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

A face recognition method based on multi-scale and multi-directional local binary patterns

A local binary mode and face recognition technology, applied in the field of image processing, can solve the problem of low recognition ability, achieve the effect of improving recognition ability, improving robustness, and speeding up recognition speed

Active Publication Date: 2020-04-07
XIDIAN UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to address the deficiencies of existing methods, and propose a face recognition method based on multi-scale and multi-directional local binary patterns, which not only describes more feature information of face images, but also retains the spatial information of images , which solves the problem of low recognition ability of the existing technology under complex lighting

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
  • A face recognition method based on multi-scale and multi-directional local binary patterns
  • A face recognition method based on multi-scale and multi-directional local binary patterns
  • A face recognition method based on multi-scale and multi-directional local binary patterns

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] refer to figure 1 , the specific implementation steps of the present invention are as follows:

[0034] (1) Face image preprocessing: Randomly extract C-type face image samples G, and cut each face image to a standard size as a training sample set X={X i ,P i}, where each type of face sample contains M Pi Vice, C represents the number of extracted face categories, G represents the total number of face image training samples, X i Indicates the i-th training sample, P i means X i category labels, M>0, G>i>0, C>P i >0; Randomly extract N face image samples and cut them to a standard size as a test sample set Y={Y j ,Q j}, where N represents the total number of face image test samples, Y j Denotes the jth test sample, Q j means Y j The category label of N>0, N>j>0, Q j to any value.

[0035] In this example, G pieces of C-type face images are randomly extracted from the standard face database. The face images in the standard face database are all of the same siz...

Embodiment 2

[0049] refer to figure 2 , the face recognition method based on multi-scale and multi-directional local binary patterns is the same as in embodiment 1, wherein as described in steps (2a) and (3a), by calculating training and testing samples W k The multi-scale and multi-directional difference relationship between each pixel point and its eight-directional pixel points, reconstructing training and testing samples All proceed as follows:

[0050] (Training samples X appearing in (2a) and (3a) i , test sample Y i Due to different representation methods, it is hereby uniformly expressed as face sample W k . )

[0051] Among them, W k Indicates face samples, W k (s,t) represents the face sample W k Any pixel in , s represents the abscissa of the pixel, and t represents the ordinate of the pixel.

[0052] a.1 as figure 2 , take a face sample W k The value W of any pixel in k (s, t), respectively, the multi-scale and multi-directional difference relationship between th...

Embodiment 3

[0071] refer to image 3 , the face recognition method based on multi-scale and multi-directional local binary patterns is the same as embodiment 1-example 2, the training and test samples to reconstruction described in steps (2b) and (3b) of the present invention Find the average value of the block by block, and concatenate the average value of each block into a row vector, which is used as the feature vector of the training and test samples Proceed as follows:

[0072] b.1 as image 3 , select reconstructed face samples sequentially from left to right and from top to bottom without overlapping The z*z pixels in constitute a pixel block Where z is an arbitrary constant, f is the number of the pixel block, and The z*z pixels in the pixel block are averaged to obtain the average value of each pixel in the neighborhood of the pixel block:

[0073]

[0074] in for pixel block The pixel value of any pixel within.

[0075] b.2 as image 3 , according to the order...

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 face recognition method based on a multi-scale and multi-directional local binary pattern, which mainly solves the problem of low recognition ability in the prior art under complex illumination. The realization steps are: select a face database, and divide training and Test sample set; use the difference between each pixel of the image and the pixel points in eight directions to reconstruct the face sample, and perform the block averaging operation to obtain the face feature vector; arrange the feature vectors in rows to get each sample Set the feature matrix; calculate the distance between the feature matrices, and get the recognition result. The present invention performs multi-direction and multi-scale weighting on the eight-direction features in each neighborhood of the face image to obtain reconstructed samples, and performs non-overlapping block, average, and series operations on them to obtain feature vectors, which increases the amount of pixel information and solves the problem of It solves the problem of high dimensionality of face samples and preserves spatial information. Improve the recognition rate of face recognition under complex lighting, which can be used for video surveillance and public safety.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to face recognition, in particular to a face recognition method based on multi-scale and multi-directional local binary patterns, which can be used for public security, video surveillance and access control. [0002] technical background [0003] Face recognition is a kind of biometric recognition technology based on human facial feature information, and it is a research hotspot in the fields of pattern recognition and computer recognition. The existing face recognition methods mainly include principal component analysis PCA method, linear discriminant analysis LDA method and local binary pattern LBP method. [0004] Both the PCA method and the LDA method are face recognition methods based on global image information. The purpose of these two face recognition methods is to extract the principal components of the image and reduce the dimension of the image. Howev...

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 Patents(China)
IPC IPC(8): G06K9/00
CPCG06V40/172G06V40/171
Inventor 刘靳陈月姬红兵孙胜男刘杨臧博
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