Face Recognition using Local Descriptors and Different Classification Schemas

• Author / Creator

  Liu,Ting

• There are two main activities in a face recognition practice: representation and classification. The main focus of this work is an analysis of image representation methods leading to better image classification scores. This study applies different feature descriptors and image segmentation techniques of image depiction, and investigates their influence on the classification results. We have proposed a number of single and ensemble classification approaches. For single classification approaches, we have considered different segmentation-based techniques of image processing, with weight-based strategies showing the most promising outcomes. In the case of ensemble-based classification algorithms we have investigated multiple criteria of importance focusing on ranking of candidates, as well as on segments and features sorted based on their prominence. We assessed and compared experimental results obtained for the FERET database. The most significant experimental results have been obtained for weighted-based strategy for single classification.

• Subjects / Keywords

  • Face recognition
  • Local descriptors
  • FERET
  • Ensemble classification

• Graduation date

  Spring 2013

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R34Q0B

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.