Profile Face Recognition Method Based on Kernel Canonical Correlation Analysis

A technology of correlation analysis and face recognition, applied in character and pattern recognition, instruments, calculations, etc., can solve problems such as incompatibility, large amount of calculation, unavoidable, etc., and achieve the effect of improving recognition rate and rapid recognition

Active Publication Date: 2018-10-30
HENAN INST OF ENG
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although they can adapt to the diversity of gestures and have a high recognition rate, they still cannot avoid the disadvantages of large amount of calculation and the need for the cooperation of the recognized person.
Therefore, they are not suitable for the application in real-time monitoring that we are interested in. In practical applications, the image acquisition of the identified person is passive and cannot meet the application requirements of face recognition technology under current uncontrolled conditions.

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
  • Profile Face Recognition Method Based on Kernel Canonical Correlation Analysis
  • Profile Face Recognition Method Based on Kernel Canonical Correlation Analysis
  • Profile Face Recognition Method Based on Kernel Canonical Correlation Analysis

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0031] Example: see figure 1 , figure 2 , image 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7 , and the present invention will be described in detail in conjunction with the accompanying drawings.

[0032] Profile Face Recognition Based on Kernel Canonical Correlation Analysis The proposed method for profile face recognition relies on the idea of ​​learning a transformation that maps profile faces to their corresponding frontal faces using kernel canonical correlation analysis. Significantly improve the recognition rate by analyzing the non-linear nature of the mapping between object poses. This can be achieved by introducing appropriate kernel functions. The selection of the kernel is not straightforward, where some tests must be performed in order to select an appropriate kernel model and then tune its parameters.

[0033] In addition, faces are described using local binary model facial components instead of raw pixels: eyes, nose, mouth and chin. In the inv...

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 side face recognition method based on kernel specification correlation analysis. For the front part and side face part of the face image to be detected, manually annotate the eyes, nose, mouth and chin, and make different postures of the front face and the side face. Degree of geometric normalization processing, use the local binary mode facial components to describe the face, calculate 64 histograms for the divided regions in the local binary mode image, and then connect the histograms of the 14 regions of the face into a vector for description , after extracting the local binary feature vector, the dimensionality reduction operation is performed by applying canonical correlation analysis, and the nonlinear generalization based on kernel principal component analysis is introduced to consider the transformation nonlinearity between the side face image and the front face image, by Mapping from the original input space to the high-dimensional feature space to perform front or side transformation, and calculate the Euclidean distance to give the identity information of the side face image; effectively realize the fast recognition of the side face under unrestricted conditions.

Description

Technical field: [0001] The invention relates to a safety monitoring of public space, in particular to a side face recognition method based on kernel specification correlation analysis. Background technique: [0002] Biometric identification is an active research area and has applications in security monitoring of various public spaces. Numerous biometric identification technologies, such as fingerprints, iris and palm prints, have been widely used in security monitoring systems. These characteristics allow to uniquely describe a person, however, in the field of public safety, it is difficult to identify and determine the identity of a person through these characteristics in the video stream as "biological" characteristics. Since 1970, face recognition technology has gradually improved in theory and technology, but it is still subject to the diversity of individual samples, such as lighting conditions, shooting angles, facial expressions, age changes, etc. In non-intrusive...

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 & AuthorityPatents(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62
CPCG06V40/172G06V40/168G06V10/50G06F18/2135
Inventor栗科峰熊欣郑吉玉王俊华王炜郝原
OwnerHENAN INST OF ENG