Unlock instant, AI-driven research and patent intelligence for your innovation.

Wi-Fi signal-based human body action layered analysis and recognition method and device

A technology of human action and recognition method, applied in the field of wireless perception, which can solve the problems of poor environmental adaptability and decreased accuracy of data-driven methods

Active Publication Date: 2020-01-07
BEIHANG UNIV
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Relatively speaking, the environmental adaptability of the data-driven method is poor, that is, its accuracy will drop significantly after the environment changes, while the model-driven method performs well in environmental adaptability

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Wi-Fi signal-based human body action layered analysis and recognition method and device
  • Wi-Fi signal-based human body action layered analysis and recognition method and device
  • Wi-Fi signal-based human body action layered analysis and recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0053] As mentioned above, the main methods of Wi-Fi-based human action recognition in the prior art can be divided into data-driven and model-driven. However, whether it is a data-driven approach or a model-driven approach, there are some issues that have not been considered. In the existing Wi-Fi-based human behavior recognition methods, the commonly used processes are: data collection, data processing, feature extraction, model training, action recognition and other steps. For the feature extraction and model training steps, the existing methods use all the extracted features to jointly train the model. That is, existing methods assume that all features play the same role in distinguishing different actions. But in fact, different features may play different roles in distinguishing different actions. Using the duration of an action as an example, it may be more effective to distinguish between walking and jumping, but jumping and sitting may not be so effective. Therefor...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a Wi-Fi signal-based human body action layered analysis and recognition method and a device. The method comprises the following steps: (A) carrying out preprocessing operation on an original signal to obtain a principal component analysis signal and a spectrogram; (B) designing an action hierarchical relationship according to action characteristics and designing corresponding characteristics; (C) training a plurality of classifiers by using the features in the step (B), and constructing a classification architecture according to the action hierarchical relationship; and(D) performing the steps A to C on the unknown signal to obtain a classification result, namely an action recognition result; meanwhile, the invention designs a Wi-Fi signal-based human body action layered analysis and recognition device, which comprises two computers which are respectively used as a transmitting end and a receiving end, a plurality of antennas, antenna extension lines and a mobile experiment table. The invention provides a complete process and method for performing human body action recognition by using Wi-Fi signals. The recognition system established by using the method canwell complete a human body action recognition task, and has robustness to environmental changes.

Description

technical field [0001] The present invention relates to a method and device for hierarchical analysis and recognition of human motions based on Wi-Fi signals, in particular to a method and device for hierarchical human motion classification based on motion hierarchical relationships, belonging to the field of wireless perception. Background technique [0002] Wi-Fi technology is a wireless local area network technology based on the IEEE 802.11 standard, and is a commonly used technology in the field of wireless communication. For Wi-Fi signals, it is essentially an electromagnetic wave, which has the conditions and potential to be used as a sensing device, so it is also used as a means of sensing the physical world. [0003] Human action recognition is an important topic in Wi-Fi perception. Its principle is to realize the recognition of human action by analyzing the different effects of different human actions on signal changes. The information used therein is Channel Stat...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62H04W4/30
CPCH04W4/30G06V40/20G06V10/462G06F18/2135G06F18/2411
Inventor 胡海苗姜永强李波蒲养林
Owner BEIHANG UNIV