, continuous detection and recognition of human movement is required. At present, the commonly used data recognition algorithms are support vector machine (SVM) and a...

AI Technical Summary

Problems solved by technology

[0006] However, in the current research and application of human gait recognition, image information, joint angles, plantar pressure or surface electromyography signa

Images