Gesture recognition method and device based on feature pyramid and support vector machine

The invention discloses a gesture recognition method and device based on a feature pyramid and machine learning. According to the method, gesture data information in a data set is utilized, a gesture data area is screened out through a skin color model, HOG feature vectors are extracted, feature ext...

Full description

Saved in:
Bibliographic Details
Main Authors YANG YINYAN, WANG RUNCHEN, HONG HUI
Format Patent
LanguageChinese
English
Published 04.04.2023
Subjects
CALCULATING
COMPUTING
COUNTING
ELECTRIC DIGITAL DATA PROCESSING
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:The invention discloses a gesture recognition method and device based on a feature pyramid and machine learning. According to the method, gesture data information in a data set is utilized, a gesture data area is screened out through a skin color model, HOG feature vectors are extracted, feature extraction and feature dimension reduction are carried out on hand information through a principal component analysis method, and extracted gesture image data serve as input of a feature pyramid; and the feature vectors and the hand features are fused by establishing a network structure, and final fused feature information is obtained. And then taking the obtained fusion feature information and the corresponding posture label as input of SVM classifier features, and performing training to obtain a classification model. According to the method, gesture feature information is fused by establishing the feature pyramid, meanwhile, the recognition model is established by utilizing the SVM, effective features are extracted
Bibliography:Application Number: CN202211712050