Facial recognition tracking method and system

A portrait recognition and portrait technology, applied in the field of image recognition, can solve the problems of inaccurate and effective portrait tracking, re-identification loss, etc., and achieve the effect of fast and accurate identification and tracking

Pending Publication Date: 2022-06-21
北京影数科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But when faces are occluded or hidden when they are flipped, or when faces exhibit high variability due to different poses and orientations or lighting, shadowing or blurring effects, their re-identification will be lost, making accurate and efficient portrait tracking

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
  • Facial recognition tracking method and system
  • Facial recognition tracking method and system
  • Facial recognition tracking method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0052] figure 1 It is a schematic flowchart of a method for identifying and tracking a person according to an embodiment of the present application. see figure 1 It is known that a schematic flowchart of a method for identifying and tracking a person provided by an embodiment of the present application may at least include the following steps S101 to S104.

[0053] Step S101: acquiring all image frames in the video sequence to be identified;

[0054] Step S102: sequentially identify and extract at least one portrait and corresponding portrait information in all image frames;

[0055]Step S103: storing the portrait information that meets the preset conditions in a pre-created tracking database, and comparing the tracking database information based on the sequentially extracted portrait information to update the tracking database;

[0056] Step S104: Using the updated tracking database to realize the tracking of the person in the video sequence to be recognized.

[0057] Bas...

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 provides a portrait recognition tracking method and system, and the method comprises the steps: firstly obtaining all image frames in a to-be-recognized video sequence; sequentially identifying and extracting at least one human image and corresponding portrait information in all the image frames; and storing the portrait information meeting the preset condition to the pre-created tracking database, and comparing the sequentially extracted portrait information with the tracking database information to update the tracking database so as to realize portrait tracking in the to-be-identified video sequence, thereby realizing accurate multi-person identification in the unconstrained video sequence. And when the face is shielded or hidden, portrait tracking and recognition can still be successfully carried out.

Description

technical field [0001] The present application relates to the technical field of image recognition, and in particular, to a method and system for tracking person identification. Background technique [0002] Portrait tracking and recognition are mainly used in security monitoring. Traditional methods have been able to perform the correspondence between people detected along the frame based on hand-crafted functions. Now, with the development and extension of technology, a new method with a specific similarity index is proposed. A network flow approach to generate an appearance-based model for estimating small trajectory similarities between candidate objects. [0003] With recent years, deep learning techniques applied to facial recognition have led to improvements in multi-person re-identification performance, including deep face recognition methods that use convolutional neural networks and triple loss functions to recognize faces along frames; The end-to-end face detecti...

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): G06K9/62G06V10/774G06N3/04G06V20/40G06V40/16
CPCG06N3/045G06F18/214
Inventor 孙建飞
Owner 北京影数科技有限公司
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