Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Identifiability face pose recognition method based on local geometrical visual phrase description

A visual phrase and local geometry technology, applied in the field of pattern recognition, can solve the problems of unfavorable gesture classification accuracy and not considering word space information, etc.

Active Publication Date: 2013-09-18
XIDIAN UNIV
View PDF3 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, the bag-of-words model does not consider the spatial information of words, which is not conducive to the improvement of the accuracy of pose classification.

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
  • Identifiability face pose recognition method based on local geometrical visual phrase description
  • Identifiability face pose recognition method based on local geometrical visual phrase description
  • Identifiability face pose recognition method based on local geometrical visual phrase description

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] refer to figure 1 , the present invention is based on the discriminative human face gesture recognition method of local geometric visual phrase description and mainly comprises the following steps:

[0033] Step 1, SIFT feature extraction.

[0034] 1a) Construct the tower-shaped scale space: For the training image I, down-sampling is performed to obtain the tower-shaped structure of the image, and the image scale determines the number of layers of the tower-shaped structure (generally 3-5 layers), and each layer I of the image tower-shaped structure is b The two-dimensional difference of Gaussian scale space of (i, j) is defined as:

[0035] D(i,j,σ)=(G(i,j,sσ)-G(i,j,σ))*I b (i,j)

[0036] in is the scale-variable Gaussian function, * is the convolution operation, (i, j) is the spatial coordinate of the image, and σ corresponds to the scale of the Gaussian function, which determines the smoothness of the convolved image. s is a constant factor;

[0037] 1b) Detec...

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

An identifiability face pose recognition method based on local geometrical visual phrase description includes constructing a word bag model on the basis of face local features, introducing space information of words in the word bag model through local geometrical visual phrases, forming feature vectors of images through the geometrical visual phrases, counting the co-occurring feature quantity by aid of inner products of the vectors, forming a training image nuclear matrix through the co-occurring feature quantity, and inputting the training image nuclear matrix into a support vector machine classifier to obtain a face pose classifier through training. The identifiability face pose recognition method can overcome the influence of illumination, shielding and offset on face pose recognition and improves face pose feature identifiability; and improves the calculating efficiency by counting the quantity of co-occurring local geometrical visual phrases in a displacement space, and ensures displacement invariance of space features based on the local geometrical visual phrases.

Description

technical field [0001] The invention belongs to the technical field of pattern recognition, and in particular relates to a face posture recognition method, which can be used for posture feature extraction and classification when the human face rotates outside the plane. Background technique [0002] Face biometric recognition technology is widely used in daily life, and face pose estimation, as a key technology, has become a hot research topic in the field of pattern recognition and computer vision. However, the test results of the Facial Recognition Technology (FERET) program in the United States show that the recognition rate of the mainstream face recognition technology drops significantly when the posture changes drastically. The face pose classification and estimation technology can establish the corresponding relationship of face features under different poses, which helps to improve the face recognition rate. In addition, face pose estimation technology provides impo...

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): G06K9/00G06K9/64
Inventor 田春娜高新波陆阳王华青蒲倩李东阳王代富郑红张相南杨二昆
Owner XIDIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com