Method for collecting and optimizing face image sample based on heterogeneous active visual network

A sample collection and optimization method technology, applied in the field of face recognition, can solve the problems of insufficient resolution of face images, insufficient positive attitude angles, and insufficient attitude angles.

Inactive Publication Date: 2012-09-26
TSINGHUA UNIV
View PDF0 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At this time, the resolution of the collected face image samples is often not high enough; the attitude angle is not positive enough, or the different attitude angles are not rich
Such face image samples are not effective in face registration and 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
  • Method for collecting and optimizing face image sample based on heterogeneous active visual network
  • Method for collecting and optimizing face image sample based on heterogeneous active visual network
  • Method for collecting and optimizing face image sample based on heterogeneous active visual network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.

[0037] This embodiment is described in detail in three parts.

[0038] (1) Evaluation Criteria for Face Image Samples

[0039] Indicators of the impact of face image samples on face registration and face recognition, including image compression rate, target distance, target expression, etc. Referring to the research in the field of face recognition and the indicators that can be changed by active vision technology, this technology selects two indicators of resolution and attitude angle, that is, the resolution of the face area and the attitude angle of the face in the image. The face area is usually rectangular, such as 40×60 in size (the unit is wide pixel×long pixel), which can be described by the width pixel value r=40. The face pose angle p is described in terms of angles from -90° to +90° from left to right.

[0040] The empirical eval...

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 method for collecting and optimizing a face image sample based on a heterogeneous active visual network, wherein the optimizing process is as follows: in every Delta t time, all cameras are performed target distribution and state regulation to finish one round of target image sample collection; the process of optimizing variables in sequence is as follows: according to distribution relationship between the camera and the target, position and facing angle of the camera in a world coordinate system, and focal distance of the camera, maximizing total experience evaluation function of all targets as the optimization target and traversing target distribution plans; under each target distribution plan, adopting the genetic algorithm, solving optimization variables to enable the target evaluation function to be the maximum; selecting the best distribution plan so as to obtain the maximum total experience evaluation function. The method improves the resolution ratio and the attitude angle of the face in the image by the active visual technology, optimizes the sample collection process by the heterogeneous active visual network, obtains face image sample with good resolution ratio and rich attitude angles in order to register and identify face.

Description

technical field [0001] The invention belongs to the technical field of face recognition, in particular to a face image sample acquisition and optimization method based on a heterogeneous active vision network. Background technique [0002] Face recognition technology has good application value and application prospects in both civilian and military fields, mainly including face image sample collection, sample image preprocessing, classifier training (also known as face registration) and sample recognition (also known as human face recognition). Face recognition) These technical links, currently there are few studies on the sample collection link. [0003] In the application scenario of human-computer interaction, there are already some studies on the sample collection link. These technologies and research usually use the rotation and zoom of the camera installed on the robot, as well as the movement of the robot itself, to collect face image samples with appropriate resolut...

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 张涛李潇涵陈宋成宇陈学东孙昊李何羿
Owner TSINGHUA UNIV
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