Loss information recovery of direct unique identification on reflective surface

Lei Lei, Weiping He, Wei Zhang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In order to address the unreadable problem of direct unique identification on metal reflective surface during product real-time tracking, in this paper, two different reflective region images are captured for the same identification, the wavelet-based homomorphic filtering is applied to enhance the contrast of reflective region, the zoning Otsu thresholding is used to get the binary image, the characteristic matrix is extracted and the unique identification information is recovered by feature fusion. The method makes the module misreading rate of product identification fall to 0.7% from 22.91%, avoids the product tracking disconnection, achieves the loss information recovery of direct unique identification on reflective surface, promotes the in-depth application of direct part marking in product life cycle real-time tracking.

Original languageEnglish
Title of host publicationProceedings - 2010 3rd IEEE International Conference on Computer Science and Information Technology, ICCSIT 2010
Pages361-364
Number of pages4
DOIs
StatePublished - 2010
Event2010 3rd IEEE International Conference on Computer Science and Information Technology, ICCSIT 2010 - Chengdu, China
Duration: 9 Jul 201011 Jul 2010

Publication series

NameProceedings - 2010 3rd IEEE International Conference on Computer Science and Information Technology, ICCSIT 2010
Volume8

Conference

Conference2010 3rd IEEE International Conference on Computer Science and Information Technology, ICCSIT 2010
Country/TerritoryChina
CityChengdu
Period9/07/1011/07/10

Keywords

  • Direct unique identification
  • Loss information recovery
  • Metal reflective surface
  • Product real-time tracking

Fingerprint

Dive into the research topics of 'Loss information recovery of direct unique identification on reflective surface'. Together they form a unique fingerprint.

Cite this