Acquiring method and detecting method of live face head pose detection regression apparatus

A head posture and face detection technology, applied in the field of image recognition, can solve the problems of poor anti-attack ability, no interaction, and unpublished posture regression regressor, etc., to achieve the effect of increasing the success rate, speed and accuracy

Inactive Publication Date: 2017-01-25
XUZHOU NORMAL UNIVERSITY
View PDF5 Cites 25 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Moreover, most face detection schemes are based on direct extraction of face image information, without interactivity, and poor anti-attack capabilities, such as photos, videos, and model camouflage, which put forward requirements for live face detection. There are mature face detection methods, and there is no public regressor for pose regression in 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
  • Acquiring method and detecting method of live face head pose detection regression apparatus
  • Acquiring method and detecting method of live face head pose detection regression apparatus
  • Acquiring method and detecting method of live face head pose detection regression apparatus

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] Such as figure 1 Shown, for the acquisition method of a kind of live human face detection head posture regressor that provides in the embodiment of the present invention, assume in the present embodiment that the feature vector of each piece of face image, the coordinate vector are all 2N dimension (N is feature Points), the following are the specific implementation steps:

[0050] Acquisition of training data

[0051] In order to obtain a suitable head pose regressor, good training data is required. Since the regressor ultimately directly affects the accuracy of the head pose estimation, and the quality of the training data directly affects the accuracy of the regressor, this application adopts the following method Obtain high-quality image regressor training data. Since it is difficult to obtain the actual pose of a human face in a color RGB image in the actual implementation process, this application uses a 3D model of a human face as a training sample, and rando...

Embodiment 2

[0098] Such as Figure 4 Shown, be a kind of face recognition method based on the regressor described in claim 1 that provides in the embodiment of the present invention, assume that the eigenvector of each piece of face image, the coordinate vector are all 2N dimensions (N is feature points), the following are the specific implementation steps:

[0099] S401. Acquire an image of a facial gesture made by a user according to an instruction sent by a terminal.

[0100] The terminal sends out an instruction, which may require the user to open his mouth, blink his eyes, raise his head, lower his head, turn his head to the left or turn his head to the right, etc. The instruction may be a single instruction or any combination of the above instructions. The user makes corresponding gestures according to the instructions of the terminal. The terminal collects images of facial gestures made by the user through the camera. The camera continuously captures multiple frames of images. ...

Embodiment 3

[0128] Such as Figure 6 As shown, it is a face recognition method provided in the embodiment of the present invention, and the specific implementation steps include:

[0129] S601 and S602 are respectively the same as S401 and S402 in the second embodiment, and will not be repeated here. Further include:

[0130] S603. Perform linear transformation on the feature points on the target face image located in S602 and compare them with a fixed threshold, and then judge whether the action is completed.

[0131] Wherein, when the value obtained after the linear transformation is within the preset threshold, the judgment action is completed, and the face recognition is successful; otherwise, the face recognition fails.

[0132] For example, by measuring whether the distance between the feature points on both sides of the mouth of the target face image and the distance between the upper and lower feature points is less than a certain threshold, it is judged whether to open the mout...

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 embodiment of the invention discloses an acquiring method of a live face head pose detection regression apparatus. The method comprises the following steps of: A, acquiring three-dimensional model data of the face as training sampling sets; B, selecting face feature points in the three-dimensional model data; arbitrarily conducting random rotation and random translation exchange to the face feature points in the space by a camera imaging model and projecting the face feature points in a camera imaging plane to acquire two-dimensional coordinate plane sets composed of two-dimensional coordinates of the feature points, corresponding Euler angles of all two-dimensional coordinate planes in the two-dimensional coordinate plane sets, and a rotation matrix of corresponding relation between all two-dimensional coordinate planes and corresponding Euler angles thereof; C, according to two-dimensional coordinates of the feature points, the Euler angles and the rotation matrix, acquiring the head pose regression apparatus by a least square method. In the embodiment, the acquiring method is beneficial to improve face identification effectiveness.

Description

technical field [0001] The invention relates to image recognition technology, in particular to an acquisition method and a detection method of a head pose regressor for live face detection. Background technique [0002] With the advent of the era of big data, the problem of personal information security has become increasingly severe, and face recognition and detection technology based on image processing has been widely used. However, the current face detection technology is aimed at a small number of face images. With the deepening of the concept of big data, image big data processing will put forward higher requirements for face detection technology. Moreover, most face detection schemes are based on direct extraction of face image information, without interactivity, and poor anti-attack capabilities, such as photos, videos, and model camouflage, which put forward requirements for live face detection. There are mature face liveness detection methods, and there is no publ...

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): G06K9/00G06K9/62
CPCG06V40/161G06V40/172G06V40/67G06F18/214
Inventor 张艳程郑鑫
Owner XUZHOU NORMAL UNIVERSITY
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