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

A multi-class human motion recognition method and recognition system

A technology of human motion recognition and human motion, applied in character and pattern recognition, instruments, computer components, etc., can solve problems such as complex algorithm construction, achieve high recognition accuracy, increase data dimension, and protect user privacy

Pending Publication Date: 2019-04-26
NANJING NORMAL UNIVERSITY
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The more action categories that need to be recognized, the more complex the algorithm is to build

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
  • A multi-class human motion recognition method and recognition system
  • A multi-class human motion recognition method and recognition system
  • A multi-class human motion recognition method and recognition system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] The present invention will be further described below in conjunction with the accompanying drawings and examples.

[0035] The present invention provides a multi-category human action recognition method based on adjacent algorithm data processing and convolutional neural network, comprising the following steps:

[0036] S1, collect the nine-axis sensor data (including three-axis acceleration sensor, three-axis gyroscope sensor and three-axis magnetometer sensor) of the smart terminal device attached to the right wrist of the human body in the case of third-party supervision and recording, and pre-attach The action category label is used as a sample when the human action recognition model is trained;

[0037] S2. Use the adjacent algorithm to process the nine-axis sensor data, and adjust the data into the input format of the convolutional neural network. The final data format is (n, m, L, d), where n is the number of data, and m is The number of axes after data processi...

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 discloses a multi-class human body action recognition method which comprises the following steps: S1, collecting action data of various others, and attaching an action class label; S2, processing the data in the step S1 by adopting an adjacent algorithm, and dividing the data into a training sample and a test sample; S3, establishing a multi-layer convolutional neural network model,and importing a training sample to adjust parameters of the convolutional neural network model to obtain an optimal convolutional neural network model; S4, transplanting the trained convolutional neural network model to a mobile intelligent terminal; And S5, obtaining human body action data by using the mobile intelligent terminal, and inputting the human body action data into the trained convolutional neural network model to obtain a human body action recognition result. The method is high in identification precision and can identify more action types.

Description

technical field [0001] The invention belongs to the field of intelligent monitoring and recognition, and in particular relates to a multi-category human action recognition method and recognition system based on adjacent algorithm data processing and convolutional neural network. Background technique [0002] Human action recognition technology has a wide range of needs in monitoring, human-computer interaction, assistive technology, sign language, computational behavioral science and consumer behavior analysis, and the monitoring, identification and analysis of human action behavior. In general, motion recognition technology is divided into two types: image-based recognition and sensor-based recognition, and sensor-based recognition is widely favored by researchers due to the convenience of data collection and the protection of user privacy. [0003] Usually, in order to maintain a high recognition accuracy, multiple sensor devices are placed on multiple joints of the human ...

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): G06K9/62
CPCG06F18/2413G06F18/214
Inventor 王焜张雷王琦王震宇滕起毛进伟
Owner NANJING NORMAL UNIVERSITY