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

Front-face-compensation-operator-based multi-pose human face recognition method

A face recognition and multi-pose technology, which is applied in character and pattern recognition, calculation, computer parts and other directions, can solve the difficult problems of multi-pose face recognition, achieve easy selection, improve calculation speed, and reduce calculation amount Effect

Inactive Publication Date: 2011-04-13
CHONGQING UNIV
View PDF5 Cites 23 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0014] The purpose of the present invention is to provide a simple and effective PCA-based multi-pose face recognition technology for the current difficult problem of multi-pose face recognition. The multi-pose face is compensated by the frontal face compensation operator, and the compensated face is used. method for face recognition

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
  • Front-face-compensation-operator-based multi-pose human face recognition method
  • Front-face-compensation-operator-based multi-pose human face recognition method
  • Front-face-compensation-operator-based multi-pose human face recognition method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] In order to better understand the present invention, specific embodiments of the present invention will be described in detail below.

[0033] The CMU PIE face database is one of the many face databases that fully considers the recognition of multi-pose faces, so the CMU PIE face database is used to explain the present invention in detail. The CMU PIE face library contains 68 people's profile faces with horizontal deflection angles of ±22.5, ±45 and ±67.5, faces with certain pitch angles, a total of 13 pose conditions, 43 kinds of lighting conditions and multiple poses under 4 kinds of expressions human face. Select the faces under various posture conditions and the corresponding frontal faces for normalization, and then take the average face to obtain the average face at each angle, and subtract the average frontal face from the average face at each pose to obtain each pose The front face compensation operator below. Now take the side face recognition with a horizontal...

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 simple and effective principal component analysis (PCA) algorithm-based multi-pose human face recognition technique, which compensating a multi-pose human face by using a front face compensation operator and using the compensated human face for multi-pose human face recognition. The attached figure in the abstract of the description is the whole multi-pose human face recognition flow chart. In the invention, when the PCA algorithm is used for human face recognition, the front face compensation operator compensates for front face profile information, namely feature face information corresponding to the large feature value resolved by the PCA algorithm, lacking in a multi-pose human face to be recognized, and reduces part of multi-pose human face information interfering with the PCA algorithm. A multi-pose human face usually lacks a front face profile which is more important information for PCA algorithm. Compared with the prior art, the technique has the advantages that: calculating by using an average face and avoiding using the method for training by forming a large matrix with human faces from a human face library, the calculation amount is reduced; thehuman face normalization requirement is low; it is easy to choose a human face library; and fewer faces are required to be trained. In addition, the algorithm used by the technique is simple, a good recognition effect can be achieved by simple addition and subtraction operation, and the technique can be used for recognizing human faces in all poses. In the invention, the method for compensating the multi-pose human face by using the front face compensation operator provides a new way for improving multi-pose human face recognition rate.

Description

technical field [0001] The present invention relates to a multi-pose face recognition technology based on PCA (Principal Component Analysis), more specifically, relates to a face compensation operator to compensate faces under various pose conditions, and use the compensated face method for face recognition. Background technique [0002] Multi-pose face recognition has been studied more in recent years, and it is known as one of the important problems that have not been well resolved in the field of computer vision and pattern recognition. Faces are easily affected by many factors, such as expressions, beards, glasses, hair, lighting, background, and the situation where only the side face is seen due to face deflection. The methods proposed for the hot issues of multi-pose face recognition can be roughly divided into There are three categories: conventional methods, three-dimensional research methods, and two-dimensional research methods. [0003] Conventional algorithms m...

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): G06K9/00
Inventor 谭晓衡张建慧陈林方杰周帅王保华
Owner CHONGQING 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