Face recognition method based on multi-task cascaded convolutional neural network

A convolutional neural network and facial recognition technology, applied in the field of facial recognition based on multi-task cascaded convolutional neural networks, can solve the problems of different shooting angles, difficult algorithm design, low accuracy, etc., and achieve strong migration ability. , good adaptability and improved accuracy

Inactive Publication Date: 2020-03-03
CHINA CHANGFENG SCI TECH IND GROUPCORP
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Because the images collected by cameras and cameras have the characteristics of different shooting angles and different head movements of the photographed targets, the algorithm design of face verification and face recognition is difficult and the accuracy rate is low.

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
  • Face recognition method based on multi-task cascaded convolutional neural network
  • Face recognition method based on multi-task cascaded convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0011] The present invention specifically comprises the following steps:

[0012] Step 1. Use the multi-task cascaded convolutional neural network to perform face detection on the original input video image and artificially photographed image, locate the area where the face is located, cut and zoom, and then obtain the eyes and mouth corners in the cut image of the same size The coordinates of five key points with the nose;

[0013] Step 2, average the five feature points in all clipped pictures respectively, and calculate the average face model;

[0014] Step 3: Perform affine transformation (Platts transformation) on the image to be aligned with the average face model obtained in step 2 as a standard, to obtain an aligned image. Let the matrix composed of five feature points of the image to be aligned be A(A∈R 5*2 ), the standard image is B(B∈R 5*2 ), the specific steps are:

[0015] Calculate the center positions (mean values) of the five feature points of the image to ...

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 relates to a face recognition method based on a multi-task cascaded convolutional neural network. Carrying out face detection on the original input video image and the manually photographed image by utilizing a multi-task cascaded convolutional neural network, positioning an area where a face is located, and carrying out clipping and zooming, thereby obtaining coordinates of five keypoints , which comprises two eyes, two mouth corners and a nose in the clipped image with the same size; respectively averaging the five feature points in all the clipped pictures, and calculating toobtain an average face model; carrying out Principal transformation on the to-be-aligned face images by taking the average face as a reference to obtain aligned images; and recognizing and verifyingthe aligned face images.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, and relates to a face recognition method, in particular to a face recognition method based on a multi-task cascaded convolutional neural network. Background technique [0002] Face verification and face recognition are research hotspots in the field of computer vision, and have a large number of application requirements in security, finance and other fields. The main idea of ​​face recognition is to locate the face by locating a series of feature points of the collected face image, and then convert the collected face into a standard face image through mathematical methods, and then exchange To verify and identify algorithms to judge. Because the images collected by cameras and cameras have the characteristics of different shooting angles and different head movements of the photographed targets, the algorithm design of face verification and face recognition is difficult and the a...

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
IPC IPC(8): G06K9/00G06N3/04
CPCG06V40/172G06N3/045
Inventor 武玉亭张晓林范宇
Owner CHINA CHANGFENG SCI TECH IND GROUPCORP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap