Hierarchical Complex Activity Recognition Method Fused with Motion and Physiological Sensing Data

Abstract

The invention discloses a hierarchical complex activity recognition method which combines motion and physiological sensory data. Including: 1) using smart devices and wearable devices to collect motion and physiological sensing data; 2) extracting statistical features from motion sensing data, and extracting structural and transient features from physiological sensing data; 3) analyzing motion sensing data The statistical features of K-Means clustering and LDA subject extraction are carried out to obtain the subject distribution of motion sensing data; 4) on the basis of the subject distribution of motion sensing data and the characteristics of physiological sensing data, corresponding classifiers are established respectively; 5) Combine the outputs of the classifiers by fractional fusion to obtain a complex activity classification model. The invention uses clustering and topic models to represent the hierarchical structure of complex activities, and integrates motion and physiological sensor data to achieve accurate complex activity recognition, and has broad application prospects in the fields of smart home, medical care, and assistance for the elderly .

Description

technical field [0001] The invention relates to the fields of pattern recognition and pervasive computing, in particular to a hierarchical complex activity recognition method for fusing motion and physiological sensor data. Background technique [0002] With the development of smart devices (such as smartphones, smart watches, etc.) The acquisition of data has become increasingly convenient, and how to use these data for activity recognition has become the focus of the industry. The corresponding technology has broad application prospects in the fields of smart home, medical care, and assistance for the elderly. [0003] Simple activities usually consist of periodic movements or single postures of the human body, such as standing, sitting, walking, running, etc. Compared with simple activities, complex activities are usually less regular, last longer, and have high-level semantics, such as eating, working, shopping, etc. Traditional activity recognition methods based on sm...

AI Technical Summary

Problems solved by technology

However, due to the different characteristics of motion sensing data and physiological sensing data, motion sensing data mainly reflect transient changes, while physiological sensing data changes relatively slowly, and there are many kinds of complex activities, feature-level fusion cannot effectively integrate two different characteristics. class data for complex activity recognition

Images