Gesture recognition method based on plum group and long-short-term memory network

A long-short-term memory and gesture recognition technology, applied in character and pattern recognition, neural learning methods, biological neural network models, etc., can solve the problems of recognition limitations, overcome the interference of environmental factors, overcome spatial complexity, and improve The effect of accuracy

Pending Publication Date: 2020-09-25
CHONGQING UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the recognition method based on computer vision, due to the difference in human skin color and the influence of the recognition environment, the recognition has great limitations.

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
  • Gesture recognition method based on plum group and long-short-term memory network
  • Gesture recognition method based on plum group and long-short-term memory network
  • Gesture recognition method based on plum group and long-short-term memory network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0054] see Figure 1 to Figure 6 , a gesture recognition method based on Lie Group (Lie Group) and long-short-term memory network (LSTM), mainly comprises the following steps:

[0055] 1) Obtain the dynamic gesture skeleton video, and extract the hand skeleton image frame by frame. The device for obtaining dynamic gesture skeleton video is Intel depth camera RealSense.

[0056] 2) Preprocessing the hand skeleton image, the main steps are:

[0057] 2.1) Unify the number of hand bone images extracted from different dynamic gesture videos to ensure that the number of hand bone images in different dynamic gesture videos is consistent.

[0058] 2.2) Normalize the hand bone images to ensure that the hand bone sizes in all hand bone images are consistent.

[0059] 3) Extract the skeletal joint point data of the hand skeletal image, and put classification labels on it. Gestures are classified according to actions, mainly including left swing (the whole hand is waved to the left), ...

Embodiment 2

[0091] A kind of experiment that verifies the gesture recognition method based on Lie Group (Lie Group) and long-short-term memory network (LSTM), mainly comprises the following steps:

[0092] 1) Data acquisition, using Intel depth camera RealSense to extract hand bone joint point information, obtain gesture information, and preprocess the data;

[0093] Collect transactions through RealSense. The hand skeleton contains 21 joint points and 20 segments of bones, such as figure 2 shown. Five gestures were collected, including left swipe, right swipe, zoom in, zoom out, and open. Each gesture was repeated 20 times by 10 experimenters.

[0094] 2) Data preprocessing is to reduce the size of each data to ensure that the data size is consistent, and then normalize the data to ensure that the bone size in different samples is consistent. All data were normalized between 0 and 1 according to the following formula:

[0095]

[0096] in, Indicates the normalized data, x i Rep...

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 gesture recognition method based on a plum group and a long-short-term memory network. The gesture recognition method mainly comprises the following steps: 1) extracting a hand skeleton image; 2) preprocessing the hand skeleton image; 3) representing a three-dimensional geometrical relationship of rigid body transformation among skeletons in the dynamic gesture by utilizing the plum group data set S(t), and converting the plum group data set S(t) into corresponding plum algebra data s(t) through logarithm mapping; 4) training an LSTM neural network model; 5) obtaininga to-be-detected gesture skeleton image, and extracting the plum algebra data s'(t) of the to-be-detected gesture skeleton image, and inputting the plum algebra data s'(t) into the trained LSTM neuralnetwork model to realize gesture recognition. The recognition method makes full use of the advantages of deep learning, adapts to the kinematics characteristics of the human body, and improves the recognition accuracy.

Description

technical field [0001] The invention relates to the field of computer pattern recognition and human-computer interaction, in particular to a gesture recognition method based on Lie group and long-short-term memory network. Background technique [0002] With the rapid development of science and technology, more and more smart devices have entered social life. People hope that these devices can be more conveniently manipulated, and respond correctly to user instructions like humans, so as to achieve real intelligence, so it is more natural The human-computer interaction method has become an urgent need at present. [0003] In recent years, due to the rapid development of technologies such as computer vision, virtual reality, and smart wearable devices, research on gesture recognition technology that is closely related to them has gradually become popular. Gesture recognition is a topic of recognizing and classifying human gestures through related algorithms. Through the recog...

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
IPC IPC(8): G06K9/00G06K9/46G06K9/62G06N3/04G06N3/08
CPCG06N3/049G06N3/084G06V40/28G06V10/462G06N3/048G06N3/045G06F18/24
Inventor 刘礼李昕廖军
Owner CHONGQING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap